TU RBIN E
STEAM
PATH
M AINTENANCE AND REPAIR
Volume 2
William P. Sanders, P. Eng.
Library of Congress Cataloging-in-Publication Data
Sanders, William P.
Turbine Steam Path Maintenance and Repair
Volume Two / William P. Sanders, P.E.
p. cm.
q.cm
Includes index
ISBN 0-87814-788-8
Copyright © 2002 by
PennWell Corporation
1421 South Sheridan Road
Tulsa, OK 74112
800-752-9764
sales@pennwell.com
www.pennwell-store.com
www.pennwell.com
Cover and book design by Robin Remaley
All rights reserved. No part of this book may be reproduced, stored in a
retrieval system, or transcribed in any form or by any means, electronic
or mechanical including photocopying or recording, without the prior
written permission of the publisher.
Printed in the United States of America
1 2 3 4 5
06 05 04 03 02
Turbine Steam Path Maintenance and Repair—Volume Two
PREFACE
The Turbine Steam Path, Damage,
Deterioration, and Corrective Options
This book has been prepared for those technical people responsible for the operation and maintenance of steam turbines.
Steam turbines represent a complex technology for units commonly designed to operate hundreds of thousands of hours while
being subjected to a severe environment and a variety of operating
phenomena capable of degrading their condition. These units are
required to continually operate in a reliable, safe, and cost-effective
manner. Under such circumstances, these units cannot maintain
their original design-specified level of performance indefinitely. All
units will deteriorate with age. Owners anticipate this, and designers
will normally leave an adequate margin, knowing that some level of
such deterioration is tolerable.
The technology of steam turbines—while mature—continues to
evolve. More accurate and time-responsive diagnostic tools and
techniques are becoming available to assist in predicting when a
unit has deteriorated to the extent that corrective action is required.
Similarly, tools are available to assist the operator in analyzing problems and determining the effective corrective action best suited to
the condition causing deterioration. The improved understanding of
unit condition and rates of deterioration now achieved, together
with advances in materials, should allow units to be maintained in a
manner that will help minimize maintenance concerns and costs.
It is the premise of this book that units “as supplied” will fulfill
two basic requirements:
xii
Preface
•
It is assumed the unit “as designed” represents an optimum
selection of component sizing and arrangement
•
It is assumed the unit “as delivered” meets design specification within the range of tolerances provided by the design
engineer, i.e., unit components have been manufactured,
assembled, tested, and installed in such a way that they are
in compliance with the original design specification
The implication of this second assumption is that if nonconforming situations or conditions arose during the total manufacturing
process (and exist within the unit), they have been evaluated by a
competent design authority in the engineering organization of the
manufacturing company and have been assessed as not having an
adverse impact on the potential performance of the unit.
In terms of turbine unit components, “design optimum” is a difficult term to define. The entire design process is one of compromise
by the designer who wants a unit to be both efficient and reliable.
These requirements often represent competing demands, forcing the
designer to select from among various elements, possibly electing to
downgrade one aspect of these requirements to meet the demands
of the other. This is done consciously and with detailed evaluation
to provide a balanced selection.
Units delivered by a manufacturer represent the supply of elements that conform to the design principles established by his or her
design function, and conform with the best technology available to
that supplier at the time the design specification was prepared.
However, the operator must recognize that the labor and material
costs involved in building a steam turbine are high, and turbine suppliers must be able to produce units at competitive levels sufficient
to allow them to achieve a profit margin enabling them to sustain
business as well as finance further development.
xiii
Turbine Steam Path Maintenance and Repair—Volume Two
Many power systems are currently experiencing significant
changes in how they operate. Pressures from deregulation, environmental concerns and legislation, and an aging fleet of power generating equipment are shifting emphasis from the installation of new
capacity to the maintenance and care of the old. There is a continuing increase in demand for electric powe,r but new capacity installation is not keeping up with it. Operators of turbine generators are
therefore required to meet this demand with their existing fleets—
aging units requiring greater care to reduce the possibility of forced
outages. The prospect exists of units experiencing extended outages
as damage is found at planned outages.
Historically, as units have aged they have tended to be used less
frequently. They are initially placed on spinning reserve and ultimately placed in reserve, mothballed, or retired—their capacity
replaced with newer, more efficient units. An advantage of this
dwindling reserve is that older units have continued to operate at
high load factors and therefore become less susceptible to the rigors
of start-up, shut-down, and the associated thermal transients.
Unfortunately, there have also been fewer opportunities for plant
maintenance to proceed with the maintenance outages required to
maintain unit operational health.
Maintenance problems associated with keeping aging units available are only going to increase. Operators who are expected to provide power on demand are going to experience even greater future
challenges of damage and deterioration. They will be expected to
identify not only the damage, but also the causative effects, and then
find immediate solutions that will not jeopardize system security.
This book examines the damage, deterioration, and failure mechanisms occurring with unfortunate consequences—on some units,
with monotonous regularity—within the turbine steam path. These
various forms of degradation can be the result of a number of factors
related to conditions often beyond the control of operating and maintenance personnel. However, even if the steam turbine is operated
xiv
Preface
precisely as intended by design, and suffers no external degrading
effects for its entire operating life, the steam environment is one that
can cause components to suffer various forms of distress. Under normal circumstances, the design process selects and defines individual
components suitable for the design operating life of the unit (normally about 200,000 hours). At a mean load factor of about 75%, this
represents a 30-year operating life.
A number of unavoidable influences affect the operating life of
the various components comprising the turbine. These include the
steam environment itself, the stresses induced in the components by
rotation, and stresses induced in various portions of the unit by
expansion of the steam through the blade passages. There are also
the effects of the high-pressure steam, causing high-pressure drops
across some components that must be contained by the casings.
External factors that can affect the reliability of components of
the steam path and act to lower the expected operating life include
the possible formation of corrosive elements at various locations
within the steam cycle or impurities gaining access from in-leakage
at sub-atmospheric pressures. There can be unit trips caused by a
number of circumstances, from system trip electrical faults to lightning strikes on power lines. Many of these factors, while possibly
occurring in a 30-year operating life, cannot be anticipated in terms
of when, where, how many, or how severe their effects might be.
The damage and deterioration that occur within the steam path
can be of several forms. It can result in a gradual material loss—the
growth of a crack—or an immediate failure causing a forced outage.
Gradual deterioration can (depending upon type and location) be
monitored and replacement parts made available, or corrective
action taken to rectify the situation before it extends to an unacceptable degree. Immediate failure is most often the consequence of
either mechanical rupture or the presence in the steam path of some
foreign object, either generated within or having gained access from
some external source (including “drop-ins”).
xv
Turbine Steam Path Maintenance and Repair—Volume Two
In writing this book, I have tried to present information that plant
personnel will be able to use to make value judgments on the type
and severity of any damage, suggest possible causes, and then consider the most appropriate corrective actions that are available. To
aid in the recognition and classifying of operational damage and
deterioration, photographs are used to illustrate unacceptable or
suspect conditions.
Many of the damaging phenomena considered in these chapters
do not occur in isolation. It is possible that several can and will
occur simultaneously, demonstrating that components are subjected to more than one degrading influence. A condition may initiate
due to one damaging mechanism introducing a condition of weakness, which then allows another mechanism to become predominant and drive a component to failure. This situation often occurs
even though the driving mechanism would not have been capable
of causing failure had not the weakness been introduced by the first,
or initiating, mechanism.
Before considering degradation and failure in any detail, it is
important to define what constitutes failure and/or deterioration.
An important consideration in any case of evaluation and condition assessment of a turbine is establishing what constitutes failure.
The definition I find most acceptable is this: A condition exists
within the unit that while it would not prevent the unit from returning to service and continuing to develop power, it could force it
from service before the next planned outage. Various other definitions exist, and the definition of failure used in any situation—and
therefore the responsibility for correction—can be controversial.
This controversy is to some extent aggravated by possibilities; e.g.,
a crack that has been determined to exist may be predicted by the
methods of fracture mechanics to be growing at a rate that would
not cause complete rupture, forcing the unit from service before
the next planned outage.
xvi
Preface
As reserve power margins diminish, steam turbines—that currently have operating periods between major maintenance outages
of three to eight years—could be forced to operate longer than
intended when they were originally returned to service. Under these
circumstances, it is difficult when making a prediction of a unit’s
future operation to be certain there will not be some major change
in its operating parameters. Parameters that can influence an acceptable definition of failure in any situation include the exact operating
period, the unit load pattern, and the steam conditions the unit will
experience over a number of years.
A simple and conservative solution to this definition of failure
would be to change any suspect component showing any crack or
unacceptable damage-or-deformation indication. This may appear
to be an expensive option, but is considerably less expensive than
a forced outage requiring weeks or months to open, repair, await
replacement parts, replace those parts, close the unit, and return it
to service.
Defining efficiency deterioration is somewhat easier. It is even
possible to quantify such deterioration in terms of reducing steam
path efficiency and unit output. What is not possible to determine is
the extent of any mechanical deterioration that may occur and cause
efficiency deterioration. This is an unknown situation not recognized
until complete mechanical rupture occurs. There is normally no
manner to predict such an occurrence—damage could be in the
incubation phase—even when an examination of the steam path is
made at maintenance outages.
During operation, certain situations and phenomena are known
to occur that have the potential to initiate damage or to cause deterioration in performance. These damaging and deteriorating phenomena can be of a continuous or intermittent nature, produced as
a consequence of transient operating or steam conditions. Such phenomena can also be the result of sudden mechanical failures of
components that cause more extensive consequential damage. The
xvii
Turbine Steam Path Maintenance and Repair—Volume Two
most commonly occurring of these degrading effects are related to
the formation of moisture in the steam path or solid foreign particles,
possibly from the boiler or scale generated within the superheater
and reheater tubes. Other sources include chemical contaminants
that are introduced or gain access to the steam path on which they
are deposited, and possibly act as corrosive elements. The other
principal degrading conditions are the operational phenomena
occurring during the operating life of the unit.
The first two chapters of volume one provide general information.
The first outlines what is considered necessary to define and constitute a maintenance strategy that represents management’s commitment to maintaining a healthy system. This chapter also outlines
means of monitoring conditions indicative of damage. The second
chapter deals with the spatial arrangement within the steam path and
the factors that affect it. This is important because the performance
(efficiency and reliability) of a turbine is influenced considerably by
the alignment of the unit and the resulting axial and radial clearances
and “laps” that are achieved in the hot operating condition.
Chapters 3, 4, 5, and 6 discuss the various phenomena known to
affect both the efficiency and structural integrity of the components.
In the second volume, chapters 7, 8, and 9 consider repair and refurbishment options currently available. Fortunately, there are ever-present advances in these technologies, and as experience is gained,
newer and improved methods develop to be applied to older units so
they can continue to operate with high levels of availability—often
with improved efficiency. Chapter 10 considers seal systems and
gland rings, and provides means of estimating the financial penalties
associated with excessive leakage. Seals are one area where operators and maintenance personnel can influence the cost of power generation and help reduce the cost of power to their customers.
The final two chapters, 11 and 12, relate to quality and the
inspection of elements being manufactured to replace damaged
components. This is an area where many engineers feel the cost of
xviii
Preface
undertaking such inspections is difficult to justify. However, what
happens when components—manufactured when they are required
in an emergency to return a unit to service—have any form of fault
and force the unit from service prematurely? In such a case, the cost
of inspection—ensuring that a supplier’s quality program is prepared
and operating properly—is well justified. It is often said, “There isn’t
time and money to do it right, but there is always time and money
to correct it.” This statement is well applied to the manufacture or
repair of components in an emergency, because the cost of a second
outage is just as high as the first, and far more embarrassing.
Because the steam turbine is a thermal machine designed to convert thermal energy to rotation kinetic energy, I have included an
appendix that provides the basic thermal relationships required to
understand the turbine and its operation.
Situation evaluation
The more susceptible areas in any turbine unit are a function of
many complex factors—individual stress levels, stress concentration,
mode of operation, and the operating environment. Individual components are also greatly influenced by the expertise with which the
parts were designed, manufactured, and assembled, and the operating transients to which they have been subjected. The diversity of
the factors that can contribute to damage precludes any generalization of cause or value. Steam path components are subjected to high
stress, both direct and alternating. Many parts operate at high temperatures and are of complex forms interacting with one another in
unpredictable ways. These factors, when combined with load and
temperature transients that occur during operation, combine to
make the steam path highly sensitive and a major source of concern
to the designer and operator.
While some concerns are common to most operators, the type of
deterioration or damage to which any component or area is subjected
xix
Turbine Steam Path Maintenance and Repair—Volume Two
normally varies from unit to unit. This accounts for the variety of
concerns expressed by maintenance staff and the different dispositions of the various nonconforming conditions that will be developed in any situation.
In many instances when corrective action is required, there is no
optimum solution that can be followed without deviation. Operation
and load demands will often negate the optimum. At other times,
costs, special tools, skills, and the availability of replacement parts
could require some form of compromise. These compromise solutions may have to be adopted from necessity, but the final disposition should provide the best balance between cost, risk, and the
immediacy of returning the unit to service.
The logical approach to maintenance and repair dispositions is:
•
Consider the available alternatives in terms of the original
design requirements of the affected components
•
Evaluate possible solutions in terms of departure from the
design-specified requirements
Many “repair” or “accept-as-is” dispositions will have only a limited effect on unit performance and can be readily accepted. Other
repairs can be proposed and accepted, representing a compromised
condition. Such options should only be accepted on the basis that
the unit will be operated with this compromised solution for as short
a period as possible, and that the selected option does not represent
a significant level of risk in the short term. If this is possible, plans
should be put into effect immediately to develop an acceptable solution that can be undertaken within a reasonable time.
The maintenance options
The satisfactory performance of a steam turbine is influenced considerably by the manner and expertise with which it is maintained and
xx
Preface
the load patterns it follows. While the plant operating engineer can
control, to a large degree, the maintenance of the units for which he
is responsible, he is unfortunately unable to exercise little influence
on operating patterns. This is a responsibility of dispatchers who have
a mandate to serve the demands of their clients rather than the turbine
generators of their system.
For maintenance to be cost-effective, it must be planned. When
signs of distress, excessive wear, misalignment, or component deterioration are detected, the need for corrective action must be considered. These corrective actions should help ensure the situation
does not deteriorate further, to the extent the unit is placed on a
forced outage status, severely load limited, or suffers an unacceptably high degree of deterioration in efficiency.
There are general maintenance requirements for any unit.
Guidance for these is provided by the designer and should be followed for all routine matters. The designer will also provide recommendations for the operating time between opening sections of the
unit for periodic maintenance and examination. During these maintenance outages, any findings that could affect unit performance
must be reviewed in relation to their possible long-term effects.
Maintenance actions
Opening a unit for maintenance provides the opportunity to
make repairs or to install replacement parts when the necessary skills
and special-purpose tools are available. Such an opening also allows
replacement parts to be ordered, which can be placed in the unit at
the current or later outage, depending upon the delivery and
required period of the outage. Replacement is made when an evaluation of any found operational nonconformance is judged to be
placing the unit at risk if returned to service without correction. A
detailed evaluation of each nonconformance should be made, and it
should indicate if, and what, actions are required.
xxi
Turbine Steam Path Maintenance and Repair—Volume Two
The principal purpose of a steam turbine maintenance inspection is to detect potential problems at an early stage. If this is not
done, relatively minor situations could progress to the extent a
forced outage or excessive loss in unit output and efficiency could
occur. During such a maintenance inspection outage, parts can be
examined visually for indications of failure, wear, or distortion. Also,
non-destructive tests can be applied to critical components to determine if their ability to continue to perform satisfactorily has deteriorated and, if so, what remedial action should be taken or planned.
A nonconformance in any part of the steam turbine unit is considered to have occurred when there are signs of mechanical failure,
excessive wear, or any form of deterioration that has the potential to
adversely affect the performance of the unit. Such nonconformances
must be reviewed for short- and long-term effects.
As soon as unit inspection indicates that a nonconforming condition has been found, it must be evaluated. The logic process of
evaluation for both availability and efficiency is considered in chapter 1. This chapter outlines avenues the maintenance engineer
should explore in deciding what corrective action needs to be taken.
There are four decisions that can be reached. In some circumstances
the decision is relatively simple, and is in fact obvious. In other situations, a decision is made based on the probability of failure, the
possible cost of repair, and ultimately, the reparation of consequential damages that are the result of not taking corrective action. These
four options can be considered:
xxii
•
scrap and replace
•
repair
•
rework
•
accept-as-is
Preface
Of these decisions, possibly the most difficult and potentially
most controversial is the latter—accept as is—a disposition that
allows a component to return to service with no effort made to correct the nonconforming condition. There are two reasons for reaching and deciding upon this course of action:
•
There is little need to make any corrections. To make them
will add no or marginal improvement to unit performance,
and the condition will not place the unit at risk
•
The cost of replacing, repairing, or reworking cannot be justified. This is often a judgment call on the part of the engineer and can only be made if he or she is aware of any risks
involved
Such a decision should not be made as a desperation measure.
The risks, if any, should be fully evaluated. The options and the probability of failure—from an extended outage to operation—must be
fully considered.
Therefore, the evaluation process can be a complex one.
Occasionally, the solution is self-evident—such as when partial failure has occurred or when excessive damage exists. The most difficult decisions are those related to suspected damage or deterioration, and those for which it is difficult to determine the cause. In
these instances of uncertainty, mature judgment is required, together with knowledge of the operating and maintenance history of the
unit. This knowledge should help in the evaluation. The information
in this book can also provide confidence in the selection of the final
disposition.
The availability of replacement parts, special skills, and tools
will often influence which decision is reached. Care must be exercised to ensure that availability or non-availability of replacement
parts does not force the owner/operator into a decision ultimately
xxiii
Turbine Steam Path Maintenance and Repair—Volume Two
causing more expense and increasing the overall risk level to an
unacceptable degree.
Often, alternatives to these potential solutions are available.
Some may degrade a unit’s rating or impose other restrictions in
terms of maximum output or the time for which a unit can be operated. The compromise correction is ultimately more acceptable over
the short-term, while the owner/operator arranges for a more palatable long-term solution.
William P. Sanders
Richmond Hill, Ontario, Canada
August, 1999
xxiv
TABLE OF CONTENTS
List of Acronyms . . . . . . . . . . . . . . . . . . . . . .ix
Foreword . . . . . . . . . . . . . . . . . . . . . . . . . . . .x
Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . .xii
Acknowledgements . . . . . . . . . . . . . . . . . .xxv
Chapter 7—Operating Damage Mechanisms and
Refurbishment Techniques for
Stationary Components
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1
Stationary Blade Row Geometry . . . . . . . . . . . . . . . . . . . .15
Operating Phenomena Affecting the
Stationary Blade System . . . . . . . . . . . . . . . . . . . . . . .36
Diaphragm Vane Repair Methods . . . . . . . . . . . . . . . . . . .59
Determination of Stage Discharge Area and Angle . . . . . . .88
The Computation of Adjustments . . . . . . . . . . . . . . . . . . .96
Diaphragm Thermal Distortion . . . . . . . . . . . . . . . . . . . .121
Repair Methods for the Diaphragm Sidewalls . . . . . . . . . .133
Correction of the Diaphragm Inner Web . . . . . . . . . . . . .141
Damage to the Outer Rings . . . . . . . . . . . . . . . . . . . . . . .148
Weld Repair of the Horizontal Joints . . . . . . . . . . . . . . .153
Stationary Blade Damage . . . . . . . . . . . . . . . . . . . . . . . .154
Components of the Casings . . . . . . . . . . . . . . . . . . . . . . .162
Casing Operating Problems and Repair Methods . . . . . . .174
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .199
v
Turbine Steam Path Maintenance and Repair—Volume Two
Chapter 8—Refurbishment Techniques for
Rotating Blades
Introduction . . . . . . . . . . . . . . . . . . . . .
Steam Path Cleaning . . . . . . . . . . . . . . .
Blade Inlet Edge Erosion Damage . . . . .
Moment Weighing of Refurbished Blades
Erosion Shield Cracks . . . . . . . . . . . . . .
Blade Trailing-Edge Erosion . . . . . . . . . .
Solid-Particle Erosion by Oxide Scale . .
Erosion Resistant Coatings . . . . . . . . . . .
Solid-Particle Peening . . . . . . . . . . . . . .
Massive Particle Damage . . . . . . . . . . .
Corrosion Effects . . . . . . . . . . . . . . . . .
Rotating-Blade Refurbishment . . . . . . . .
Water Induction . . . . . . . . . . . . . . . . . .
Fretting Corrosion . . . . . . . . . . . . . . . . .
References . . . . . . . . . . . . . . . . . . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
. . . . . .201
. . . . . .204
. . . . . .206
. . . . . .242
. . . . . .254
. . . . . .255
. . . . . .257
. . . . . .264
. . . . . .271
. . . . . .273
. . . . . .281
. . . . . .291
. . . . . .303
. . . . . .304
. . . . . .305
Chapter 9—Damage Mechanisms Arising from
Operation and Refurbishing Techniques for
Rotating Components
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .309
The Rotating Components . . . . . . . . . . . . . . . . . . . . . . . .311
Coverband Damage, Repair, and Refurbishment Methods .362
Tie Wires Damage, Repair, and Refurbishment Methods . .410
Fusion Techniques for Rotating Blades and
Stage Hardware . . . . . . . . . . . . . . . . . . . . . . . . . . . .427
Common Rotor Damage Mechanisms . . . . . . . . . . . . . . .433
Bends Induced in the Turbine Rotor . . . . . . . . . . . . . . . . .464
Blade Root Steeples and the Wheel Rim . . . . . . . . . . . . .480
Corrective Action for Rotor Rim Damage . . . . . . . . . . . . .498
Rotor Weld Repair . . . . . . . . . . . . . . . . . . . . . . . . . . . . .508
Considerations of the Weld Repair Process . . . . . . . . . . .527
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .544
vi
Table of Contents
Chapter 10—Seals, Glands, and Sealing Systems
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .549
Steam Path Seals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .551
Functions of the Steam-Sealing System . . . . . . . . . . . . . .557
Steam Leakage Through Labyrinth Seals . . . . . . . . . . . . .559
Quantifying Labyrinth Leakage
(Applying the Method of Martin) . . . . . . . . . . . . . . . .567
The Economics of Seal Maintenance . . . . . . . . . . . . . . . .595
Forms of the Seal Knife Edge Discharge Coefficients . . . . .604
Form of the Gland Rings . . . . . . . . . . . . . . . . . . . . . . . . .618
Forms of the Seal Strip and its Trimming . . . . . . . . . . . . .627
Insertion and Securing of Seal Strips . . . . . . . . . . . . . . . .630
Seal Strip and Gland Ring Materials . . . . . . . . . . . . . . . .641
Gland System Operating Problems . . . . . . . . . . . . . . . . .643
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .651
Chapter 11—Quality Assurance for Replacement
and Refurbished Steam-Turbine Components
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .653
Responsibility for Quality . . . . . . . . . . . . . . . . . . . . . . . .657
Definition of Quality . . . . . . . . . . . . . . . . . . . . . . . . . . .658
Definitions of Performance . . . . . . . . . . . . . . . . . . . . . . .660
The Design Specification . . . . . . . . . . . . . . . . . . . . . . . .663
Reverse Engineering . . . . . . . . . . . . . . . . . . . . . . . . . . . .666
The Quality Assurance Program . . . . . . . . . . . . . . . . . . .677
The QA Manual . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .679
The Engineering Review . . . . . . . . . . . . . . . . . . . . . . . . .680
The Responsibility and Administration of a QA Program . .683
The Inspection and Test Plan . . . . . . . . . . . . . . . . . . . . . .688
Purchaser Assurance of Quality . . . . . . . . . . . . . . . . . . .689
Product Surveillance . . . . . . . . . . . . . . . . . . . . . . . . . . . .690
Nonconforming Situations . . . . . . . . . . . . . . . . . . . . . . . .699
Available QA Program . . . . . . . . . . . . . . . . . . . . . . . . . .704
The Machining of Turbine Components . . . . . . . . . . . . . .705
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .710
vii
Turbine Steam Path Maintenance and Repair—Volume Two
Chapter 12—The Manufacture and Inspection
Requirements of Steam Turbine Blades
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .711
Radial Alignment of Rotating Blades . . . . . . . . . . . . . . . .712
Blade Manufacturing Techniques . . . . . . . . . . . . . . . . . . .723
The Blade Manufacturing Processes . . . . . . . . . . . . . . . . .726
Profile and Cascade Tolerances . . . . . . . . . . . . . . . . . . . .749
Profile and Placement Errors . . . . . . . . . . . . . . . . . . . . . .765
Passage Swallowing Capacity . . . . . . . . . . . . . . . . . . . . .770
Special Processes Applied to the Vane . . . . . . . . . . . . . . .774
Blade Root Tolerances . . . . . . . . . . . . . . . . . . . . . . . . . .775
Factors Influencing Blade Pitch Errors . . . . . . . . . . . . . . .803
Requirements to Accommodate Stage Hardware . . . . . . .812
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .813
Appendix—Thermodynamics and the Mollier
Enthalpy-Entropy Diagram for Water/Steam
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
The Physical Properties of Water/Steam . . . . . . . . .
The Gas Equations . . . . . . . . . . . . . . . . . . . . . . . .
The Heating and Expansion of Steam . . . . . . . . . . .
The Entropy of Steam . . . . . . . . . . . . . . . . . . . . . .
Reversibility . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Steam Properties and Diagrammatic Representation
The Basic Power Cycles . . . . . . . . . . . . . . . . . . . .
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
viii
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
. . .815
. . .818
. . .831
. . .833
. . .852
. . .857
. . .859
. . .873
. . .883
Chapter
7
Operating Damage
Mechanisms and
Refurbishment Techniques
for Stationary Components
INTRODUCTION
The stationary components of the steam path are not subject to the
same level of stress as a consequence of rotation and centrifugal loading. However, they can still be in a high temperature/high pressure
environment and will therefore be subject to loads sufficient to affect
their operating life. In addition, the alignment these components are
able to maintain relative to the rotating components during their operating life can be affected by steam conditions and various operating
characteristics. When some form of deterioration is found in stationary components (and possibly progressing to an unacceptable level),
it is necessary to evaluate the situation and to take corrective action.
To undertake such correction, procedures must be developed.
1
Turbine Steam Path Maintenance and Repair—Volume Two
It is necessary to develop and assess suitable procedures for both
monitoring and correction so that at each outage the various affected components can be examined, critical dimensions recorded, and
a non-destructive examination (NDE) of critical regions undertaken.
These condition reviews should be an integral part of all outages.
Any nonconforming condition must be monitored so the condition
can be corrected when deterioration has occurred to the extent the
unit cannot be returned to service without continuing to operate at
risk, or with a significant reduction in the operating efficiency.
Two major stationary components possess conditions that must
be examined because corrective action is most often necessary—the
casing (including the gland seal housings) and the diaphragms, or
stationary blades. (These stationary vanes are often referred to as partitions.) Both components can be subject to pressure, temperature
differentials, and transients sufficient to cause distortion. In addition,
these components will have steady stresses developed in them as a
consequence of pressure differentials.
The engineer responsible for turbine maintenance must establish
programs for monitoring stationary components. With stationary
components there is unlikely to be dramatic failures similar to those
associated with rotating components. However, the consequence of
stationary component deterioration can be just as damaging in terms
of forcing the unit from service and the costs and delays of correcting
the situation. In determining what should be monitored, the equipment manufacturer will define the basic requirements of the individual components. However, these can be summarized as follows.
Diaphragms are designed to carry and locate the stationary blade
rows within the steam path (see chapter 2 in vol. I). Diaphragms can
have large pressure differentials developed across them. They also
develop a torque on the individual vanes that will rotate them in their
circumferential locating slot if not constrained by suitable keys and
pins at the horizontal and vertical positions. Diaphragms and stationary blade rows operate at high temperatures for considerable
2
Operating Damage Mechanisms & Refurbishment Techniques for Stationary Components
periods and are supported and located in the casing only at their
outer diameters. Diaphragms and stationary blade rows also have a
strength, or support, discontinuity at the horizontal joint.
Note: Diaphragms and stationary blades perform the same function. By definition, diaphragms comprise an outer ring that locates in
the casing, a stationary blade row, and an inner ring or web. The stationary blade row normally comprises individual blades that locate
in the casing or blade carrier directly.
The diaphragms are normally used throughout the steam path of
an impulse design unit, whereas the stationary blade rows are normally inserted into the casings of the high and intermediate pressure
sections of reaction designed units. The basic design of these two
stages was considered in chapter 2.
The most commonly encountered (and most damaging) mechanisms in diaphragms and stationary blades are the following:
•
Compounds deposited on surfaces of steam path components that are normally inert may be composed of materials
that can become corrosive under the appropriate environmental circumstances
•
Material loss associated with solid-particle erosion caused by
exfoliated boiler scale
•
Damage caused by impacts with solid particles, either carried into the turbine or (more likely) originating within the
blade system, steam chests, and/or valves
•
Damage induced by water formed by condensation and then
accumulated into larger droplets and deposited upon the surfaces of the steam path components. This water forms a surface
layer, flows through the steam path, and will cause various
forms of material loss
3
Turbine Steam Path Maintenance and Repair—Volume Two
•
Distortion that causes the diaphragm to modify from a circular to an elliptical form. This adjustment may cause the horizontal joint to either open or close. This effect, if observed,
is monitored by measuring the diaphragm bore diameters at
the horizontal and vertical centerlines
•
The high-pressure differential that exists across the vanes and
inner web can cause an elastic deformation of the diaphragm.
At high temperatures there can also be a plastic deformation,
which will cause a dishing. This is normally checked by suitable “drop checks”
The casing is the main structural component of the turbine and
contains the rotating components. The casing also contains and provides alignment for the stationary steam path components. The following common deteriorating mechanisms can affect the turbine
inner and outer casings as a consequence of the combined effects of
temperature and pressure:
4
•
The casing can distort from the true circular form. The casing
must maintain its circular form along the length of the axis.
Distortion in the horizontal or vertical direction can affect
concentricity of the stationary blade rows. This distortion is
more significant if the stationary blades are mounted directly into the casing. In the instance of a diaphragm type construction, casing distortion may not necessarily affect the
steam path concentricity. However, it could affect the vertical and horizontal position of the diaphragm
•
There is often a tendency for a casing to “hump” and assume
an “upward bow” in the cold condition. The casing must
remain flat at its horizontal joint
•
The horizontal joint must remain closed and provide an
effective seal against steam leakage that would allow the
steam to bypass the stages or even complete sections
Operating Damage Mechanisms & Refurbishment Techniques for Stationary Components
•
As a consequence of large temperature swings, the casing
will be subject to the effects of “low-cycle fatigue.” This will
most often introduce cracks into the high temperature
regions at small fillet radii, where considerable stress can
concentrate as these temperature changes occur
Lower stress levels in the stationary components lend themselves
more readily to refurbishment. Many of the deteriorating situations
that are encountered are readily correctable.
Stationary blade definitions
Chapter 2 provided definitions for the various portions of the
rotating blades. In this chapter, repair techniques are discussed for
the damage and repair of the stationary blade rows. It is therefore
appropriate that similar definitions be provided for the stationary
components. The names given to different portions of the steam path
elements differ from manufacturer to manufacturer, which makes it
difficult to be consistent, and can cause some level of confusion
when describing various aspects of both damage and repair. In this
text, the following definitions will be used:
Diaphragms
Diaphragms are manufactured by a number of processes, and
have the primary function of expanding the steam and guiding it into
the following row of rotating blades. In its simplest form, it comprises an outer ring, a row of stationary blades, and an inner ring
designed to provide a pressure barrier between the stationary blades
and the rotor. The major components [Fig. 7.1.1(a)] follow:
An outer ring. This ring is located in the casing or blade carrier.
Horizontal joint keys and crush pins are used to hold the diaphragm
in both vertical and horizontal alignment to the rotor and achieve
the alignment requirements outlined in chapter 2.
5
Turbine Steam Path Maintenance and Repair—Volume Two
Root
Attachment
Outer
Ring or Ledge
Outer
Ring
Profile
(b)
Tip
Seals
Vane or
Partition
Tip
Diameter 'Dt'
Profile
Inner Band
or Ring
Skirt
Web or
Inner Ring
Root
Diameter 'Dr'
(c)
(a)
Shaft
Seals
Shaft
Seals
Fig. 7.1.1—Definitions of the stationary blade row components.
K
L
w
ψ2
Br
Bs
ψ1
D
D
Dp
R
Outer
ring
s
s
r
r
Fig. 7.1.2—A water catcher half produced as part of
the outer ring on a diaphragm.
6
Vane or
Partition
(d)
Operating Damage Mechanisms & Refurbishment Techniques for Stationary Components
This outer ring can also be extended to provide radial seals
above the rotating blade row [Fig. 7.1.1(b)]. It is also possible (in the
water region) that a portion of the water catcher can be built into the
outer ring (Fig. 7.1.2).
The vane. The vane expands the steam and directs it into the following rotating blade row. As in the rotating blades, the profile forms
the pressure surface of one expansion passage and the suction surface of the adjacent passage. Vanes are subject to the same requirements of dimensional control and surface finish as the rotating elements. The definitions used to describe the vane profile are shown in
Figure 7.1.3.
In many stages of the high and intermediate (reheat) pressure
expansions, the radial height of the vane is not sufficient to justify
other than a cylindrical vane of constant profile. Other, longer stages
have a profile that “varies” with radial height, although on many
stages the profile remains the same but of reducing chord (Fig. 7.1.4).
Inlet
nose
Chord
Axial
width
Pressure
face
Suction
face
Discharge
tail
Fig. 7.1.3—Definitions of portions of the vane profile.
7
Turbine Steam Path Maintenance and Repair—Volume Two
Fig. 7.1.4—A stationary vane of constant profile, but
with a reducing chord.
The means of attachment of the vane to the outer and inner sidewalls is dependent upon the method of manufacture. This vane is
often referred to as the partition.
To remove as much kinetic energy as possible from the expanding steam, the previous rotating blade row would have been
designed to discharge the steam as close to axial as possible. Therefore, the profile of the stationary blades is normally axial—i.e., the
inlet angle “α0” is equal to 90 degrees.
An inner ring or web. The inner web provides the pressure barrier across the stationary row. This web is also designed to carry radial seals at its inner diameter to minimize the quantity of steam that
leaks past the stationary blade rows. Similarly, this web will often
have a radial seal produced just below the root diameter “Dr” [Fig.
7.1.1(c)]. These strips are intended to limit the amount of steam that
re-enters the main steam flow, which can introduce efficiency losses. This seal helps force the steam to flow through pressure balance
holes produced in the wheels of the rotor.
8
Operating Damage Mechanisms & Refurbishment Techniques for Stationary Components
The seal system produced between the inner surface or the inner
web and the rotor are designed to minimize the leakage quantity that
bypasses the stationary blade row, as this steam does not expand
through the blade rows and therefore produces no power.
Stationary blades. The stationary blades are mounted directly
into the casing or blade carrier and perform the same function as
the stationary blade row of the diaphragm. Figure 7.1.1(d) shows a
single blade element that has the same radial height as the
diaphragm vane of Figure 7.1.1(a). The tip diameter “Dt” and the
root diameter “Dr” are the same on both steam path elements.
This design is normally used on the high and intermediate pressure rows of a reaction machine in which there tends to be limited
axial space and therefore the root block performs no function other
than to secure the blade in its carrier. Caulking is also normally used
to secure the blade. The inner band or ring can be integrally produced with the vane and root portion, or it can be attached by riveting using the same methods as on the rotating blade coverbands.
This inner ring will also normally carry radial seals to minimize
steam leakage.
Nozzle plate. In addition to the diaphragm and stationary
blades, the nozzle box is that portion of the main unit structure into
which the steam enters the steam path via inlet pipes and valves.
After entering the steam chamber, the steam flows around an inner
belt (or a portion of a belt) and enters the first row of stationary
blades. This row of stationary blades is contained in a structure
called the nozzle plate.
There are various methods of forming this steam chamber and
nozzle plate. The methods used depend upon the inlet steam conditions and the method of controlling the unit (either throttle- or nozzle-controlled). In a throttle-controlled unit, the basic inlet is a continuous belt and steam flows around the complete 360 degrees. In
9
Turbine Steam Path Maintenance and Repair—Volume Two
the nozzle-controlled unit, individual portions of the inlet are connected to individual stop and control valves.
These inlet belts can form a portion of the casing (Fig. 7.1.5) or
a self-contained chamber (Fig. 7.1.6), which are assembled separately, aligned, and welded into the casing.
Fig. 7.1.5—Steam admission nozzle boxes. With
this design the nozzle boxes are an integral part of
the casing.
The nozzle plate is the structure that carries the first row of stationary blades. This plate is designed to receive steam as it enters the
high-pressure section delivered from the boiler. Steam enters from the
main inlet pipes, passes through the main stop and control valves,
and enters an inlet belt that passes around the rotor. This chamber is
10
Operating Damage Mechanisms & Refurbishment Techniques for Stationary Components
Fig. 7.1.6—A nozzle chamber with segmented steam-admission ports.
The inlet pipes are connected to the nozzle block, which is a selfcontained forging.
constructed in such a manner that no inner web is required on the
nozzle plate. Therefore, the nozzle plate becomes a device attached
to the steam chamber at both its inner and outer diameters.
Depending upon its design, a 360-degree nozzle plate may span
the upper and lower halves of the inlet and discontinue only at the
horizontal joints. In smaller rated units, the nozzle plate may span
11
Turbine Steam Path Maintenance and Repair—Volume Two
only the single (normally the lower) 180-degree half of the inlet. In
some nozzle-controlled unit designs, it may also be produced as a
number of individual segments, each admitting steam from a single
valve and individual chest.
When steam enters the main steam inlet belt and flows around
the inlet perimeters (the annulus arc fed by an inlet pipe), it enters
the first row of nozzles or stationary blades. This stage is normally
designed for a high-pressure drop and there is, therefore, a considerable temperature and pressure gradient across it. Basic considerations biasing the designer to select this high-pressure drop for this
stage include the following:
12
•
This stage should not experience leakage past the blade row
because of its method of attachment to the nozzle box. This
means a large pressure drop can be accommodated without
increasing the leakage potential
•
If a large pressure drop is designed into this first stage, and
the steam is contained within a separate nozzle box, temperatures and pressures to which the casing is exposed are
reduced to conditions at the first stationary row discharge.
This reduces the pressure/temperature duty on the casing
•
These stages are designed with large diameters and generally possess a velocity ratio of blade velocity/steam velocity that results in a large pressure drop. If the first stage is a
two-row wheel (a Curtis stage), the velocity ratio will be of
the order 0.25-0.32. Therefore, the enthalpy drop and the
pressure and temperature drop will be large, reducing the
duty on the casing considerably. This will also require
fewer stages (after the control stage) to complete steam
expansion
Operating Damage Mechanisms & Refurbishment Techniques for Stationary Components
Casing definitions
The turbine casing—essentially a cylindrical vessel—is the main
stationary component of each turbine section. It encloses the rotating portions of the unit and locates the stationary blades either
directly or through the location and support of an inner casing,
which itself carries the stationary blades and/or diaphragms.
The principle component of the casing are the shells, which provide
the mechanical strength and carries and locates other elements such as
packing heads, diaphragms, and the inner casing or blade carriers.
The casing is normally split along its centerline at the horizontal
joint. This is to facilitate assembly and provides access to the rotor
and internal stationary portions of the unit. The shell halves are connected through a bolted flange at the horizontal joint. It acts to contain the steam and maintain its connection to the steam path blades.
Casings may also provide locations for internal-gland packings or
portions of the steam seal system. They could be equipped with
internal moisture collection and drainage systems if moisture is present in the steam. In the case of minor failures, the high-pressure
shells should also be capable of containing missiles generated from
the rotor.
Both the upper- and lower-half casings can be arranged to provide connections for welded stub pipes. External pipes are connected to these stubs, allowing steam to be extracted for regenerative
feed heating or other cycle or process use. Such steam is extracted
from the main steam flow. Other pipes, used to introduce or extract
steam to other parts of the cycle, may also penetrate the casing. Usually, pipes connected to the upper-half casings are joined through
flanges or other devices. This allows quick disassembly at outages
and reconnection without the use of any form of heating or metalfusion techniques.
13
Turbine Steam Path Maintenance and Repair—Volume Two
Special provisions are necessary in the casing to admit the
high-pressure, high-temperature steam and to make provision for
the differential expansion that occurs between the various portions of the shells. Such differential expansion occurs because of
different temperatures (temperature gradient) along the axial
length of the casing, and also because of the different rates at
which the various parts of the turbine heat and cool with main
steam temperature changes (see chapter 2, vol. I ). For doubleshell construction, it is necessary for the main inlet pipes to pass
through the outer casing and introduce steam to the main steam
inlet belt or nozzle box.
High-pressure casings are normally supported at each end
through arms that are produced integrally with and extend from the
casing to pedestals located adjacent to and between the casings or
sections. Transverse and axial keys are used to maintain alignment of
the shells at these pedestals. Usually such keys have been hardened
by nitriding and are located on the bottom vertical centerline to
ensure correct alignment is maintained at all loads and during transient operating conditions.
Low-pressure casings are designed to both contain the steam
and also to minimize the “inward leakage” of air when exhaust
pressure is sub-atmospheric. Low-pressure turbine exhaust creates large volumetric flow, so these low-pressure casings are usually produced by fabrication; because such fabrications are not
structurally strong, it becomes necessary to support them for their
entire perimeter at their horizontal joint or at a similar location
below this joint.
Turbine casings comprise a number of individual components
which, when assembled, allow the unit to operate safely, achieving
high levels of reliability and efficiency. Principal among these components are the following:
14
Operating Damage Mechanisms & Refurbishment Techniques for Stationary Components
•
Shells—the main structural components produced by casting
or fabrication—in some designs, by a combination of both—
dependent upon the experience and preference of the
designer
•
Shaft end packing head—is attached to the shells and carries
the gland rings, which are located where the rotor passes
through the shells. Gland rings minimize the outward leakage of the steam or the inward leakage of air
•
Inlet section—the inlet to the steam path must be designed to
allow free access of the inlet pipes as they transport steam to the
nozzle box. They must also minimize steam leakage that will
occur at those locations and must be designed to permit movement between the inlet pipes and the main body of the shells
•
Explosion diaphragm on low-pressure sections—in the lowpressure exhausts there is a need to provide for the rapid
removal of steam from the internals of the casing in the event
there is a sudden and high rate of pressure increase due to
some transient condition
•
Diffuser at exhaust from the last stage—in an effort to maximize the energy extracted from the steam, the final rotating
blade is arranged to exhaust into a diffuser, normally produced as part of the casing fabrication
STATIONARY
BLADE ROW GEOMETRY
The primary function of the stationary blade row is to provide
controlled expansion of the steam from a high energy level to a
lower one; convert thermal energy to kinetic energy, and direct the
15
Turbine Steam Path Maintenance and Repair—Volume Two
resulting high-velocity steam jet into a row of rotating blade elements. These represent the primary functions.
However, there are other requirements that must be considered.
The stationary blades must also be able to accept steam from the
previous rotating blade row without incurring high incidence losses
and then redirect this steam from the inlet direction through the turning angle “θ” to discharge it without incurring excessive profile losses in the process.
Note: The turbine is designed to contain a number of stages
with a specific ratio of “U/Co” (blade tangential velocity to steam
adiabatic velocity). The enthalpy drops in each stage establish the
stage inlet and discharge pressures, and the stationary blades are
designed to pass the required quantity of steam within those inlet
and discharge pressures.
Stationary blade two-dimensional considerations
Consider the basic convergent nozzle shown as Figure 7.2.1.
Figure 7.2.1(a) shows development of the passage shape through
which the steam must expand, from an initial width of “Wi” to a final
or discharge width “Wd.” Figure 7.2.1(b) shows the developed width
sensed by the steam as it flows through and between the passage
walls. If the nozzle shape formed by these passage walls is maintained as a constant at various radial heights, the discharge area at
any position “g-g” can be found as the product of width “g-g” and
the radial height “H.”
This nozzle form would possibly be suitable if the steam were to
enter and discharge without a need to be deflected from an incoming angle to a different angle at discharge. However, it is also necessary for the nozzle to deflect the steam from an incoming angle
(most often an angle that is close to the axial) to an angle which, to
achieve maximum efficiency, should be as small as possible (consis-
16
Operating Damage Mechanisms & Refurbishment Techniques for Stationary Components
Fig. 7.2.1—Development of the passage shape in a convergent
nozzle, with no turning angle.
tent with providing adequate nozzle discharge area and enabling the
steam to enter the rotating blade row).
Consider the mean of the passage width line “k-m” shown as Figure 7.2.2. Steam enters the nozzle at an angle that is in a substantially axial direction “A” and discharges at an angle “α1”, which is
small and inclined to the tangential direction “T.” It is also required
that this nozzle have a discharge width “Wd” to provide the required
design discharge area.
The details of the profile can be established in the following
manner—an axial width “Y” is set consistent with strength requirements, and an inlet flow width “Wi” is calculated. It is then necessary to establish a profile shape that will provide both the turning
angle “θ” and the discharge width “Wd” that are required.
17
Turbine Steam Path Maintenance and Repair—Volume Two
Fig. 7.2.2—The steam turning angle of a nozzle passage.
The actual determination of a profile shape is often undertaken
by trial and error methods. Engineers in this task gather considerable
experience. Computerized systems of detailed design and evaluation
have somewhat simplified the assignment, and the engineer is able
to achieve greater optimization in establishing suitable geometries
for the vane profiles.
In examining the requirements shown in Figure 7.2.2, several
factors are worth attention:
18
•
The discharge width “Wd” will, as drawn, be formed as the
gap between two adjacent profiles. Figure 7.2.3 shows the
discharge tail between two such surfaces
•
Because the vane discharge tail cannot extend beyond the
discharge edge “a-a,” the actual throat is formed as shown at
the discharge point
Operating Damage Mechanisms & Refurbishment Techniques for Stationary Components
Fig. 7.2.3—Details of the vane discharge tail.
Fig. 7.2.4—The developed steam expansion passage, forming a convergent passage between two stationary profiles.
19
Turbine Steam Path Maintenance and Repair—Volume Two
•
The actual discharge angle of the vane will be somewhat different from the steam angle, and there is some small deviation (angle “Γ” as shown). The extent of this deviation
depends upon the curvature of the discharge tail suction surface. The discharge tail has some small thickness (“b”)
Figure 7.2.4 shows two profiles and details of the passage shape
formed between them, including line “k-m” and the passage width
variation from “Wi” at inlet to “Wd” at discharge.
Stationary blade three-dimensional
considerations
The row discharge area is established as the product of the mean
(or effective) throat and the radial height of each stationary blade
row opening. Consider several forms of vane and the variation of
throat “Wd” along the radial height.
Vanes of constant section. Such a vane is shown in Figure 7.2.5. The
profile is established and used along the entire radial height “H.” This
type of vane is used for those stages with small radial height-to-mean
Fig. 7.2.5—The radial variation of discharge opening along a vane of
constant profile.
20
Operating Damage Mechanisms & Refurbishment Techniques for Stationary Components
diameter ratios (H/Dm) of less than about 0.20. In such stages there is a
relatively small increase in pitch from the root “Pr” to the tip “Pt.”
An examination of the steam passages shows that the throat (now
given the more familiar symbol “O”) is formed on a curved tail of the
suction face. Because this is the smallest flow section it is clear that
the throat, and therefore the discharge angle, changes along the radial height. This is an acceptable condition and can be allowed for by
the designer, who determines the discharge area “Ad” by integration
along the radial height. (This is discussed in a later section.)
Vanes of constant but reducing profile. A similar profile used for
vanes of larger radial height-to-mean-diameter ratios employs a vane
of constant profile, but one that changes its chord “C” and other
principal dimensions along the radial height. Such a vane-stacking
diagram is shown in Figure 7.2.6. Five sections are shown at equally spaced radial locations from the tip “t” to the root “r.” In this figure the root section is the narrowest at “Wr” and the tip section “Wt”
is the widest where “W” is the axial width.
Fig. 7.2.6—Vanes of constant profile, but reducing chord “C”.
21
Turbine Steam Path Maintenance and Repair—Volume Two
In this design the throat “O” is formed on approximately the
same, but scaled position of the suction surface, maintaining the
same ratio of “O/P” and, therefore, a constant discharge angle along
the total radial height. The discharge area “Ad” can be determined
using the same method as in Figure 7.2.1. At each radial location the
vane has the same setting angle (“ ξ ”).
Vanes of twisted profile. Modern machining techniques allow
stationary vanes to be produced to a true vortex form, at a cost not
significantly higher than many other types.
Fig. 7.2.7—The vortex vane, with a profile of varying section.
22
Operating Damage Mechanisms & Refurbishment Techniques for Stationary Components
Figure 7.2.7 shows a typical vortex or twisted-type vane, where
the inlet nose has been adjusted to accommodate the inflowing
steam from the previous rotating blade row at different angles (“Vt,”
“Vm,” and “Vr”). The total discharge area is found in the same manner as the previous forms, and the throat characteristic can vary
along the length of the vane to accommodate design requirements.
Vane tilt
The radial disposition of the vane is selected to improve efficiency, both by maximizing the energy conversion process and also
by minimizing secondary losses that could occur by introducing vortices and turbulence, both of which induce losses. The vane can be
tilted in both the tangential and axial directions to improve the performance of the stage.
Tangential tilt. In many longer stages the stationary vanes are
given a tilt in the tangential direction of rotation. This is done to suppress the radial component of flow and to guide the steam flow into
the following rotating blade row in a more axial direction. This effect
is shown in Figure 7.2.8, where a tilt angle “ Ψ ” has been used.
This tilt “ Ψ ” often requires some special considerations at the
horizontal joint, and while causing some small increase in manufacturing costs, it is more than offset by the savings available in fuel
costs. In the diaphragm half shown in Figure 7.2.8, the true horizontal joint “H-H” is shown along with the actual joints (“h-h”),
which enable the vanes to be carried in the outer and inner rings.
Since both halves are symmetrical, the joints close.
Axial tilt. The stationary vanes will often increase in chord (and
therefore, axial width) as their position from the inner root diameter
increases towards the outer root diameter. To minimize the axial gap
between the stationary and rotating blade rows consistent with
maintaining adequate clearance at all radial positions, the stationary
23
Turbine Steam Path Maintenance and Repair—Volume Two
Fig. 7.2.8—Tangential tilt ‘ ψ’ in the stationary blade row.
blade is given an axial tilt. If a minimum axial clearance “Ca” is
required, the tilt shown in Figure 7.2.9 can be used.
This tilt is shown as a forward tilt “ ζ” from the true radial position “R-R” towards the root of the blade. This tilt allows an axial gap
(“Aa”) to be maintained between the discharge edge of the stationary
blade row and the inlet to the rotating blade. Because there are no
centrifugal effects on the stationary blades, this tilt will not induce
any additional stresses in the vane. This tilt and the proximity of the
edges will also allow minimum laps “Lo” and “Li” to be used.
Compound radial tilt. For large radial-height blades with high
discharge mach numbers and a large radial flow component within
the axial flow, the vane can be given a forward curvature to suppress
this radial flow. This suppression causes an advantageous mass flow
distribution. The outer flow portion of the vane is given an advanced
forward curve. This blade then promotes flow through the center
region of the blade path, which is more efficient.
24
Operating Damage Mechanisms & Refurbishment Techniques for Stationary Components
Fig. 7.2.9—The axial tilt in a diaphragm.
Stationary vane profile details
The profile is required to deflect the steam through a suitable
angle from inlet to discharge and to provide a discharge area sufficient to ensure energy conversion in accordance with the requirements of the designer. The profile’s form is established by several
methods, including hand and computer calculation and detailed
design layout to ensure that the form of the passage is acceptable.
There are certain details of the profile that the designer considers in
order to maximize energy conversion efficiency. These include the
following:
25
Turbine Steam Path Maintenance and Repair—Volume Two
The inlet nose. The steam entering the stationary blade row is
moving with a relatively low velocity because kinetic energy has
been extracted from it in the previous rotating blade row. In the
majority of stages it is also entering in a principally axial direction.
This means that the inlet nose is at about the 90-degree position.
This is not always so; the stationary vane form must be selected from
the requirements defined by the velocity triangle that is produced in
the stage design process. Various forms of nose can be used. The
intent of the nose is to accept the steam and cause it to divide and
flow down either the pressure or suction faces of the profile. The
radius and immediate form of the nose is important, as it should
introduce minimal flow disturbance likely to cause the boundary
layer to separate.
Figure 7.2.10 shows a typical nose form. It is produced from an
inlet radius “n” blending to a radius “a,” and then to “c” on the suction face and a radius “b” on the pressure face. This type of inlet is
used in many existing stages in current operation. A major consideration with this form of inlet is that at the points of blending from one
radius to another—or at any other location—there should be no sudden change in the radius of curvature, as this can promote boundary
layer separation.
Fig. 7.2.10—Details of the inlet nose.
26
Operating Damage Mechanisms & Refurbishment Techniques for Stationary Components
Discharge tail. The discharge tail thickness “b” (Fig. 7.2.11)
should be as thin as possible, recognizing that there can be high
loads induced there, and that it is necessary to maintain a nominal
thickness to avoid crack development. On some stages—where the
possibility of material loss by erosion occurs—it is possible to
increase the tail thickness to extend the life of the vane.
An aero/thermal design requirement is to maintain the discharge
tail relatively straight. This helps prevent premature separation of the
boundary layer. Excessive curvature on the tail suction surface can
create a large “wake” region downstream of the row (Fig. 7.2.11),
thereby inducing losses in the stage. The separation angle “Γ” must
be maintained at a minimum. It is probable the flow will be able to
sustain the boundary layer for a short distance beyond the throat
(“O”), but it will not be able to continue to turn with the curvature
of the tail suction face.
Fig. 7.2.11—Flow separation of a curved discharge tail.
27
Turbine Steam Path Maintenance and Repair—Volume Two
Surface curvature and finish. Surface curvature should be such
that there is no sudden change from one radius to another. This is
difficult with a profile defined from blending radii, but at the position of change, the blending should be as smooth as practical.
When machining and hand polishing produce vanes, it is possible to make this transition; and by finishing the surface in the
major direction of the chord, the condition of the vane can be
improved considerably. The design requirement of surface finish
must be observed.
Stationary blade row dimensional requirements
A number of different manufacturing processes are used in the
production of the impulse and reaction stationary blade rows. For
this reason, it is not possible to consider the requirements of each, or
the details of each step that should be monitored during manufacture. This section will discuss the row geometric characteristics and
consider the more important aspects of each.
In the impulse stage design, about 90% of the stage heat drop is
converted in the stationary blade row to velocity energy. For this reason there is every incentive for the manufacturer to control the quality of the stationary blades and ensure engineering requirements are
achieved in terms of dimensional conformance, surface finish, and
structural integrity.
Attaining design requirements to a large extent depends upon the
means of manufacture. Design requirements are normally defined
with the manufacturing process to be used and practical limitations
considered in the specification of acceptability. While specified tolerances on each of the major parameters may vary, the following (as
shown in Fig. 7.2.12) are considered the most essential in achieving
an acceptable product.
28
Operating Damage Mechanisms & Refurbishment Techniques for Stationary Components
Fig. 7.2.12—Stationary vanes and the principal control dimensions.
The blade opening (throat) “O.” The throat and its variation
along the radial height of the vane are fundamental in establishing
the efficiency of the blade row. Consistency from opening to opening means an effective opening (“Oe”) is necessary to establish the
individual aperture areas and, in total around the row, to achieve the
stage discharge area.
When a blade row is manufactured for a diaphragm, or when
individual blades are assembled in a casing, the opening will vary
from outer to inner section. It is therefore necessary to monitor and, if
required, to adjust these individual values to achieve the correct area.
The vane pitch, “P.” The pitch between blades is established in
terms of the diameter being considered and the number of blades in
the row. The actual pitch achieved in any row and its variation from
passage to passage depends upon the method of manufacture.
For individual body blades, there should be little variation since
the pitch will depend upon the machined root block, which can be
held within very tight tolerances. There is of course the possibility of
vane lean, which will cause a cumulative error along the radial
29
Turbine Steam Path Maintenance and Repair—Volume Two
height. This can often be corrected by means of minor adjustment or
positioning with an attached coverband.
For cast, welded, and fabricated pieces, locating the individual
stationary vanes from some marked position and attaching them to
inner and outer rings usually achieves the pitch.
The number of pitches is selected to accord with the design
requirements, achieve discharge area and angle, and complete a half
circle (180 degrees). The final and most important measure is the
cumulative pitch around the half diaphragm. This is measured and
controlled to ensure joint blades are in their correct position so pitch
tolerances at the horizontal joint are achieved.
The ratio “O/P.” The second most important function of the stationary blade rows is to assure the steam is discharged at the correct
angle into the rotating blade row.
For the rotating blade the discharge angle was defined as “β2”
and determined as the ratio of opening to pitch. Similarly, for the stationary row, the discharge angle “α1” can be defined as:
Therefore, this ratio “O/P” is critical to the total stage performance. It is possible for the area to be correct, but if the variation of
throat is outside the design specification, this will modify the discharge angle variation along the radial height and will also modify
the radial steam flow distribution.
The vane-setting angle, “ ξ.” The blade as adjusted to its correct
setting angle “ ξ ” is designed to provide an expansion passage sufficient to produce the correct throat—the desired ratio of “O/P” and
the width “W” (Fig. 7.2.13). However, any change in the setting
30
Operating Damage Mechanisms & Refurbishment Techniques for Stationary Components
angle will modify both the throat and the width “W.” The modification of the throat will modify the ratio “O/P.” While some level of
error can be tolerated, this must be controlled within design tolerances. Normally, variation in the width either positive (“+dW”) or
negative (“-dW”) can be accepted in a diaphragm, but must be considered more carefully in a stationary blade assembled to the casing
in a reaction unit. These blades may have a carefully controlled axial
clearance that may not be able to be compromised.
Fig. 7.2.13—Showing the effect of varying the vane setting angle ‘ ξ’.
Vane tilt angle, “ Ψ.” It is also possible with some forms of construction to adjust the vane tilt angle “Ψ” (Fig. 7.2.8) from the true
radial line. This tilt causes a radial inward pressure on the steam and
reduces the radial flow component. It also minimizes the nozzle
effect on the downstream side of the vane. It is further possible that
the tilt angle will vary along the radial height for those vanes with a
compound radial tilt.
Inlet and trailing edge setback, “dai ” and “dao.” The blade lattice is normally built up by whatever form of construction is used so
31
Turbine Steam Path Maintenance and Repair—Volume Two
that the inlet and discharge nose lie in a true tangential plane. In fact,
there is some degree of tolerance in the axial shift “dai” and “dao,”
allowing the vanes to be either “proud” or “recessed” on both edges.
This effect is shown in Figure 7.2.12.
Sidewall discharge diameters, “Dot ” and “Dor.” The sidewall
diameters “Dot” and “Dor” [shown in chapter 2 as Fig. 2.12.1(a) for
an impulse design and Fig. 2.12.1(b) for a reaction stage] show the
importance of maintaining these diameters so the correct lap can be
achieved in these stages.
The requirements discussed above define the optimum form and
arrangement of the stationary blade row elements. However, for various considerations (such as mechanical strength and assembly) it is
not always possible to achieve these in the final design, and it
becomes necessary to make compromise selections of various
parameters and requirements to achieve an acceptable stationary
row. These changes include the following:
Vane cross-section irregularities. This section has discussed the
essential nature of the vane sections—the need to ensure that both
are the same from element to element, and mounted so as to ensure
the expansion passages generated between them are as identical as
possible within manufacturing limitations.
There are however, three situations where this identical form of
passage is not achieved. The design engineer is aware of this and in
fact has specified the steam path so that this does not occur. These
situations include the following:
•
32
Nozzle box—the nozzle box is located at the inlet end of the
high-pressure section, and accepts steam entering the unit
from the boiler or steam source. In nozzle-controlled units
this box is designed to admit steam to a selected quadrant of
the entire 360-degree inlet. By design, each of these steam
Operating Damage Mechanisms & Refurbishment Techniques for Stationary Components
admission quadrants is isolated from its neighbors so that the
steam entering it is discharged through the nozzles into the
rotating blade row over a very defined arc
Therefore, at the inlet to this stationary blade row, there are
walls produced at the end of each inlet passage (Fig. 7.2.14) where
there are differences in the expansion passage shape. These differences are not significant in terms of passage flow characteristics,
but represent a departure from the complete similarity from one
passage to the next.
Pe
Pz
Fig. 7.2.14—Showing the endwall shapes of a nozzle segment.
If these normal nozzle vanes have an axial width “W,” and the
nozzle plate is made marginally thicker at “Wu,” then there is a
small difference (“Wu-W”) that is a short guide for the steam entering from the inlet plenum. The nozzle plate is attached to the inlet
plenum; the inlet pitch to the plate is shown as “Pe.” At discharge
the total pitch is “Pz,” which is equal to the sum of the individual
nozzle discharge pitches. Normally “Pe” is greater than “Pz”
because of the sidewall taper at the inlet ends
•
Extended vanes (for axial strength)—the vanes used in certain stages that have a high-pressure drop across them must
33
Turbine Steam Path Maintenance and Repair—Volume Two
be evaluated for their axial deflection under the effects of
pressure and steam momentum loads. If these loads become
excessive, it is common to use extended axial width vanes
(Fig. 7.2.15). Such vanes are used on a portion of the total in
each stage and extend the axial depth of the steam path from
“W” by an amount “E” to “Wu”
Fig. 7.2.15—Extended section vanes, used for axial strength, within the
blade annulus.
The extended portion of these vanes is selected to preserve (to
the greatest extent possible) the aerodynamic form of the expansion
passage; it has little or no effect on the flow distribution of the steam.
If, however, the incoming steam has an inlet angle significantly different from that of the extended vane metal angle, there could be
some effect on the “swallowing capacity” of the individual passages
•
34
Throttle controlled units at their horizontal joint—units
designed for throttle control allow steam admission over the
complete inlet arc on the first stage in the high-pressure section of the unit. This mode of admission removes the highimpact forces that the first stage experiences in the nozzlecontrolled units. This is why the first-stage rotating blades are
normally designed for lower stress levels
Operating Damage Mechanisms & Refurbishment Techniques for Stationary Components
The first-stage stationary row will still receive steam from the
inlet plenum, which because of design-required discontinuities at the horizontal joint will introduce a small “dead
band” at this location. In Figure 7.2.16 the dead band is
shown as the tangential length “Db,” which occurs at the
horizontal joint “h-h.” No steam is admitted over this small
arc and therefore there is no steam load on the rotating
blades. This change in steam load is sufficient that the rotating blades experience a 2/rev stimulus. Under certain circumstances this high-frequency load change can result in
cyclic damage and rotating blade failure.
A solution some manufacturers have adopted is to remove
material from this dead band region (Fig. 7.2.16), producing
only a small dead band of tangential width “w.” The removal
process must be controlled so as not to introduce too much
curvature onto the outlet surface, thereby producing an
excessive wake. However, the net effect should be to lower
the influence of these impact loads on blade life.
Fig. 7.2.16—The method of reducing the effect of ‘dead band’ in a throttle controlled unit.
35
Turbine Steam Path Maintenance and Repair—Volume Two
OPERATING PHENOMENA
AFFECTING THE STATIONARY
BLADE SYSTEM
Phenomena that cause stationary blades and diaphragms to deteriorate are, in many instances, the same as, or related to, those
affecting the rotating components of the steam path, i.e., many of the
problems encountered in rotating blades are present in the stationary blades, as well. However, these phenomena may manifest themselves in a different form or there may be a different appearance or
form of the damage.
Because the stationary blade components are not normally as
highly stressed, repair methods and materials are often available
allowing them to be refurbished, and repairs can be undertaken on
components that appear to have suffered irreparable damage. The
principal factors or influences causing deterioration of the stationary
blade system follow.
Chemical deposition
Common among problems affecting stationary blade passages is
chemical deposition on the vanes and sidewalls. This deposition
results in the buildup of layers of various compounds within the
expansion passage. Such deposition will cause a surface roughening
and therefore a frictional loss in the steam path. Although the deposits
will accumulate on the stationary blading, the patterns are different
from those seen on the rotating components. The probable reason for
this difference is that during the passage of the steam through the
rotating blades, the contaminants have a radial flow component,
causing greater quantities to be deposited on the outer sections of the
steam path. This is not the case in the stationary components.
36
Operating Damage Mechanisms & Refurbishment Techniques for Stationary Components
The radial flow component in the steam transporting the
chemical contaminants causes the main chemical deposit to occur
on the underside of the coverband and outer regions of the rotating blade. This radial flow phenomenon also accounts for the
higher deposits on the outer sidewalls of the stationary blades,
although the stationary components have a more even flow distribution. Figure 7.3.1 shows deposits on the stationary vane portion
of a high-pressure stage diaphragm as seen from the horizontal
joint. Figure 7.3.2 shows deposits on a lower-pressure stage—definite accumulation patterns at the opening—and Figure 7.3.3
presents deposits that occur around the strengthening stub attaching a blade portion at the horizontal joint.
Fig. 7.3.1—Showing similar deposits on a lower pressure stage. There is significant
deposit in the throat region.
37
Turbine Steam Path Maintenance and Repair—Volume Two
Fig. 7.3.2—Concentration of deposits in the inner flow portion of the expansion passage.
Fig. 7.3.3—Deposits on the strengthening stub of a horizontal joint blade.
38
Operating Damage Mechanisms & Refurbishment Techniques for Stationary Components
The even deposition pattern in the stationary blade rows will
result in a somewhat larger frictional energy loss, because all areas
of the vane and expansion passage can be affected. There is, therefore, a greater likelihood that when a stationary blade passage is
examined, it will be noted that the deposit is more evenly distributed
over the entire radial length of the vane and inner and outer sidewalls. There is also a tendency for solids to precipitate from the
steam near the opening of the passage, mainly onto the convex or
suction surface of the stationary blade vane.
When a unit is opened, the opportunity exists to blast clean the
stationary blades and the inner and outer sidewalls to remove as
much of the chemical deposit as possible. It is suggested that when
deposits are found, samples be collected and analyzed to determine
their chemical composition and establish if aggressive compounds
are present. If gaps are present at the junction of the vane and sidewalls (produced by casting or welding cast-type diaphragms), no
special effort should be made to remove deposits from the gap.
That’s because this deposition is of more benefit than harm—it
removes a flow-disrupting device, i.e., deposits in this region can
help minimize the turbulence caused by steam flow into and out of
the gaps. The exception to this is when chemically aggressive compounds are found in the deposits.
Controlling the cleaning procedure is important, as this can damage the discharge tails of the vanes (see chapter 6, Fig. 6.4.1). This
damage can result from the use of an air pressure that is too high, too
hard, or too large a grit, by exposing the surfaces for too long to the
effects of the grit, or from holding the nozzle too close to the vane.
The procedure for blasting should be defined, calibrated, and
applied under careful conditions not causing severe distortion of the
vanes. The requirements for blast cleaning the stationary elements
are the same as outlined in chapter 6 for the rotating blades.
39
Turbine Steam Path Maintenance and Repair—Volume Two
Solid-particle impact damage
Stationary components (as rotating components) are subject to
solid-particle impact damage.
Generally, the consequential harm associated with impact damage suffered by stationary blades is not as severe as that on the rotating blades because there is normally a larger impact velocity
between rotating blades and any debris than there is between debris
and the stationary blade rows. Also, the rotating blades tend to have
craters formed on the inlet edge, whereas stationary blades and
diaphragm vanes have a greater tendency for the damage to occur
predominantly on the outlet or discharge tail. This latter area is less
robust and damage can occur on either the pressure or suction face.
Despite the lower impact velocities (compared to rotating
blades), impacts can result in craters of various size and form, produced on the surface of the vanes, or massive damage, including significant material rupture, which requires replacement of a portion of
the vane or stationary blade row. These impact craters will be a
source of continuing energy loss and efficiency degradation. If large
enough, such areas of damage can also be a possible source for the
development of flow disturbances in the steam, which are capable
of inducing vibration damage in following blade rows. Such a level
of damage is not a common situation though its influence should be
considered in the event of unaccountable failures after units are
returned to service with uncorrected damage.
Damage that occurs on the stationary blade rows requiring remedial action is most often associated with the discharge tail. The cross
section of the stationary blade profile—which is thinner than the
rotating blade—tends to be deformed to a greater extent than other
portions of the profile. In addition, damage at the discharge tail can
influence both the outlet area and angle at which the steam is discharging from the stage at that point.
40
Operating Damage Mechanisms & Refurbishment Techniques for Stationary Components
Diaphragms and stationary blades are often damaged by particles generated within the steam path from some mechanical rupture
upstream of the stage. These detached pieces break loose and transport through the steam path, and though these are broken or
chopped into smaller pieces by the action of the rotating blades, they
can often rebound between the stationary and rotating parts of the
unit during their passage through the stage. Such multi-impacts can
cause bending and cracking of the vane and discharge-opening distortion. When it occurs, this type of damage occurs in a relatively
short period of time and can often be detected, or inferred, by the
change of pressure measured at various locations within the unit. If
performance monitoring is carried out, this type of damage can be
more readily detected because such damage will have an effect on
the unit efficiency level.
It is difficult to quantify this damage because there tends to be little consistency of measurable damage from one incident to another.
Possibly the most suitable method of quantification is the level of
repair required to return the components to a fully serviceable condition. Damage is termed light, medium, or heavy, according to the
definitions.
•
Light—this damage can normally be corrected to an acceptable degree by hand dressing, using a file and/or emery
cloth. Such correction can even include bending the discharge tail to achieve an acceptable condition to a degree.
Figure 7.3.4 shows a blade row that has suffered impact
damage at the outer regions for steam flow and also shows
evidence of some larger impact that has deformed the discharge tail. Figure 7.3.5 shows similar damage; in this case,
damage has progressed to a slightly more severe condition
but can probably be returned to acceptable parameters by
dressing and minor bending
41
Turbine Steam Path Maintenance and Repair—Volume Two
Fig. 7.3.4—Relatively light impact damage in the outer regions of a vane. There is also
evidence of mechanical tears in the vane.
Fig. 7.3.5—Light peening damage of greater severity in the outer regions of steam flow.
42
Operating Damage Mechanisms & Refurbishment Techniques for Stationary Components
•
Medium—this damage is similar to that shown in Figures
7.3.4 and 7.3.5 but has increased to the extent it cannot be
repaired by hand dressing and bending. Figure 7.3.6 shows
the outer section of a vane in which damage has ruptured the
vane at the outer sidewall. Similarly, Figure 7.3.7 shows a
slightly different form of this same level of damage; here the
vane does not exhibit such severe crater damage but it has
been torn from the sidewall and the opening has closed as a
consequence. This form of damage (with the closed opening)
is normally caused by a piece of free debris in the axial gap
between the stationary and rotating blade rows. There are
medium pits and dents that must be repaired. It is normal with
this type of damage to remove the discharge tail and rebuild
Fig. 7.3.6—Medium damage, being most severe in the outer regions of steam flow, which
also exhibits a radial crack starting at the sidewall.
43
Turbine Steam Path Maintenance and Repair—Volume Two
Fig. 7.3.7—Localized medium type damage at the outer regions of steam flow.
•
Heavy—this damage is severe and normally requires the
removal of the complete discharge tail and weld refurbishment. There are situations, particularly with large, low-pressure stages, where the complete discharge tail can be
removed and replaced by weld attaching an insert. Such
damage is most often repaired by total weld deposit, certainly for small radial-height vanes
Figure 7.3.8(a) shows a control-stage nozzle plate in which a
piece of by-pass valve seat (Stellite 6B) has detached and lodged in
an opening. Another piece had passed into the axial gap between
the stationary and rotating blade row and caused extensive damage
to the stationary row vanes [Fig. 7.3.8(b)].
44
Operating Damage Mechanisms & Refurbishment Techniques for Stationary Components
Fig. 7.3.8(a)—Mechanical damage on a nozzle block. A piece of pilot valve seat.
Fig. 7.3.8(b)—Damage caused by a piece of pilot valve seat that has passed through the
control stage stationary vanes.
45
Turbine Steam Path Maintenance and Repair—Volume Two
It is not always possible to be sure that such damage conditions
exist from condition monitoring alone. Any measured change in
state-line efficiency from “enthalpy drop tests” could be due to a
number of different conditions related to conditions such as surface
texture seal clearances, and will provide no evidence of mechanical
damage. Unlike rotating elements, such damage causes no change
in vibration levels and is therefore not easy to detect.
Solid-particle erosion
After the boiler superheater delivers steam or returns it from the
reheat section, the first stages of a turbine can be subjected to relatively heavy damage as the result of the gouging action caused by
exfoliated scale particles carried over from the boiler. (This is discussed in chapter 4.)
This erosion of the stationary blade row can be a slow process,
and although its effects and presence can be inferred from both pressure and state-line efficiency measurements, because it is a gradual
deterioration, monitoring must be undertaken with test-quality
instrumentation. It is often difficult to differentiate between the
effects of this damage and damage due to chemical deposition.
Figure 7.3.9 shows a portion of a stage from a high-pressure stationary blade row. This row has suffered material loss from the vanes
discharge tail. This loss affects both the discharge area and angle and,
if not corrected, it will cause a significant operating fuel cost penalty. The scale that enters the row collides with the pressure face of the
vanes, and because of its hard abrasive nature, it removes metal from
this surface (see chapter 4, Fig. 4.8.5).
There is a tendency for the scale entering the stationary and
rotating blade rows to migrate to the outer steam flow diameters.
Such scale will, therefore, cause more severe damage in that
region. This effect can be seen in Figure 7.3.10, where material
46
Operating Damage Mechanisms & Refurbishment Techniques for Stationary Components
Fig. 7.3.9—Severe material loss due to solid particle erosion on the discharge edge of the
vanes of a high pressure stage.
Fig. 7.3.10—Extensive damage in the outer flow section of a high pressure stationary
element. This damage is caused by a combination of SPE and solid particle impact.
The damage has been severe on the thinned portions of the vane.
47
Turbine Steam Path Maintenance and Repair—Volume Two
has been lost primarily in the outer diameter positions. In this
stage the eroded vane sections have been further damaged by
solid-particle impact.
The material loss that occurs on the vane discharge tail is of two
basic forms (chapter 4):
•
A gouging or grinding action that removes material at a relatively linear rate. Such a material loss is shown in Figure 7.3.11
•
A brittle or chipping type fracture common in austenitic materials. Figure 7.3.12 illustrates partial loss by the chipping condition
These types of failure and base materials—i.e., martensitic or
austenitic steels—can be repaired. However, the weld filler rod and
stress relief requirements will be different for the two steels.
Fig. 7.3.11—Severe material loss as a result of a gouging removal from the discharge edge.
This stage shows evidence of an earlier repair, some of which material appears to have
been lost at a faster rate than the base material.
48
Next Page
Operating Damage Mechanisms & Refurbishment Techniques for Stationary Components
Fig. 7.3.12—Brittle or ‘chipping’ type failure of a nozzle block discharge tail.
Fig. 7.3.13—Erosion damage causing material loss from the outer section of a stationary
vane inlet nose. This loss is causing an ‘undercut’ of the vane, and could weaken it against
creep type damage.
49
Chapter
8
Refurbishment
Techniques for
Rotating Blades
INTRODUCTION
The rotating blade is the steam path element that seems most susceptible to damage in normal operation. This is partially because
both water and any mechanical debris impacting the blades will,
because of the high tangential velocities of the rotating blades, develop sufficient force to possibly cause severe mechanical deformation.
In addition, the stress levels in the rotating blades are so great that
when combined with irregularities of steam flow, they create the possibility of causing dangerous levels of vibration. There are also discontinuities in the blade shape, creating a potential problem of stress
concentration in many regions. These factors, together with the possible presence of aggressive chemical compounds and the probability of externally imposed shock loads, indicate the continual potential
for failure or serious damage to these components, resulting in an
201
Turbine Steam Path Maintenance and Repair—Volume Two
interruption of service. Many millions of rows of blades are in service, and their failure is relatively rare. However, they do remain the
component causing more outages than any other in the unit.
Some of the performance-deteriorating mechanisms do not
require repair or replacement. However, they may require the turbine be cleaned. Even this upgrading action must be performed to
accepted standards to prevent deterioration of the unit.
This chapter considers the damaging mechanisms and deteriorating performance (discussed in chapters 4 to 7), and any other
phenomena, in terms of their ability to be repaired. In some
instances these procedures are described in general terms. However, many of these repair and refurbishment techniques represent relatively new techniques, which continue to develop as newer welding techniques and improved materials become available.
None of the procedures described in this chapter should be considered absolute. It is normally necessary to review each failure or
incidence of damage separately, and determine the local condition,
which must be considered within the review. Refurbishment/repair
technology continues to develop, making new procedures and techniques available. These newer techniques may offer a superior or
more cost-effective repair. If welding techniques are employed in the
procedure, these must be selected with care relative to the materials
involved, and then their application must be calibrated, specified
and controlled, because the blade materials (and possibly the stage
hardware) can be affected by the preheat temperature; the rates of
temperature change for both heating and cooling for the repair and
stress relief.
The deteriorating mechanisms affecting the blade vane will be
examined relative to their effect on the blades. Consideration will be
given as to how these damaged or deteriorated blades can be rectified to achieve an acceptable condition to the extent the component
can be returned to service. Other rotating components of the turbine,
202
Refurbishment Techniques for Rotating Blades
including the blade root are considered in chapter 9. The root is
conveniently left to chapter 9, because any repair or corrective
action taken on it must be considered in relation to the transfer of
load between the blade root and rotor; any change must be reviewed
in terms of the effect on the entire blade.
When corrective action is required, there is often more than one
procedure available. If alternative refurbishment techniques are
available, they are considered and described. It must be recognized
at the outset however, that because of the diversity of form of the
components of the steam path, and from manufacturer to manufacturer, it is unlikely any repair procedures can be recommended as a
remedy for all maintenance problems. In practice, it is necessary to
examine each turbine, and make an assessment of what repairs can
be achieved for a reliable and economic solution. Then undertake
these in relation to the requirements of the design.
Many of the observed deterioration effects of operation will
require components to be replaced. Other situations, either because
of the extent of damage or their location, will suggest certain repair
procedures can be used without jeopardizing the integrity of the
components that have suffered damage. This condition review normally requires mature judgment of the situation, and an assessment
of any risk involved. This is particularly the case when repair procedures require large amounts of heat to be used, because if not controlled, they can modify the metallographic structure of the materials.
The possibility of modification to the blade or stage hardware
materials is especially important in components subjected to high
temperatures, pressures, and operational stresses. In many instances,
using the existing repair procedures with advanced materials and
methods returns the component to a condition making it superior in
quality to that which was originally supplied. The basis for the superior nature of these repairs, is that since the component was new,
both materials and welding procedures have advanced to the extent
product quality can be improved. This is not an adverse reflection of
203
Turbine Steam Path Maintenance and Repair—Volume Two
the original design or designer, but a demonstration of advances in
state-of-the-art modifications for product improvement. Also for a
minimum cost and delay, the owner/operator can return his/her unit
to service confident it will perform at a level reflecting the adequacy and acceptability of the repair procedure.
There are occasions when, although repairs to existing components may be possible, it is faster and more cost effective for the
operator to purchase new components. The factors influencing this
depend considerably upon the delivery period for the replacement
components, the repair time, the cost of any forced or extended outage, and the availability of materials that are needed to construct the
components. No general rules for such a decision exist; each case
must be judged individually. The owner should be aware however,
that when parts are replaced, the damaged components can often be
refurbished and carried as inventory spares for possible future use.
For this reason, components should always be removed without
destruction, if possible.
STEAM PATH CLEANING
An essential part of turbine maintenance is the cleaning of the
steam path. The steam path surfaces will become coated with various compounds carried over into the turbine unit as a consequence
of operation. Although this cannot be avoided, it can be reduced.
These deposits have potential degrading aspects to the performance
of the turbine unit.
Chemical deposits on the steam path components, both the
vanes and sidewalls, cause a deterioration of the stage efficiency.
This efficiency loss occurs as a consequence of inducing flow separation and turbulence into the steam flow. When a turbine is opened
204
Refurbishment Techniques for Rotating Blades
for inspection, and the steam path becomes accessible, it is a normal
procedure to clean the components, normally by blast cleaning,
although hand techniques can also be used effectively. However, the
latter tends to be slower and more costly.
Despite the amount of time the manufacturing department of a
turbine supplier devotes to preparing the steam path elements to
achieve design requirements, these surface conditions will not be
maintained for extensive periods after the unit is placed in service.
The requirements of steam path cleaning were discussed in
chapter 6. It is necessary to consider the methods available for cleaning and how they may be applied to the rotating blades. The available cleaning methods include:
•
blasting with a suitable medium
•
water washing
•
hand cleaning using a suitable solvent
Blast cleaning. Blast cleaning is the most commonly used
method of removing deposits from the rotating portions of the unit,
once it is available for cleaning. This is essential as it allows nondestructive examination and removes even the most persistent
deposits. The considerations listed in this chapter must be observed.
Modern usage prefers aluminum oxide as the most suitable blasting
media. These procedures result in both rotor and blades being suitable for NDE, and return to service.
Water washing. Although this procedure is not as commonly
used as blasting, it has been used with varying degrees of success,
when a unit is open with the rotor and diaphragm removed. Water
washing will dissolve soluble deposits in the unit, but the insoluble
deposits must be removed by mechanical force, which requires a
high-pressure jet. This slower method is made faster with higher
205
Turbine Steam Path Maintenance and Repair—Volume Two
pressures, which again introduces some level of risk associated with
bending the edges of the blades (if they are thinned).
The use of “on-line” water washing is found to be effective on
geothermal units that operate on steam, which is initially in the wet
region. It has in general been of little use on large utility units with
high initial steam conditions, and any efficiency gain has not been
sustained for a period sufficient to justify the use of such a procedure.
One possible advantage to water washing is that it can remove
some of the chemically aggressive compounds from hideouts, if they
are water-soluble. However, it should not be concluded that all compounds will be removed. This water washing should only be considered a diluter action. It is not a complete remedy.
Hand cleaning steam path elements. Hand cleaning methods
tend to be a slow process, but can be justified in situations where certain surfaces are inaccessible, or the use of blasting for extended periods would be required, and such extended exposure to the effects of
blasting could possibly damage sections of the blade system.
BLADE INLET
EDGE EROSION DAMAGE
Stages operating with a two-phase flow of water/steam can be
subject to moisture impact, and under certain conditions will suffer
erosion damage. Blades can operate successfully for many years and
the erosion severity should not be such that they need to be repaired
or replaced during a normal operating life, providing no unanticipated conditions occur. Manufacturers normally predict erosion life,
and expect the blades to be satisfactory for a 30-year operating life.
However, a level of erosion penetration can be reached when this
206
Refurbishment Techniques for Rotating Blades
impact type damage has proceeded to the extent remedial action
must be taken, or the blade would suffer material loss beyond the
point at which repairs could be effective. Under such conditions the
blades could be operating at risk. Certainly there exists the possibility of missiles being generated from the blade because of high stress
concentration and the rate at which erosion is proceeding.
It is not possible to make any general statement regarding the
need for remedial action or when it is necessary. Rather, each stage
and blade within a stage must be judged on its own merit. When,
from previous observations, it is deemed necessary to take corrective
action and a turbine is to be repaired, then such repairs should be
undertaken before the erosion pits penetrate the shield and progress
into the base or parent material to any significant degree.
Most often the need for blade or shield replacement occurs
when a shield has been lost, or secondary erosion or leading edge
and/or cracking is present. Figure 3.8.9, in chapter 3 shows heavy
(secondary) erosion at a blade inlet edge. This erosion may often
require repairs to the blade before the entire shield life is expended
by normal erosion. Such replacement is normally required only
when localized penetration of the erosion, or a crack propagates too
far into the base material. In the case of secondary erosion, if it
becomes necessary to repair or take remedial action, efforts should
be made to trace the source of secondary collection or concentration
and see if corrective action is possible. In the case of crack propagation, it is necessary to determine if the crack has extended into the
blade material. This will determine the form of repairs that must be
undertaken.
Note: Secondary erosion is defined as the local erosion where
moisture appears to have been concentrated, causing excessive penetration at one radial location. This is normally seen on each blade
in the row.
207
Turbine Steam Path Maintenance and Repair—Volume Two
Moisture-impact erosion removes most material from the rotating blade inlet edge in its outer sections. This damage, while causing some small reduction in stage efficiency, normally poses no danger in terms of blade failure, unless some form of secondary damage,
resulting from other phenomenon is also present. When the efficiency loss is small, it is difficult to justify (with any degree of certainty)
the need for corrective measures. Therefore, the need to make
repairs or refurbish the blade is a consequence of the secondary erosion damage, which can become more severe.
The secondary damaging mechanisms or phenomenon that
require corrective action can take several forms, but can generally
be considered to be the following:
Detachment of a braze attached erosion shield. If a braze
attached shield detaches, there is consequent exposure of the softer
material of the blade vane. After a relatively short period of operation,
the blade material will erode at what can be an accelerated rate.
With some longer shields, it is possible they will have been
attached as a number of individual segments on each blade. It is possible, with these designs, for individual segments of the shield to
detach, leaving the blades with localized areas exposed, and protection intact in others. Figure 3.8.12 of chapter 3 shows the consequences of such an incident where the outer segment of a shield has
detached allowing heavy erosion to occur in the blade material in
the outer sections of the blade.
Local erosion penetration. It is possible for moisture collection
points to exist in the stationary portion of the steam path. These locations tend to concentrate the moisture detachment points, and therefore the radial location at which the moisture makes contact with the
rotating blade inlet edge. This will result in severe secondary erosion
penetrating the erosion shield. Such damage is shown in Figure 3.8.9
in chapter 3.
208
Refurbishment Techniques for Rotating Blades
Cracking originating in a weld-attached shield. With weldattached shields, cracks can initiate at the blade inlet edge. Such
cracks normally originate in the HAZ, and propagate towards the
inlet edge. However, a small proportion of these cracks will run
either radially along the heat-affected zone (HAZ), or across the vane
towards the discharge edge. Figure 8.3.1 shows such a crack, and
Figure 8.3.2 the form of the weld-attached shield, at the position “Z”
(the point at which these cracks originate), and the degree of protection required to prevent erosion in the softer blade material. Figure 8.3.3 shows an exposed crack, of the type shown in Figure 8.3.1
in the shield of a last stage blade element.
Fig. 8.3.1—A crack initiating at the leading (inlet) edge of a last
stage blade with a weld attached erosion shield.
209
Turbine Steam Path Maintenance and Repair—Volume Two
Fig. 8.3.2—The crack initiation point in the HAZ of the weld attached
erosion shield.
Fig. 8.3.3—The exposed section through a leading edge crack
showing the failure surface.
210
Refurbishment Techniques for Rotating Blades
TAPPS
Cracks initiating in the braze material. Figure 8.3.4(a) shows a
crack originating at the inlet nose of a blade with a braze-attached erosion shield. This figure shows the blade cross-section, and in Figure
8.3.4(b) a magnification of the inlet nose. At this location there had been
a fault in the braze, a porosity hole, which filled with water and collected corrodents that helped initiate a corrosion fatigue failure.
TAPPS
Fig. 8.3.4(a)—A high cycle fatigue crack initiating at the inlet nose of a last
stage blade.
Fig. 8.3.4(b)—The initiation sites on a blade inlet nose as shown in figure 8.3.4(a),
the cracks initiating in the attaching braze material, where a porosity hole existed,
and collected corrosive ions.
211
Turbine Steam Path Maintenance and Repair—Volume Two
A number of initiation sites can be seen on the inlet edge, but
the primary failure appears to have been at the inlet nose tip.
Thermally hardened inlet edges. Some manufacturers provide
erosion protection in the form of thermally hardened, vane inlet
edges. This method is known to provide good protection. However,
should there be a failure to maintain steam reheat temperature, an
increased moisture content will result, or a crack developing in the
transition zone between the hardened and unhardened material,
then corrective action must be taken. In this situation, because there
are no externally attached shields, some drastic procedure may need
to be adopted.
A previous solution has been to “crop” the blade below the
crack, and also the blade that is diametrically opposite; this is necessary to preserve dynamic balance.
When a situation of damage, or need for repair or refurbishment
exists, engineering decisions are required concerning the condition,
its severity, and the most appropriate corrective action. There are
various procedures that should be evaluated, and where appropriate
corrective action is required. These available corrective measures
include the following:
212
•
Braze attach new shielding to blade material. This is the most
common solution, and can be undertaken in the field with
the rotor in the unit, if necessary
•
Braze attach shielding, with weld metal build up. If the blade
material has suffered erosion, it could be possible to deposit
small amounts of a suitable filler material such as Inconel 82
and hand dress. However, this could require a short period
of localized stress relief for the blade material
•
Weld attach an erosion resistant bar nose to the blade inlet
edge. If the shield and blade material have been penetrated
extensively, there are considerable advantages to replacing
Refurbishment Techniques for Rotating Blades
the entire inlet nose with a material better able to resist erosion. This is considered later in this chapter
•
Weld attach nose, as above, and build up base material to
provide a suitable attachment seat or area. This is considered
later in this chapter
•
Raw weld deposit, using a Stellite 6B® stick. This deposit is
made directly onto the blade material
Weld attaching a solid bar inlet nose
This is the procedure used for the weld attachment of a preformed
erosion shield, or solid bar nose, to the inlet edge of a blade vane, the
vane having suffered damage to the extent the existing shield needs to
be replaced to ensure the blade can be returned to service and continue to operate at an acceptable level of performance.
Preparation of the blade vane. Upon receipt of the blades, portions of them are used to establish both the vane geometry, and the
position of the vane placement on the blade root platform. This
includes establishing the vane setting angle “ ξ” at various radial positions. The information gathered must be statistically significant, and
sufficient to allow a shutter gauge to be constructed. This gauge must
contain sufficient planes of measurement so that each blade can be
placed in it and measured for geometric conformance at completion
of the repair procedure. Gauge planes should not be more than about
6.0" apart in the radial direction. A typical shutter gauge is shown as
Figure 8.3.5. The blade shown here has its inlet nose removed.
The inlet nose will often have been partially destroyed by erosion, or removed or deformed by some other phenomenon. This portion of the vane will therefore need to be reconstructed from the
existing material. It is normally possible to gauge the outer portion of
the blade form from the existing vane material, so that shutter gauges
in the outer regions can be reproduced.
213
Turbine Blading (UK) Ltd.
Turbine Steam Path Maintenance and Repair—Volume Two
Fig. 8.3.5—A blade having had its inlet
edge removed, is located in a shutter
gauge for dimensional checking prior
to repair.
After construction of the shutter gauge, the remaining blades are
placed in it, and gauged for conformance. Any significant variation
in the main vane characteristics, or the parameters shown in Figure
8.3.6, will be noted as this is indicative that the original blade has
either been deformed during operation, or was produced outside
manufacturing tolerances. Significant deviation from the mean values as measured in the shutter gauge should be noted by blade number. Each blade must be completely gauged. When this has been
completed, and details recorded, the inlet edge will be prepared for
the attachment of the bar nose. This nose is traditionally Stellite 6B®.
However, it is also possible to employ a steel tool, which is hard and
can possibly provide superior protection. If the blades being refurbished are from a unit operating on steam produced in a BWR (a
214
Refurbishment Techniques for Rotating Blades
direct nuclear cycle) application, Stellite 6B® is not recommended
because Stellite 6B® has a high cobalt content and can become
radioactive when working on steam delivered directly from the reactor. This makes their rework more difficult.
β1
T
C
-dT
β2
b
Fig. 8.3.6—The major characteristics of the
vane profile.
The nose material removal process requires the inlet edge be cut
away, and the cutting process must not generate any excessive
amount of heat. For this reason a normal cold cutting procedure is
required. The material removed must include the entire inlet edge
and any remaining erosion shield. If the shield was attached by
welding, the material removed should be sufficient to eliminate the
HAZ. It is therefore necessary to ensure the following:
•
Any HAZ material from the original attachment should be
removed. This is important if the shield is a “wrap-around”
type, as shown in Figure 3.8.7(d) in chapter 3. The back face
must be examined by hardness testing to ensure all hard
material is eliminated
215
Turbine Steam Path Maintenance and Repair—Volume Two
•
The form of the bar to be attached should account for any
discontinuities in the blade vane. Such discontinuities can
include tie wire bosses or holes, together with local thickening or thinning at the blade tip
The inlet edge preparation can take several forms depending
upon the extent of damage, the form of the blade, the rate of vane
twist, blade material, and the amount of inlet removed. The form to
be used in any application will be selected based upon the requirements of the blade geometry, and previous experience.
In the run-out region at the base of the shield, the radius should
be smooth, and large to the extent it will avoid any significant discontinuities. The final shape will depend upon the form of the shield
and vane at the run-out position.
If the blade vane contains holes (intended to allow tie wires to
pass through) special design considerations are required to allow sufficient material to remain after the weld attachment. Such a preparation is shown in Figure 8.3.7(a), with details of the “cut-out” in Figure 8.3.7(b). Here the selection of radius “r” and thicknesses “d” must
be made to minimize the effect of stress concentration, which could
occur in this region. A blade, in a shutter gauge, with the inlet edge
removed is shown in Figure 8.3.5. Here the removed material can be
seen to be discontinuous above the tie wire hole. This is to allow a
hole to be replaced in the 12% Chrome material (AISI 403) coupon
that is attached in the outer region, and at the same time remove the
HAZ away from the hole that is to be produced.
Inlet edge material of the same form will be removed from each
blade so each will have the same geometric form in the region
where the shield is to be attached. The shield can be of the same
form and, depending upon the vane twist, have the same turning
angle. If some form of weld preparation chamfer is to be used, each
blade must have the same geometry in this area.
216
Refurbishment Techniques for Rotating Blades
’r’
(b)
Gap ’g’
’d’
Gap 'g'
Stellite
bar nose
Contour
around
tie wire
hole
Tie wire
hole
Blade
vane
(a)
Fig. 8.3.7—A blade with its inlet nose
removed, and contoured to avoid the tie
wire hole.
Preparation for the welding process. Before any attempt is made
to attach a new shield by welding, the region on the vane adjacent
to the “cut-out” inlet edge must be cleaned of any deposits or grease.
This is to be done using a suitable degreaser, wire brush, and solvents that do not contain any corrosion-producing compounds.
After production of a weld preparation on the blade vane and
shield, and before the main attachment weld is made, the pre-twisted Stellite 6B® nose is attached to the vane at two or more radial
positions as shown in Figure 8.3.8(a), by a “tack weld.” This attachment and positioning is required to achieve the following objectives:
217
Turbine Steam Path Maintenance and Repair—Volume Two
•
The gap “g” between the shield and vane material is established at the desired value for the entire length. This gap is
critical, and must be checked at the radiused run-out region
at the base of the shield
•
The weld-prep angles “α” on both the Stellite 6B® nose and
the vane must be correct, within a specified tolerance. This
is normally checked by some form of gauge. A typical weld
prep is shown in Figure 8.3.8(b)
•
The shield to be attached must be positioned so that sufficient material is available, that the nose form, with the correct radius and blending to the pressure and suction faces
can be reformed by machining or hand work at completion
of the weld attachment
The tack welds must be spaced and be of an adequate length,
otherwise they will not be able to restrain the shield position during
the full welding process. These welds should be attached first at the
tip section, then on alternate pressure and suction sides of the
vane/shield interface, maintaining gap “g” to the end of the shield.
Tack welds should not be made at the lower run-out radius at the
base of the shield, or opposite any tie wire holes.
To undertake weld attachment, the blade is placed in a welding
fixture, which is produced from compatible material to that of the
vane. This fixture is to provide a precision fit, locating the blade at
several positions on its contour and sufficient to retain both the vane
setting angle “ ξ” and correct radial alignment during the welding
process. The fixture clamping must be designed so as not to interfere
with the welding access during shield attachment.
Prior to welding, both the fixture and blade should be slowly
preheated to a temperature of 400-600°F. (The exact temperature
depends upon the blade material.) This heating is to be undertaken
using an electric furnace. The temperature ramp rate is not to exceed
218
Refurbishment Techniques for Rotating Blades
200°F/hr. The furnace heating should be such that the entire blade
fixture assembly is at the correct preheat temperature.
This preheat temperature should be monitored and maintained
by suitable methods, such as tempil sticks or contact thermometers.
If the temperature falls too much, the entire assembly should be
reheated to the original temperature.
Gap "g"
Positioning
tack welds
Shield of erosion
resistant material
Fig. 8.3.8(a)—The shield with a “tack weld”
holding the erosion shield at the correct
location prior to full weld attachment.
Gap
α1
α2
Blade vane
material
Shield
material
Fig. 8.3.8(b)—The weld preparation on the
vane and shield for the attachment of a full
nose erosion shield.
219
Turbine Steam Path Maintenance and Repair—Volume Two
The welding process. The welding process should be completed
using “gas tungsten arc welding” (GTAW), with a high frequency
start, foot pedal control, and “direct current straight polarity” (DCSP)
capability. The power supply should be capable of supplying an
output up to 250 amps at 20 volts.
The argon shielding gas should be of adequate purity and of
welding grade. The tip size and gas flow rate should be adequate to
fully protect the molten puddle. A thoriated tungsten electrode of
suitable size and tip shape should be used.
The weld attachment of the shield should be completed by the
use of continuous pass weld deposits, using Inconel 82 filler material, deposited alternately on the pressure and suction surfaces. Each
pass is to be started at the tip section, and terminated by a gradual
arc break, which is achieved by manipulation of the foot pedal current decay control.
The crater on the tack ends of each weld pass should be ground
out. Only uncontaminated stainless steel wire brushes and alumina/silicon carbide grinding wheels should be used.
The weld deposit is completed by alternate passes and dressing
on the pressure and suction faces of the vane/shield interface. On
each pass the weld deposit should begin at the tip section, progressing to the base of the shield run-out region.
Attention should be paid to the base of the Stellite 6B® inlet
edge insert run-out. At this position, the weld should run into a block
of steel of the same chemical composition as the blade material, and
the weld should terminate in this material rather than on the blade
vane. A blade with an attached nose is shown in Figure 8.3.9. Here
the inlet edge is to be attached by welding and will be stress relieved
and dressed to the final form by hand controlled grinding.
Post weld heat treatment. At completion of the weld deposit
process, the refurbished blade should be placed into an oven for post
220
Turbine Blading (UK) Ltd.
Refurbishment Techniques for Rotating Blades
Fig. 8.3.9—A blade with the inlet edge
removed, and the bar nose shield ready
for attachment.
weld heat treatment. If the post weld heat treatment is undertaken
only when an oven load is ready, the oven temperature should be
maintained above 400ºF while the blades are loaded, i.e., the welded blades should not lose their preheat prior to stress relief. In the
stress relief oven the blades should be suspended or stood in a vertical position in their fixture (Fig. 8.3.10). This minimizes the probability and extent of distortion during heating. The stress relief process
should be undertaken in a vacuum furnace, to prevent oxidation of
the blade vane and root surfaces.
When the oven is full, the temperature should be raised at a
ramp rate not to exceed 150°F/hr. to a temperature suited to the
stress relief process (1,200 to 1,275°F). The temperature should be
221
Turbine Steam Path Maintenance and Repair—Volume Two
maintained at this value within +/-25°F for the full stress relief cycle.
This temperature is to be continuously monitored, and held for not
less than one hour for each 1.0" of maximum section thickness, or
four hours minimum.
Turbine Blading (UK) Ltd.
When this initial heat soak is complete, the blades should be
cooled to 1,100°F, at a rate of 15 to 25°F/hr. The blade can then be
cooled to 400°F at a cooling rate not in excess of 125-150°F/hr. At
that time the blades can be cooled in still air to room temperature,
being careful to exclude all external drafts.
Fig. 8.3.10—A blade after nose attachment
ready for stress relief.
Inlet edge profile finishing. When the welding and stress relief
operations are complete, the weld deposit area and inlet edge must
be dressed to achieve the final profile requirements. This shaping
should be accomplished by using uncontaminated alumina/silicon
222
Refurbishment Techniques for Rotating Blades
carbide wheels. The surface finish should be consistent with the surface on the original blade, and have a main direction that is radial.
The blades are to be mounted to the profile fixture and checked
for conformance to their original geometry. The inlet edge refurbished portion of the vane can be checked during the final shaping
process by means of specially prepared profile gauges.
Weld inspection. Each blade that has been refurbished should be
100% inspected by liquid penetrant testing techniques and radiography. This testing should be undertaken in accordance with the requirements of American Society of Mechanical Engineers (ASME) standard.
Turbine Blading (UK) Ltd.
Figure 8.3.11(a) shows a blade before weld refurbishment, and
Figure 8.3.11(b) shows the same blade after attachment of a complete nose of Stellite 6B®. Here the nose inlet form has been completely restored.
Fig. 8.3.11—In (a) is shown a last stage blade with evidence of moisture impact erosion
on the outer flow sections. In (b) the damaged portion has been replaced with a solid
bar nose of Stellite 6B®.
223
Turbine Steam Path Maintenance and Repair—Volume Two
Final moment weighing. When this entire procedure is complete, the blades must be moment weighed, and depending upon the
change in mass, a new disposition on the rotor may be required. This
is influenced to some extent by the root or fastening form, and the
need to return the blades to their original rotor position.
Weld attached inlet edge, and build-up of
base material
A special case of the welded inlet edge shield occurs where there
has been extensive damage caused by erosive penetration into the
blade material below, or under the shield. This blade material,
which had been removed, provides the shield-connecting surface,
on which a weld preparation might be required. In this case the
blade material must be restored prior to attachment of the Stellite
6B® nose shield.
Figure 3.8.12 (chapter 2) shows a blade that has suffered extensive damage as the result of erosive material loss from the outer section of the blade, after losing the braze-attached outer portion of the
erosion shield. In this instance weld refurbishment was undertaken
in two separate steps. First, a coupon of material that contained the
chemical and mechanical properties consistent with those of the
blade material was attached. The second step attached a solid inlet
edge shield of Stellite 6B®. The steps in this attachment are shown
in Figure 8.3.12. The requirements for vane geometry, weld preparation, welding, the post weld heat treatment, and inspection apply
to both phases of the process.
It is also possible, if relatively small amounts of vane material
have been lost, to re-establish the blade by weld deposit using a filler
rod consistent with the material properties of the blade material.
However, there are possibly other options, such as removing slightly larger portions of the vane until an acceptable base on which to
attach a shield is obtained.
224
Turbine Blading (UK) Ltd.
Refurbishment Techniques for Rotating Blades
Fig. 8.3.12—The three steps in the repair of the
blade shown as Figure 3.8.12.
If blades are joined into groups by weld attachment or some
other process, this can make the refurbishment more complex. The
groups can be broken into individual blades, which must then be
reconnected after refurbishment, or the refurbishment can be completed by weld repairing the groups. Figure 8.3.13(a) shows a welded group of four blades that are damaged locally by secondary erosion at their inlet edge. In Figure 8.3.13(b) this same group has been
restored by the weld attachment of solid Stellite 6B® inlet edges to
each blade, without breaking down the groups. After weld attachment the requirements of stress relief and geometric adjustment must
be observed.
225
Turbine Blading (UK) Ltd.
Turbine Steam Path Maintenance and Repair—Volume Two
Fig. 8.3.13—Shown in (a) is a four blade group with eroded inlet edges. In (b) these
same type of groups after weld repair with a solid stellite nose.
Similarly, in further developments, it is now possible to weld
repair the blades without their removal from the rotor. This has significant advantages in terms of time and cost. However, the quality
control process requires strict observance, as does the delicate control of technical processes, in terms of shaping and stress relief.
Figure 8.3.14(a) shows a blade that has lost a considerable portion of its tip section inlet edge by moisture-impact erosion. Figure
8.3.14(b) shows this same blade after the damaged material is
removed, and the blade is rebuilt. Figure 8.3.14(c) shows the coupon
that is to be weld attached to the blade material. Figure 8.3.14(d)
shows the results of coupons being weld attached, and in Figure
8.3.14(e) is the rotor at completion of the repair process.
226
Refurbishment Techniques for Rotating Blades
Fig. 8.3.14(a)—Exhaust blades which have suffered a
heavy material loss in the outer flow sections.
227
Turbine Steam Path Maintenance and Repair—Volume Two
Fig. 8.3.14(b)—The blades of Figure 8.3.14(a) after removal
of the damaged tip material.
228
Refurbishment Techniques for Rotating Blades
Fig. 8.3.14(c)—The partially rebuilt blades, and the coupon that is to be attached.
Fig. 8.3.14(d)—Stress relief of the blades after coupon attachment and vane shaping.
229
Turbine Steam Path Maintenance and Repair—Volume Two
Fig. 8.3.14(e)—The rotor after completion of the repair.
Braze attached shield
In both the power and the industrial sectors of steam turbine
repair of erosion damage, the most commonly encountered shield
replacement technique is removing the old, damaged shields, and
replacing them by brazing on new shields.
This is an operation that can be undertaken in the field, even
with the rotor sitting in the lower half of the casing. However, if the
rotor is still in the casing, ensure that excess braze material and flux
do not drop into the casing. Also ensure that all old braze material
is removed from the blade before brazing new shield material.
There are various requirements for achieving an acceptable
braze jointing of an erosion shield to the inlet edge of a blade.
Regardless of the procedure or braze materials used, it is essential to
conform to the following basic requirements, which are common to
all procedures:
230
Refurbishment Techniques for Rotating Blades
•
The surfaces to be joined must be clean—these surfaces must
be prepared by removing any oxide surface scale, grease, oil,
or dirt. Cover the surface in such a manner that the formation
of oxides is prevented during the brazing process. An oxide
scale would prevent the formation of a strong joint, and the
oxidization would be accelerated by the application of higher temperatures
There are various methods of cleaning and removing oxide
scale. The most common method in blade application includes solvent cleaning, using such products as chlorinated hydrocarbons and
petroleum-based solvents. The use of alkaline and acid cleaning is
also common. Alkaline cleaning is acceptable, but is often a slower
process, and is more suited to producing a smaller number of joints.
Acid or mechanical cleaning is a successful process for the range of
stainless steels used for most blades. However, it is necessary to control the concentration of any acid used and to be sure all traces of
the acid are removed from the surface when cleaning is complete,
and before brazing.
•
The surface can also be cleaned by blasting—however, the
blasting medium used should be one that does not leave a
deposit on the surface because this could impede the brazing
process. Nonmetallic blasting media should not be used to
clean the surface, unless special care is taken, and some other
cleaning or washing is used to remove any surface deposits
In addition, any surface that cannot be cleaned should be considered as non-bonded, and the holding strength will be limited to
those surfaces that are cleaned.
•
Fluxing of the joint surfaces must be undertaken to promote
an acceptable bond—the primary function of the flux is to
combine with, and even stop the formation of products that
would prevent an acceptable joint from being formed during
the brazing process. Secondary requirements or advantages
231
Turbine Steam Path Maintenance and Repair—Volume Two
to the use of flux include the formation of a barrier, which
helps prevent an oxide scale from forming on the braze surface, and also the formation of a sealer cap on the joint to
keep it clean and lower the heat loss immediately after completion of the brazing process
•
Preheating of the joint to the correct brazing temperature must
be achieved—this preheat temperature is selected so the heat
in the blade metal is sufficient to melt the braze material, but
not so high as to heat the blade material to the extent its
mechanical properties are influenced by a change in its metallographic structure. The steels used for long low-pressure
stages, are those most likely to be affected by brazing; they
tend to be the fine-grained, low carbon martensitic, or precipitation hardened types. The primary precaution is to avoid
heating the blade material too rapidly, and to a temperature
that will modify its material structure causing a hardening,
which could eventually result in stress corrosion cracking
The braze filler material must receive heat from the blade material to melt it. If the filler is melted by heat from the brazing flame,
then it will cool immediately after the flame is removed, because
heat will soak into the blade material from the molten filler. This will
prevent a flow of the filler and a suitable bond will not be formed
between the joint faces.
•
232
Joint geometry is of critical importance as its strength is
dependent upon the thickness of the braze material connecting the surfaces—in turbine blades the normal direction of
the load is to produce a shear stress in the braze joint material. The approximate relative strength of the joint is shown in
Figure 8.3.15. In this case the optimum gap between surfaces
would be between 0.002" and 0.004". This is a difficult clearance to maintain, and fixtures and clamps are often employed
Refurbishment Techniques for Rotating Blades
Fig. 8.3.15—Braze joint strength as a function of braze gap.
It is common for the braze attachment of erosion shields to use
a “wafer” of filler material that is about 0.003" thick. However,
clamping may still be required, but this must be applied carefully,
because as soon as the braze material wafer melts, there is a tendency for the shield that is being attached to move to an equilibrium
position that may not be optimum in terms of gap clearance. This is
particularly so because the shield is thin, and required bending to
match the form of the vane inlet edge.
•
Post brazing activities for martensitic steels are minimal—
however, they are important and should be followed, as they
will normally improve the total quality of the joint and therefore blade row
It is important to carefully cool the blade after the brazing
process. Ensure the blade is brazed in a facility that is not subject to
excessive drafts where sudden chilling of the joint could occur. It is
often better to clad or cover the joint with some form of insulating
material to slow the cooling process. It is also necessary to clean and
remove any excess braze material from the surfaces of the joints.
Cleaning of the joint at completion of brazing is important to remove
233
Turbine Steam Path Maintenance and Repair—Volume Two
both excess flux and excess braze material, particularly the material
that has deposited as beads in the region of the joint.
Cleaning should be completed as soon as possible after the brazing and cooling process. Clean by washing the surfaces with a jet of
wet steam or hot water, or by mechanical means, using a wire brush
or fine emery cloth.
•
Post braze inspection is important to ensure the joint is
strong, and meets engineering requirement—in the case of
brazing erosion shields, an ultrasonic testing (UT) inspection
will give an indication of the percentage of attachment that
has been achieved. Such UT inspection is not possible however, if the blades are still mounted on the rotor, and the
process must be controlled to ensure acceptable quality.
Should the braze surface contain holes, as seen in Figure
8.3.16, they should be filled by hand puddling, as they represent locations where corrosive elements can collect
Braze attached shield, with base metal build-up
If an existing shield is attached by brazing and the shield has
been penetrated locally, or has been lost, and the operator wishes to
retain the brazed shield form, it is possible to re-attach a new shield
in the field.
Depending upon the amount of blade material lost (in terms of
the remaining braze attachment area), it is possible to rebuild the
vane material in-situ, by making small weld deposits of a suitable
filler material. The filler material must be consistent with the parent
material of the blade, or an Inconel filler can be used. At completion
of this weld deposit and surface dressing, a shield is braze attached
to the restored vane.
If this form of repair is to be undertaken, having lost only one portion of a segmented shield, it is necessary to remove the entire shield,
234
Refurbishment Techniques for Rotating Blades
Fig. 8.3.16—Holes at the shield/vane interface of a
brazed on erosion shield.
re-profile and clean the entire inlet edge, and attach a complete new
shield. If the new shield is to be reattached by brazing, it may be necessary to weld rebuild and stress relieve the section locally before
attaching the new shield. If the new shield is to be attached by welding, then the repaired vane and shield can be stress relieved together.
235
Turbine Steam Path Maintenance and Repair—Volume Two
There are two forms of geometry that are common in braze
attached shields; these are shown in Figure 8.3.17. The “build-up”
and “replacement” procedures used depend upon the form of shield
used on the original blade. There may be situations where blades
with an initially recessed shield will be repaired in-situ, and the final
shield form results in the proud nose being used. This is because it
is difficult to construct the shield recess on a blade vane with it
mounted on the rotor. The disadvantages of the aerodynamic losses
associated with this form of shield could be more than offset by the
savings in time and cost of in-situ repair.
Fig. 8.3.17—Forms of the braze attached erosion shield.
Raw weld deposit
As a temporary measure, a “raw weld” stick can be deposited on
the inlet edge of blades to provide protection until more appropriate
236
Refurbishment Techniques for Rotating Blades
measures can be taken. This repair does not represent a total solution.
The material that is deposited can cause “carbon depletion” at the
blade material/deposit interface, making the weld brittle and susceptible to cracking. This form of repair can be effected by the use of a
Stellite 6B® weld rod, and with the blade “in-situ.” It is necessary to
preheat the blade inlet edge, bringing it to a temperature suited to the
parent material on which the weld is to be deposited. The Stellite
6B® rod is then deposited using an oxyacetylene technique. A stress
relief operation is necessary upon the blade/weld deposit. An important feature of this repair is control of the cooling cycle after the stress
relief heat soak is complete. With this form of protection there is a
high possibility of microcracking during operation. The bar nose inlet
edge repair is preferable, as it is more durable and it lessens the risk
of cracking. The oxyacetylene repair is not recommended for more
than about one year of service, but may be useful in extending the
life of a blade row until replacement blades are available.
Figure 8.3.18 shows 44" 1,800 rpm exhaust stage blades that
were temporarily weld repaired to prevent serious erosion penetration until new blades could be made available. After removal, the
blades would be available for permanent repairs.
Dependent upon the amount of raw weld material that has been
deposited, the inlet edge can be re-profiled. This procedure is most suited to making repairs for heavy secondary erosion, where local penetration of the shield has occurred, and the operator is not able to replace
the entire blade, or even replace the entire shield at that outage.
A major advantage of this form of repair is that it can be undertaken within a relatively short time, say during a normal maintenance outage, and can provide protection of the blade so it can be
permanently repaired at a later date. It will also allow the unit to
continue to produce full power. (The exhaust stage can account for
up to 10% of the unit output.)
237
Turbine Steam Path Maintenance and Repair—Volume Two
Fig. 8.3.18—Raw weld deposit on long exhaust blades.
Dressing an eroded surface
There is always a temptation for the operating engineer to dress
smooth an eroded surface if penetration damage has occurred, but
the shields do not require any refurbishment, or replacement. This
should not be attempted. By dressing the material, the rough surface,
238
Refurbishment Techniques for Rotating Blades
which captures and retains a portion of the impacting water, is
destroyed. It is these “retained pools” of water held at the surface by
the irregularities that cushion subsequent impacts, lowering the
material loss rate.
Repair of off-shield erosion (on the vane)
As described in chapter 3, there are situations where the erosion
damage extends beyond, behind, or below the erosion shield. This
type of damage at serious levels is not common, but for a welldesigned blade and turbine, operating at rated steam conditions with
a suitable shield, such damage should not occur. A shield replacement program, as described in this chapter, often extends coverage
because the eroded portion of the shield and the blade HAZ is cut
away. However, the refurbisher of the blades does not remove more
material than is necessary to achieve an acceptable weld joint.
For this reason, it is not always certain that protective coverage
will be extended. If an examination of the blades indicates that the
potential for “off-shield” erosion exists, and the new inlet edge form
will not correct this condition, even with its extended coverage, then
alternative shield forms can be considered.
A full inlet edge repair can more effectively cover the following
forms of “off-shield” erosion.
Beyond-shield erosion. This form of erosion can be protected
against by the use of an extended width shield. However, blade
mechanical characteristics should be investigated for changes. It is
normal when a refurbisher attaches full inlet edge shields to perform
vibration tests to determine characteristics before attachment of the
shield, and then test it again after attachment to ensure there has been
no significant change in the fundamental, or lower order harmonics.
Below-shield erosion. If there is significant erosion below the
shield, there could be advantages to extending the length of the
239
Turbine Steam Path Maintenance and Repair—Volume Two
shield down the radial length of the vane. The required extension
can be judged from the level and pattern of erosion on the blades
that are to be refurbished. Considerations of changes in vibratory
characteristics are the same as for the “beyond-shield” correction
described above. Before making a decision to extend a shield, the
damage should be assessed in terms of the hours of operation on the
blade, and the probable consequences of not making an extension.
Between shield segment erosion. This type of damage occurs on
blades with a brazed multi segment shield. This will normally be
corrected by the use of a full inlet edge, full-length shield.
Pressure surface erosion. Should this form of erosion occur, a
full inlet edge (bar nose) shield would possibly rectify this condition,
depending upon the extent of the erosion on the pressure surface.
Such damage is unlikely to be a cause for major concern.
Tenon erosion. The tenon erodes as a consequence of moisture
centrifuged to the outer diameters of the blade row, which rebounds
in the radial space between the coverband and the casing inner surface. Such erosion is shown in Figure 3.8.14 in chapter 3.
This type of damage cannot be tolerated if the moisture removes
tenon material to the extent that the joint between the coverband and
tenon becomes visible for an extensive length, say 10% of the total
periphery of the tenon. In this case refurbishment must be undertaken. There are several options that should be considered, including
the weld rebuild of the existing tenon material to provide more shear
section, the removal of the coverband and a weld rebuild to form a
new tenon, and undercover brazing or welding to provide greater
attachment strength of the coverband to the blade outer platform.
Figure 8.3.19 shows the result of undercover brazing on a last
stage blade, where tenon material had been lost. If brazing attachment is attempted, it must be ensured the brazed joint is 0.002" to
0.004" in thickness, and has a generous radius to ensure a strong
240
Refurbishment Techniques for Rotating Blades
Fig. 8.3.19—Under cover brazing to attach a coverband where tenon material loss has
been experienced.
joint. It is also possible to undertake further peening of the existing
material. However, it is necessary to ensure this does not “work
harden” the tenon material, making it more brittle. The most appropriate action in any situation will normally depend upon the amount
of joint exposure, and the operating temperature of the stage.
The repair and rebuild of tenons is covered in greater detail in
chapter 9.
241
Turbine Steam Path Maintenance and Repair—Volume Two
MOMENT WEIGHING OF
REFURBISHED BLADES
For longer blades it is important the “moment weight” of the
individual blades is measured, and then a determination made of the
most appropriate distribution of the blades around the rotor rim so
as to minimize the total “out-of-balance” forces. These forces need
to be balanced by the addition or removal of weight from the rotor
itself. This achievement of balance is necessary to minimize the field
balance adjustment that is required on the rotor.
For the longer blades each element is individually moment
weighed, and then an optimum disposition around the rotor is determined. This moment weight data must be provided to the blade purchaser in the form of a drawing or other method, and is delivered as
part of the blade supply. If this information is not provided, the
blades are of no use, and as such the delivery is not complete.
After determination of the most appropriate location of the refurbished blades, it must be considered whether these long blades are,
depending upon the form of the root fixture, best returned to their
original position. In this case the “out-of-balance” forces in the original position should be determined, and if these forces are beyond
those that can conveniently be balanced on the rotor, then some
minor trimming should be considered.
Note: If the refurbished blades have a root form that requires
tangential loading, their assembly may be a little more complex, as
it may be necessary to make a tangential shift so that the closing
blades are in the optimum position. This should not however, make
the mounting order different.
242
Refurbishment Techniques for Rotating Blades
The moment weighing process
There are calibrating/measuring devices available that quickly
and accurately determine the moment weight of individual blades.
These devices have a moment arm that can be adjusted to accept the
individual blades, and read directly the moment weight. However,
the basic principle of moment weighing is relatively simple and can
be undertaken by anyone with a single accurate scale, (calibrated to
measure to at least 0.5 oz.), and two knife edges on which the blades
can be supported in a repeatable manner.
This determination of moment weight, the optimum position in
blading sequence on the rotor, and the residual unbalance can be
undertaken in the following manner.
Blade numbering
Each blade should be assigned a discrete number from “one” to
the total number of blades “Z” in the row. (Note: if extra blades are
supplied as inventory spares, they should be identified and moment
weighed also.)
Blade weighing. Each blade should be accurately weighed, supported on an accurate scale. It is preferable to support the blade vertically, to ensure the weight is applied correctly on the scale pan, as
shown in Figure 8.4.1, and determine the weight “Wb.”
Blade moment. The blades should then be supported on two
knife-edges, as shown in Figure 8.4.2, and the tip reaction “Tt”
recorded for each. In setting up to make these measurements, it is
necessary to ensure the blades are supported at a position corresponding to the discharge root diameter, on plane “X-X.”
Data analysis. After recording these two pieces of data, the following analysis can be made, using the total information:
243
Turbine Steam Path Maintenance and Repair—Volume Two
•
The measured blade weight “Wb.” This must be accurate to
the nearest 0.5 oz
•
The recorded tip reaction “Tt,” with the setup as shown in Figure 8.4.2. This reaction should be accurate to the nearest
0.01#
It is necessary to locate the measuring device at the tip of the
blade as close as possible to the tip section. If the blade has tenons
for the attachment of a coverband, the knife-edge should be located
on the tenons as shown in Figure 8.4.2.
h
Wb
Fig. 8.4.1—Blade weight
determination.
244
Refurbishment Techniques for Rotating Blades
Fig. 8.4.2—Moment weighing the blade.
The fixed knife-edge should be located at the root, at diameter
position “Dr,” i.e., the knife-edge must be located at the same position on all blade elements.
•
The known root diameter “Dr” of the blade at discharge
•
The measured blade height “h” at discharge
The measured weight and tip reaction are as shown in Figure
8.4.2. This diagram shows the distance “Kr” above the root diameter, which represents the position of the center of gravity.
From this diagram, and taking moments about plane “X-X” at the
root diameter “Dr,” the value of “Kr” can be found.
Taking moments about “X-X” gives:
Wb . Kr = Tt . h
Therefore:
Kr =
Tt . h
Wb
245
Turbine Steam Path Maintenance and Repair—Volume Two
Then taking moments about the shaft center “O,” Figure 8.4.2,
the following equation is obtained:
Mw = WB (Dr/2 + K)
Where “Mw” is the moment weight of the blade. Substituting for
“Wb” from the previous equation gives:
Therefore, from knowledge of the measured weight and tip reaction when supported from the root diameter, the effective moment
weight of each blade can be determined.
Example 8.4.1
Consider a blade with a vane height of 18.00 ", mounted to the
rotor at a mean diameter of 60". One element in a group of 60 blades
is weighed and found to have a weight of 10.091#. When supported on a knife-edge the tip reaction is 4.190#. These forces, weights,
and distances are shown in Figure 8.4.3.
Dm = 60.00"
x
18.00
Dr = Dm - h = 42.00"
Kr
0
Tt = 4.190#
x
Wb = 10.091#
Fig. 8.4.3—Moment weight of theFigure
blade in8.4.3
Example 8.4.1.
Moment weight of the blade in Example 8.4.1.
246
Refurbishment Techniques for Rotating Blades
Apply this data to find “Kr” using the previous equation. Therefore:
.h
Kr = TtWb
x 1800
Kr = 4.190
10.091
Kr = 8.474 inches
The moment weight of this blade can be found from the application of the equation for Mw giving:
Determination of tangential position
To select a suitable tangential position of the individual blades,
and minimize the “out-of-balance” force the following procedure is
suggested:
Blade listing. A listing of the blades is prepared, ranking them
from heaviest (moment weight) to lightest. This ranking should
include any available spare blades. The available spares are
removed from the group of blades. (It is suggested blades representing averages for the various ranges of weight available are removed.)
First blade selection. Select the heaviest blade #1 and place it
at the “top dead center” (TDC) position, which is position (1) of
Figure 8.4.4.
247
Turbine Steam Path Maintenance and Repair—Volume Two
Fig. 8.4.4—The determined blade disposition around the wheel.
Blade placement. Place the next heaviest blade #2 opposite
position (1), at the “bottom dead center” (BDC) in position (31).
The next blade #3 is placed next to #2, one pitch “l” counterclockwise.
The next blade #4 is placed adjacent to blade #1 and one
pitch counterclockwise.
The next blade #5 is placed adjacent to the #4 blade in a
counterclockwise direction.
This process is completed placing blades adjacent to the next blade
in a counterclockwise direction until all positions have been filled.
248
Refurbishment Techniques for Rotating Blades
Example 8.4.2
Table 8.4.1 lists 60 blades arranged in descending order of
moment weight, determined by the methods outlined in example
8.4.1. These moment weights vary from 293.474 #-inches to
280.791 #-inches.
Using the method outlined above, the following disposition of
the blades, as shown in Figure 8.4.4 was determined as suitable.
The “out-of-balance” moment resulting from the individual
moment weights from the 60 blades can be resolved into vertical
and horizontal components. This is done as shown in Table 8.4.2,
and as shown, results in a total out-of-balance force “Fh” of 0.8449
#-inches.
Units lb-inches
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
293.474
293.118
292.841
291.887
291.756
291.415
291.340
290.906
290.730
290.464
290.370
289.936
289.873
289.562
289.562
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
289.536
289.431
289.101
289.060
288.663
288.615
288.480
288.469
288.252
288.154
288.079
288.069
287.503
287.503
287.349
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
287.331
287.076
286.994
286.627
286.563
286.239
286.080
285.616
285.605
285.070
285.002
284.471
284.415
283.992
283.988
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
283.550
283.535
283.303
283.276
282.999
282.984
282.606
282.210
281.700
281.408
280.888
280.870
280.765
280.742
280.719
Table 8.4.1—60 Blade Moment Weights
249
Turbine Steam Path Maintenance and Repair—Volume Two
The sum of the individual horizontal forces is found from:
Similarly the sum of the vertical forces can be found from:
Using this information the out-of-balance force “Fh,” its phase
displacement “ω” can be found.
The resultant force “Fh” can be found from the resolution of the
vertical and horizontal components of the out-of-balance forces.
The angle of inclination “ω” of the resultant out-of-balance
forces “Fh” lies in the second quadrant, and can be found from:
The direction of the out-of-balance force “ω” can be found from
resolution of the vertical component of -0.667 #-inches, and the hor-
250
Refurbishment Techniques for Rotating Blades
izontal component of +0.519 #-inches. For example 8.4.2 this angle
“ω,” in the second quadrant can therefore be found from:
This is shown in Figure 8.4.5.
In the matter of determining the most suitable position of a number of blades, programs exist in the facilities of both original equipment manufacturers (OEMs) and other manufacturers, which are
superior to this simple hand method. However, owners with replacement blades available, but no balance data can yield good results
with this method.
+ Vert.
Resultant
= 0.845 #-ins
Hor =
0.519 #-ins.
- Hor.
+ Hor.
ω°
Vert =
-0.667 #-ins.
ω = 37.9 °
- Vert.
Fig. 8.4.5—The resultant “out-of-balance” force and phase
angle ‘ω’.
251
Turbine Steam Path Maintenance and Repair—Volume Two
Posn.
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
Bld
1
59
58
55
54
51
50
47
46
43
42
39
38
35
34
31
30
27
26
23
22
19
18
15
14
11
10
7
6
3
2
60
57
56
53
52
49
48
45
44
41
40
37
36
33
32
29
28
25
24
21
20
17
16
13
12
9
8
5
4
Mom Wt. Angle a
293.4740
0
280.7420
6
280.7650
12
281.4080
18
281.7000
24
282.9840
30
282.9990
36
283.5350
42
283.5500
48
284.4150
54
284.4710
60
285.6050
66
285.6160
72
286.5630
78
286.6270
84
287.3310
90
287.3490
96
288.0690
102
288.0790
108
288.4690
114
288.4800
120
289.0600
126
289.1010
132
289.5620
138
289.5620
144
290.3700
150
290.4640
156
291.3400
162
291.4150
168
292.8410
174
293.1180
180
280.7190
186
280.8700
192
280.8880
198
282.2100
204
282.6060
210
283.2760
216
283.3030
222
283.9880
228
283.9920
234
285.0020
240
285.0700
246
286.0800
252
286.2390
258
286.9940
264
287.0760
270
287.5030
276
287.5030
282
288.1540
288
288.2520
294
288.6150
300
288.6630
306
289.4310
312
289.5360
318
289.8730
324
289.9360
330
290.7300
336
290.9060
342
291.7560
348
291.8870
354
Sin a
0.0000
0.1045
0.2079
0.3090
0.4067
0.5000
0.5878
0.6691
0.7431
0.8090
0.8660
0.9135
0.9511
0.9781
0.9945
1.0000
0.9945
0.9781
0.9511
0.9135
0.8660
0.8090
0.7431
0.6691
0.5878
0.5000
0.4067
0.3090
0.2079
0.1045
0.0000
-0.1045
-0.2079
-0.3090
-0.4067
-0.5000
-0.5878
-0.6691
-0.7431
-0.8090
-0.8660
-0.9135
-0.9511
-0.9781
-0.9945
-1.0000
-0.9945
-0.9781
-0.9511
-0.9135
-0.8660
-0.8090
-0.7431
-0.6691
-0.5878
-0.5000
-0.4067
-0.3090
-0.2079
-0.1045
Res. Hor.
0.0000
29.3455
58.3743
86.9599
114.5777
141.4920
166.3426
189.7219
210.7187
230.0966
246.3591
260.9132
271.6370
280.3009
285.0568
287.3310
285.7749
281.7740
273.9794
263.5295
249.8310
233.8545
214.8439
193.7548
170.2003
145.1850
118.1424
90.0290
60.5886
30.6102
0.0000
-29.3431
-58.3962
-86.7992
-114.7851
-141.3030
-166.5055
-189.5667
-211.0442
-229.7544
-246.8190
-260.4244
-272.0782
-279.9840
-285.4218
-287.0760
-285.9280
-281.2204
-274.0507
-263.3313
-249.9479
-233.5333
-215.0891
-193.7374
-170.3831
-144.9680
-118.2505
-89.8949
-60.6595
-30.5105
Σ 0.5192
Table 8.4.2—Blade Disposition and Resolved Moments.
252
Cos. a
1.0000
0.9945
0.9781
0.9511
0.9135
0.8660
0.8090
0.7431
0.6691
0.5878
0.5000
0.4067
0.3090
0.2079
0.1045
0.0000
-0.1045
-0.2079
-0.3090
-0.4067
-0.5000
-0.5878
-0.6691
-0.7431
-0.8090
-0.8660
-0.9135
-0.9511
-0.9781
-0.9945
-1.0000
-0.9945
-0.9781
-0.9511
-0.9135
-0.8660
-0.8090
-0.7431
-0.6691
-0.5878
-0.5000
-0.4067
-0.3090
-0.2079
-0.1045
0.0000
0.1045
0.2079
0.3090
0.4067
0.5000
0.5878
0.6691
0.7431
0.8090
0.8660
0.9135
0.9511
0.9781
0.9945
Res. Vert.
293.4740
279.2041
274.6296
267.6349
257.3458
245.0713
228.9510
210.7076
189.7320
167.1749
142.2355
116.1660
88.2602
59.5798
29.9607
0.0000
-30.0361
-59.8929
-89.0213
-117.3309
-144.2400
-169.9052
-193.4463
-215.1865
-234.2606
-251.4678
-265.3521
-277.0808
-285.0469
-291.2368
-293.1180
-279.1812
-274.7323
-267.1404
-257.8117
-244.7440
-229.1751
-210.5352
-190.0251
-166.9263
-142.5010
-115.9484
-88.4036
-59.5124
-29.9990
0.0000
30.0522
59.7752
89.0445
117.2427
144.3075
169.6719
193.6671
215.1672
234.5122
251.0919
265.5951
276.6680
285.3804
290.2880
Σ -0.6665
Refurbishment Techniques for Rotating Blades
An alternate means of moment weighing
Other means of determining the moment weight of the blades,
using one or two accurate balances are also available. Figure 8.4.6
shows the required setup for recording the reaction with one or two
balances “Bt” and “Br,” which record the readings of weight “Tt”
and/or “Tr.”
X
Y1
Y2
h
Dr/2
Kr
O
Bt
Tt
Wb
X
Br
Tr
The 'B's are the two balance scales on which the blade is placed,
the measured weights being 'Tt' and 'Tr'.
Fig. 8.4.6—The moment weighing of a blade using two balances ‘Bt’ and “br’.
With only one balance “Bt” available, and taking moments
about the lower balance:
With two balances “Bt” and “Br” available, the value of “Kr” and
“Wb” can be found from:
The sum of “Tt” and “Tb” is the weight of the blade “Wb.”
253
Turbine Steam Path Maintenance and Repair—Volume Two
EROSION SHIELD CRACKS
If erosion shields are attached by welding, there is a possibility
that cracks could initiate in the HAZ, particularly when the weld
filler material that is used for the shield attachment is not sufficiently ductile. These cracks result from both residual stresses and those
induced by rotation. The cracks can also result from possible differences in the expansion coefficients between the blade material, filler
material, and the erosion shield. One of the major advantages of
using an Inconel as weld material is that Inconel is ductile and therefore provides a ductile barrier between the shield and blade material. This weld filler ductility allows for differences in thermal expansion, and the internal stresses associated with temperature changes.
The Inconel can therefore accommodate the consequential plastic
deformation, which otherwise would lead to crack initiation.
Note: However, whenever a weld is made to hard alloys (e.g., an
AISI 410 or Stellite 6B®), a suitable stress relief operation should be
performed.
If AISI 410 or a similar material has been used as the weld filler
material, and a crack initiates in the HAZ, corrective action is
required. This condition cannot be allowed to persist when the unit
is returned to service. The most appropriate action is to weld a full
inlet edge attachment in place of the erosion shield. This is however, a time consuming process, often requiring the blade be removed
from the rotor. An alternative solution is to cut out the cracked
shield, and prepare to replace or otherwise correct the problem at
the next outage.
It is possible to operate blades with complete inlet edges missing
for extensive periods. However, there must be a limit to the number
of blades within a row that can be modified by having shields
removed. It must be considered that the removal of a shield can
modify the natural frequency of the blade, and also affect any blades
254
Refurbishment Techniques for Rotating Blades
with which it is batched. Removing the inlet nose is not considered
ideal; therefore it is prudent to make arrangements for repair at the
first outage opportunity.
BLADE TRAILING-EDGE
EROSION
Trailing-edge erosion is becoming relatively more common,
occurring on the last stage blades of units that have operated at light
loads with operative water-cooling sprays (see chapter 3). This erosive damage that occurs as a consequence of the root recirculation
of water is caused most often from the hood cooling spray water.
In the region of this damage, towards the root section, the blade
vane stresses are tending to a maximum. Therefore, such erosive
damage occurs at a radial location on the vane, where stresses, due
to both direct centrifugal loading and dynamic loading due to steam
and centrifugal bending effects, are high and can be close to their
limiting values in terms of achieving a suitable “factor of safety.”
When trailing-edge erosion is observed, a first action should be to
examine the spray nozzles at the outer edge of the exhaust blades to
ensure they are orientated in their design direction, and don’t spray
directly towards the blades.
If this form of damage is found, the depth of penetration should
be measured, and the condition monitored with depth readings at
repeatable locations as often as possible. Review the extent of penetration to determine the depth to which the gouging (erosion cutting) has penetrated into the blade material. A critical consideration
of this damage is that the root section discharge edge is, by design,
relatively thin. Therefore if grooving exists, it will lead to stress concentration, and failures can be induced.
255
Turbine Steam Path Maintenance and Repair—Volume Two
No efforts should be made to weld repair these blades, since the
gouging exists in a high stress region, and the elimination of residual
stresses cannot be guaranteed. While a crack, should one exist, may
appear to be successfully repaired, further cracks will probably form
and propagate from the welded region soon after the unit is returned
to service.
Similarly no efforts should be made to hand dress any damage
that is found. If cracks do exist, the blades should be considered in
need of replacement. While stages have continued to operate with
extensive trailing-edge erosion, this is not considered an acceptable
practice, and eventually these elements will fail and the unit will be
forced from service. Blades should be purchased as soon as it is
determined that the grooves are progressing at a rate allowing the
groove to penetrate the complete thickness of the discharge edge
replacement.
Note: As a general rule, if penetration has proceeded to a depth
of 1/16" (or half the trailing edge thickness if the blade trailing edge
thickness is less than 1/8") then the blade should be monitored, and
preferably replaced at the first opportunity. This, although an expensive option, is probably far less onerous to the operator than the possibility of massive blade failures in the exhaust stage and the consequential damage to both the turbine and condenser. There is no firm
technical basis for this recommendation. However, at this time material rupture has not been found prior to these conditions having been
reached.
It cannot be predicted at what radial position on the trailing-edge
erosion that cutting and cracks will initiate. Knowledge of the natural frequencies the position of “nodal lines” and regions of maximum
stress is probably valuable. However, it is difficult (if not impossible)
to establish all of this data with the blades mounted to the rotor, as
individual differences on the blade vanes will modify these positions
from blade to blade.
256
Refurbishment Techniques for Rotating Blades
When this condition of erosion is found, there is every reason to
monitor its progress, and possibly purchase replacement blades for
installation at a suitable opportunity.
SOLID-PARTICLE EROSION
BY OXIDE SCALE
This form of damage is normally most severe in terms of material loss in the stationary blade rows (see chapter 4). However, the
rotating blades do suffer material loss, and such loss needs to be
monitored to estimate the rate of deterioration, and the point in the
maintenance cycle at which replacement blades should be made
available.
The damage that occurs on the stationary vanes and sidewalls
tends to cause a deterioration in expansion efficiency, whereas the
damage sustained to the rotating blades will cause a deterioration in
efficiency, and can also expose the rotating blade to mechanical
deterioration, possibly leading to failure. There are three distinct
areas where rotating rows should be examined for damage, and
some effort made to quantify and record the conditions.
On the rotating blade inlet edge. At this location there is material loss due to the “gouging action” of the scale particles as they
flow through the stage. A portion of these particles pass between the
blades, cutting and removing material from the vane inlet edge suction surface as they do. Another portion of the particles rebound
between the stationary and rotating blades again, removing material
from the inlet edge and nose. In general this damage is apparent, and
is more severe towards the tip, or outer flow sections of the blade.
257
Turbine Steam Path Maintenance and Repair—Volume Two
On the rotating blade tip section suction face. When solid-particle erosion damage is found on a stage, it is necessary to examine the
pressure face just under the coverband. The design of coverbands used
on control stages is often an integral coverband, forming an inner band
with an outer coverband and secured by tenons formed on the outer
surface of the inner coverband. This form of design is shown in Figure
8.7.1. With this form of damage the inner integral coverband is “undercut” at the tip section just below the integral portion of the inner
coverband, as shown in Figures 4.8.9 and 4.8.10 in chapter 4. This
damaging effect weakens the attachment of the coverband system,
increasing both stress and stress concentration in this region, and modifying the natural frequencies of the blades. The dimensional modifications will also have an adverse influence on stage efficiency.
Casing
Diaphragm
Outer
Ring
Cr
Rotating
Blade
Vane
x
Stationary
Vane
L1
Ca
L2
Fig. 8.7.1—The double cover, with the inner band
forming an axial seal, and radial seals on the outer
band.
Damage to the tenons attaching the outer coverband. The scale
that is centrifuged out from the stationary blade row, or reaches the
casing sidewalls after repeated impacts between the stationary and
rotating blade vanes, will pass over the blade coverband, causing
erosion damage on the rivet heads formed from the tenons.
If fox-holed tenons are used, as on many control stages, this
becomes significantly less of a problem as the coverband provides
protection for the tenon heads.
258
Refurbishment Techniques for Rotating Blades
If the coverbands do not include a fox-holed attachment, this
should be considered as a corrective action. However, to make this
modification it is necessary for the coverband to have adequate radial depth to allow attachment. Or the coverband thickness can be
increased to allow a foxhole to be used. In either case an analysis of
the stresses induced should be considered.
If the material loss at the tenons is occurring on a row that does
not have an integral inner coverband, the covers can be secured by
the use of undercover brazing, or welding the underside of the coverband to the blade tip. This can, if applied correctly, strengthen the
attachment, and allow the rotor to be returned to service.
These forms of solid-particle erosion deterioration on rotating
blade vanes are a form of damage that cannot at present be repaired,
but weld deposit can be applied to rebuild tenons.
Preventive and corrective actions
These forms of solid-particle erosion deterioration of stationary
and rotating blades are a type of damage that cannot, at present, be
repaired except to apply welding to rebuild tenons. However, there
are methods available to make the rotating blades more resistant to
this type of damage, and when damage is present to lower its rate of
progression. These include:
•
The use of erosion-resistant coatings on the blade vanes and
sidewalls. This layer provides a hard, more erosion-resistant
surface. This is achieved by coating the affected surfaces
with a thin layer of material that is much harder than the
blade material, and is therefore able to resist the erosive
effect of the scale. These coatings modify the blade material
substrate by diffusing elements from the coating into the
blades and sidewalls
259
Turbine Steam Path Maintenance and Repair—Volume Two
•
To operate the unit in a full arc admission mode, at all loads.
This reduces nozzle and blade erosion by distributing the
particles to all passages at a lower velocity. It also reduces
the high impact forces developed on the weakened control
stage blades. However, this does introduce significant efficiency losses into the unit, particularly those that operate at
partial loads for extended periods
•
Changes in inlet stage nozzle geometry have been found to
accelerate the scale most effectively, and help minimize the
material losses. These changes include increasing the axial
gap from stationary row discharge to rotating row inlet. Also,
modifying the nozzle passage outer sidewall contour has
been found to allow the scale to accelerate closer to the
steam velocity
By increasing the gap between the nozzle discharge and the
rotating blade inlet, there is greater axial space for the scale particles
to accelerate; therefore they have a less severe impact effect on the
following rotating blades.
•
The introduction of blades designed to modify the distribution of the steam flow that carries the particles into the blade
rows, and therefore modifies the effect of impact
These various and other methods of reducing this damage will
be considered. However, at this time this is an evolving science and
rapid changes in technique are occurring as experience is gathered.
It is unlikely that any one solution to this problem will be found in
the immediate future. However, both the operator and designer can
implement certain actions and changes to limit the extent of this
damage within any time frame. The actions to be used are based on
palliative measures in both the boiler and turbines. These include
the following:
260
Refurbishment Techniques for Rotating Blades
Acid washing of the boiler. This is the periodic chemical cleaning of the boiler tubes. This cleaning is expensive, but experience
has indicated that in their new condition it will normally require
about five years of operation to produce sufficient scale on the tubes,
which will detach and cause the damage noted on the stationary and
rotating blades. If chemical cleaning is employed, it will need to be
repeated at about five year intervals.
The initial cleaning is the most expensive for the owner, as the
initial wash requires the engineering and installation of cleaning
lines. This piping system, once installed, can be reused on subsequent washes.
Improved boiler tube materials. This is the use of austenitic (or
similar) steel that provides a resistance to the formation of thick
oxide scale on the boiler tube surfaces. This steel would be used for
the superheater and reheater tubes. Unfortunately, there are disadvantages associated with the use of austenitic steels, in that their heat
transfer capabilities are inferior to those of the low carbon steels, and
therefore a larger heat transfer surface is required in the boiler.
There are systems now available for applying a high chromium content deposit on the surface of the boiler tubes. These treated tubes are better able to resist scale formation for considerable
periods, and can be retained for the life of the normal unit.
Processes exist for “chromizing” new units and a “chromate”
treatment of existing units. While these processes can be an
expensive investment, the probable cost recovery in terms of
reduced maintenance and efficiency losses could soon allow the
investment to be recovered.
Use of full arc admission. The use of full arc admission at unit
start-up ensures scale admission is more evenly distributed around
the entire inlet annulus of the first stage nozzles. It also tends to
lower the steam velocities at entry to the rotating blades. The principle time for scale exfoliation is at start-up, particularly at cold starts.
261
Turbine Steam Path Maintenance and Repair—Volume Two
Therefore, if during the start-up period the scale can be distributed
around the complete inlet annulus, damage levels will be reduced.
Many examples exist showing that erosive damage is most severe in
the nozzle or stationary blades that are the first to admit steam to the
turbine section. This is certainly because these stationary blade elements pass the greatest amount of scale that is more abundant at
start-up. This severe damage can be seen in chapter 4, Figure 4.8.1,
for an austenitic material, and in Figure 4.8.5 for martensitic vanes.
Many manufacturers now recommend and are designing bypass
systems into their units. This system allows the initial steam to bypass
the steam path at start-up, and flow directly to the condenser. This
“bypass operation” is continued until the danger of scale exfoliation
and carryover has been reduced.
Most large, advanced condition units are now built to operate at
full arc admission at start-up, thereby admitting steam to the entire
ring of the first stage stationary blades. This action allows the total
damage to be spread around the entire inlet ring, rather than be concentrated in one area, or segment, of the inlet annulus.
Monitoring boiler tube temperature. It is important to carefully
monitor boiler tube temperature, and limit it to a level that will prevent or minimize oxide scale formation. This requires rigorous control
of temperature in the superheater and reheater walls to provide assurance that scale formation is minimized. With modern computer control methods on new units, this is an easier objective to achieve.
Vane material improvements. Attempts have been made to manufacture the stationary blades and nozzles of a more erosion-resistant material. Thus far, no steel has been found with superior resistance to the extent it can be considered a complete solution. However, some success has been achieved by the use of Stellite 6B®
inserts on some vanes that have been weld repaired. However,
unless the filler material has a comparable hardness, the HAZ will be
attacked and lose a disproportionate amount of material.
262
Refurbishment Techniques for Rotating Blades
Nozzle geometry changes. Changes in inlet stage nozzle geometry have been found to accelerate the steam most effectively and
help minimize material losses.
Axial gap increase between blade rows. By increasing the axial gap
between the nozzle discharge edge and the rotating blade inlet, there is
greater axial space for the scale particles to be accelerated to a velocity
nearer that of the steam. The increased velocity of these particles has a
less severe impact effect on the following stage rotating blades.
The use of protective coatings. These coatings are used to coat
or impregnate the surface material of the vanes, providing a harder
more resistant surface.
There has been considerable activity in identifying and providing a coating that can be applied to both the stationary and rotating
blade elements, which will provide a surface better able to resist the
erosive damage. Various coatings capable of providing an acceptable level of protection are available, from work undertaken by Electric Power Research Institute (EPRI), and other companies.
The rate of erosion is considered to be influenced by several factors, which are beyond the control of operators and possibly even
engineering design:
Scale particle size. The rate of material removal from surfaces is
influenced by the size of the particles impacting them. Size does
affect metal removal rates, erosion being almost insignificant with
particles smaller than 5 microns, increasing to a maximum with particles of a size 50 to 100 microns.
Particle hardness. Particle hardness also influences erosion loss rate,
and varies at about the 2.3 power of scale hardness. The temperature of
the blade also has some influence on erosion damage. It has also been
suggested that on impact, there is a localized heating and annealing, or
even melting at the point of impact. If this were so, it would tend to negate
considerations of other environmental temperature effects.
263
Turbine Steam Path Maintenance and Repair—Volume Two
These considerations do little to cause turbine designers and operators to feel they can place any great confidence in designing away
from the solid-particle erosion phenomenon. Design considerations
are such that small vane angles are sought to reduce thermodynamic
and aerodynamic losses, and maximize stage efficiency. This requirement aggravates the material loss by solid-particle loss impact. Present
knowledge suggests that careful monitoring of the level of deterioration of both the stationary and rotating blades is important to help plan
blade replacements and repairs (by welding). Planning for a convenient outage reduces the negative effect on system security. It is necessary to ensure deterioration does not occur to the extent a forced outage condition is induced.
EROSION RESISTANT
COATINGS
There are a number of materials and processes available for coating stationary and rotating blades, which are intended to extend their
life in an erosive environment. These coatings can improve the material’s surface resistance to solid-particle erosion by producing a protective layer of material, which is able to resist the damaging effects of the
scale. These various coatings can be applied by different processes. It
is the combination of the various materials and processes that introduce the different levels of protection available to the industry. The
maintenance engineer should consider these combinations when the
use of such a coating is being evaluated.
There has been considerable research to identify and provide a
coating that can be applied to both the stationary and rotating blades,
while providing a surface better able to resist the erosive damage. Various coatings are available, from work undertaken by EPRI and other
companies, capable of providing protection suitable for application to
both stationary and rotating blade elements.
264
Refurbishment Techniques for Rotating Blades
Coating technology applies protection through the modification of
the blade material surface by application of substrate chemistry. It does
this by two alternative methods:
•
Diffusion coatings (diffusion alloying)—This is the diffusion of a
resistant material into the substratum of the area to be protected. This is a surface conversion process in which the surface
substratum reacts with the diffusion material, normally a boride
compound, to form a more protective layer. While there are a
limited number of such materials that can be applied at this
time, these processes offer a cost effective means of covering
complex forms. The diffused coatings tend to be life limited in
many turbine applications because the coatings tend to be thin
(0.001-0.004" ), and for long cycle times between outages they
may be considered to have limitations. However, the increased
durability of the nozzle greatly reduces blade damage
Research continues on these diffusion processes and materials, and
there is the probability that suitable long life materials and processes
will eventually become available.
•
Overlay coatings—This is coating the surface to be protected
with a layer of material that adheres to it. The deposited material does not rely upon reaction with the blade material substrate, although normally there is some. These overlays make
available complex coating materials, which can offer high
resistance to erosion
The basic elements used for coatings are, at higher temperatures
aluminum compounds, which form an aluminum oxide, Al2O3, and at
lower temperatures chromium compounds, which form a chromium
compound, Cr2O3. For solid-particle erosion protection, chromium
carbide is currently the most popular and effective choice.
While all the technologies discussed below may not be currently applied to steam turbine parts, they are mature processes and are
265
Turbine Steam Path Maintenance and Repair—Volume Two
available if and when needed. The protection technologies that are
currently available include:
Electron beam physical vapor deposition. The basic elements
used for coatings are, at higher temperatures aluminum compounds,
which form an aluminum oxide, Al2O3, and at lower temperatures
chromium compounds that form a chromium compound, Cr2O3. For
solid-particle erosion protection, chromium carbide is currently the
most poplar and effective choice.
By robotic control it is also possible to achieve a uniform surface,
or to vary thickness to increase coverage in the more susceptible areas.
Plasma coating process. Like the electron beam vapor deposition process, the plasma coating process is an overlay “line of sight”
treatment, in which the material to be deposited (a powder form) is
heated above its melting point and accelerated towards the surface
to be covered. As the powder particles impact the surface to be covered, they form a series of “splats,” which overlap and fuse to each
other, and are attached to the surface by mechanical bonding.
While the temperature of the plasma may reach 50,000°F, the
coated part remains relatively cool during the coating process and
therefore the mechanical properties of the blade material are not
affected. This process can also be undertaken in a vacuum, which
improves the coverage and purity of the material deposited. Figure
8.8.1 shows the plasma coating process, and Figure 8.8.2 a 200x
etched surface of CoCrAlYSi coating.
Chemical vapor deposition. This process allows metallic, intermetallic, and refractory compounds to be deposited. At this time this
process has little application to steam turbine components for solidparticle erosion protection, but the ability to cover complex shapes
at a relatively low cost may make it applicable at some future time.
Pack cementation. Pack cementation is a method in which the
parts to be protected are packed into a mixture of powders contain-
266
Refurbishment Techniques for Rotating Blades
Fig. 8.8.1—The plasma coating process.
ing aluminum or chromium, and then heated to a temperature for a
period of time sufficient to form the protective coating required. This
method is employed for diffusion alloying.
267
Turbine Steam Path Maintenance and Repair—Volume Two
Pack cementation allows complete coverage of all surfaces in
which it is required to achieve coverage. This method is economical
and areas that are not to be covered can be masked. Figure 8.8.3
shows a packed retort containing the parts to be coated, with the
retort and parts being loaded into a furnace for heating.
Fig. 8.8.2—A 200x section of a plasma
coat.
Gas phase coatings. This process is similar to the pack cementation process, except the component to be coated is not surrounded
by the powdered mixture. In this process the powder mixture is converted to the gaseous phase and made to surround the component at
an elevated temperature.
This process is limited to aluminum and chromium coating, but
it does allow control of coating thickness and microstructure by
eliminating any reliance on the contact areas, as is required with the
pack cementation process.
268
Chromalloy
Refurbishment Techniques for Rotating Blades
Fig. 8.8.3—The pack cementation process; the samples to be coated are loaded into a
furnace.
Considerable research is being undertaken on coatings suitable
for steam path components. Progress has been made in terms of
establishing acceptable coatings based on the cost, suitability of coverage, and life. One material has been reported to offer superior
properties, and its wear resistance is superior to many other product
coatings. It is about 35 times more resistant than an uncoated AISI
422 material. Figure 8.8.4 shows a curve of comparative resistance
for various coating, and the base material AISI 422. The coating
material used in the EPRI material SPE-8515-HT, is a complex
chromium-containing compound (FeCrAlY), which is plasma
sprayed onto the component, and is 0.008-0.015" thick, and is suitable for protecting both stationary and rotating blades.
269
Turbine Steam Path Maintenance and Repair—Volume Two
Fig. 8.8.4—Comparative solid particle erosion resistance rates for various coatings.
This EPRI coating (designated SPE-8515-HT) is produced by a
plasma spraying process, and after deposition it is heat treated at
1,000°F. This heat treatment increases the surface hardness as shown
in Figure 8.8.5.
Fig. 8.8.5—Effect of heat treatment exposure time on erosion resistance.
270
Refurbishment Techniques for Rotating Blades
SOLID-PARTICLE PEENING
Peening, or small particle impact damage, occurs on the rotating
blades as a consequence of small hard particles, or debris that are
free to rebound within the blade path system (see chapter 4). This
type of damage causes a deformation of the surface layer of the
vanes, producing small craters, or mechanical deformation. These
deformations disrupt steady flow, and cause losses due both to the
separation of the boundary layer, and the turbulence this creates.
There are various sources of these small particles, and these are
discussed in chapter 4. In summary the following are the most likely:
•
The result of mechanical failure of some portion of the steam
path, upstream, being “chopped” into smaller particles,
which are then free to move with the steam
•
Debris carried into the unit from the boiler. This is normally
weld bead, of a size that can pass through the steam strainer
•
Debris entering the unit from some external source
•
Small parts left in the unit during an outage
The normal appearance of this damage is a small indentation,
possibly with a small rim or lip at some portion of the crater edge, as
shown in Figure 8.9.1. This type of damage causes a loss in stage
efficiency and therefore output. The most appropriate corrective
action is to dress the crater rim, removing the deformed material that
is above the edge of the crater. The dressing effect is shown in Figure 8.9.2. No effort should be made to remove any original surface
material, and as much of the original surface contour as possible
should be preserved. The lips of the craters are best removed using
fine emery cloth wrapped around a file end.
271
Turbine Steam Path Maintenance and Repair—Volume Two
Direction of Particle
Impact
Dress lip to remove
high spots
Original surface
Figure
8.9.1impact crater.
Fig. 8.9.1—Typical section through
a small
Typical section through a small impact crater.
Fig. 8.9.2—A blade vane after dressing the high
spots to remove proud lips.
272
Refurbishment Techniques for Rotating Blades
After dressing, the blade vane will not have been completely
restored, and there will still be energy losses associated with the
remaining craters presence. However, dressing of the crater lip will
have helped to reduce the losses.
In some instances these craters have been “filled-in” by the
deposit of an Inconel filler material. This will however, require a
stress relief of the blade material, and the maintenance engineer
should evaluate if the potential gain is justified.
MASSIVE PARTICLE DAMAGE
Massive particle damage is normally the consequence of
mechanical rupture of steam path components upstream of the rotating blade row, failure within the same row and the consequential
damage caused by large particles trapped within the stage, or failure
of components within the valve system, downstream of the steam
strainer. If inspection indicates massive damage has occurred to the
blade vane due to the passage through, partial passage through, or
lodgment within the steam path of a large particle, immediate corrective measures are normally necessary.
Failures producing large particles that can pass into the rotating
blades have a high probability of causing severe damage to the rotating blade row; this damage is discussed in chapter 4. Figure 4.3.6
shows a row of control stage rotating blade, which has suffered
severe damage to the extent the blades must be replaced.
Often this type of damage is aggravated because the particles are
too large to pass through the blade path, and have not been (or are
too hard to be) broken into pieces small enough to migrate down the
steam path. The fact that such articles cannot move downstream
may in fact prevent further damage from occurring to other rows.
273
Turbine Steam Path Maintenance and Repair—Volume Two
Unfortunately, often a sufficient number of pieces of debris are small
enough to migrate, while the larger particles are contained. Therefore the steam path will suffer damage on several rows.
The massive particles that are contained between stages rebound
in the axial gap between the stationary blade outlet and the rotating
blade inlet, and the level of damage shown in Figure 8.10.1 occurs,
reducing in severity further down stream.
When massive particle damage has occurred, it is necessary to
review the blade condition before any decision is made concerning
the remedial action that should be taken. Alternatives exist. If it is not
possible to restore the condition of the existing blades to the extent
that there is confidence they can continue to operate until the next
planned outage, then it becomes necessary to replace them. It may
not be possible to weld repair, because the probability of successful
welding is remote, and at this time the capability to rebuild does not
exist. If however, there is a need to return the unit to service, certain
options are available. These include the following:
Fig. 8.10.1—Massive particle damage.
274
Refurbishment Techniques for Rotating Blades
To continue to operate the unit in its damaged condition.
Depending upon the nature of the damage and the possibility of any
material fracture at the impact points, this will be an operation with
some level of risk. Stage efficiency will certainly be reduced. It is normally prudent to specify, after an engineering evaluation, the maximum period for which the blades can continue to operate in this damaged condition, and at the same time order replacement blades so
repairs can be affected when the unit is next removed from service.
This option represents a management (rather than engineering)
decision. All an engineer can expect to contribute is an assessment
of the condition and the probability of and estimated time to failure.
Note: The setting of a maximum period of operation relies entirely on the experience of plant engineers. While the blades can be
dressed and straightened, the need to undertake NDE at completion
of adjustment exists. Unfortunately, even the act of straightening can
cause further damage, which may not be apparent from NDE results.
By bending and dressing. Depending upon the extent and location of the damage, efforts can be made to correct the situation by
bending, dressing, and NDE. However, this can be risky, and
although it can assist in returning the unit to service for a short period, it should only be done after very careful examination of the condition. In this instance, new blades should be ordered to allow permanent repairs to be undertaken as soon as possible.
With these two options an evaluation should also include a consideration of the consequential damage that would be induced if failure does occur during the proposed operating period.
Removal of the blades. This can be done, and the rotor can continue to operate for some period, normally at reduced output, and
possibly with reduced initial and reheat steam conditions. Normally replacement blades should be ordered. However, if due to predicted load factors, or the age of the unit, it is intended to continue
275
Turbine Steam Path Maintenance and Repair—Volume Two
to operate in this condition, it will be necessary to provide a pressure reducing baffle in place of the blades.
Since the pressure drop associated with high reaction stages are
relatively small, it is possible to remove a rotating blade row and
continue to operate without further modification of the steam path.
However, with the impulse type stage, if rotating blades are removed
it will normally be necessary to incorporate a pressure-reducing baffle in place of the blades to ensure the energy level of the steam
entering the following stage is not too high.
Note: The pressure-reducing baffle is used to replace the stationary blade row or diaphragm to ensure the energy is removed
from the steam. This baffle consists of a perforated plate and is
designed to make the steam flow more uniform.
This baffle allows the unit to continue to operate with the design
steam conditions and therefore minimizes the loss of output.
In the case of last stage blades, it is often possible to remove an
outer portion of the blade or blades and operate in this manner for a
period. However, replacement blades should be ordered for installation as soon as possible. If the unit is returned to service without a
row of blades, it is advisable that the root portion of the wheel be
covered by either the blade root or dummy roots to protect it against
attack from chemical corrosion, water, debris within the steam path,
or other phenomena that could degrade its condition. There is normally a considerable reduction in efficiency of this stage, which is
particularly evident at part load.
If the outer portion of only one or two blades is damaged, these
can be removed along with other blades diametrically opposite their
rotating position, in order to retain balance.
Note: If rotating blades are damaged it is normally best, particularly in high temperature stages, to machine off the vanes, leaving
the root portion present to cover the wheel portion of the attachment
276
Refurbishment Techniques for Rotating Blades
and minimize the possibility of rotor damage by oxidation or other
phenomena that might affect it.
In the event of massive particle damage, it is usually necessary
to replace the blades. Not to do so will require the unit be operated
at risk, and there will certainly be a reduction in unit efficiency while
the damaged blades are used. The possibility of failure can be
reduced by returning the unit to service and operating in a mode that
will not induce higher than necessary stresses in the row. The operating parameters that must be controlled include:
For control stages
Control stages in partial admission turbines are subject to highlevel impulse loading. This loading results from the blades passing
alternately into arcs of steam admission and dead bands. To minimize the stresses produced on these blades, the following operating
limitations are suggested:
•
Operate at full arc admission at all loads. This reduces the
impact loading on the control stage rotor blades. To do this
the unit will be modified to throttle control, and some load
limit may be applied
•
If the unit has sliding pressure control, this should be used at
all loads, but again a load limit will probable be required
For last stage blades
Last stage blades are subject to high stress levels, and an assessment of the damage level must be made to determine the best temporary corrective action possible. These blades are tuned elements,
and any corrective action must consider the effect of mechanical
dimensional changes on blade frequency, and the possible consequences this will have in returning them to service. There are certain
factors in modifying long blades that should be considered:
277
Turbine Steam Path Maintenance and Repair—Volume Two
•
The effect on coverband and tie wire batching patterns. Consider if these can be retained if some blade elements are
removed or cropped
•
The position of erosion shields relative to the damage.
Cracks can initiate in any specially prepared recess made to
accommodate the erosion shield
•
The effect of any tie wire admission holes present in the
vane. If a blade is to be cropped how will this affect the stress
in this area
Note: If a last stage blade is damaged and is removed, another
blade diametrically opposite must also be removed to preserve
dynamic balance. Also, the removal of these two blades introduces
two large “holes” in the row outlet. This can cause large variations
in the pressure at inlet to the row, and the blades on either side of
the “holes” will have large and turbulent steam bending forces
developed on them. In time progressive failures will occur. For this
reason, the situation should be corrected as quickly as possible.
It will also be necessary with blade removal to adjust any tie
wires and coverbands to eliminate the possibility of these elements
spanning two pitches, or alternately making batches too short.
For all stages
In addition to the special considerations for the control and last
stages blades, all blade rows with damage will operate with a
reduced load producing capability, and probably with increased
stress and stress concentration. To minimize the stresses induced in
these elements, the following actions should be considered:
•
278
Maintain unit inlet and reheat temperatures at their design or
lower delivery conditions. Do not exceed design temperature for any reason
Refurbishment Techniques for Rotating Blades
•
Maintain initial and reheat temperatures at constant values,
and prevent temperature excursions as much as possible
•
Maintain unit load at a sensibly constant value (not exceeding an agreed level, set as a consequence of the damage).
System load swings should be picked up by other units within the system
•
Minimize overspeed transients. The requirements for periodic valve testing should still be observed, but such transient
conditions should be avoided as far as possible
By exercising this level of control over the unit steam conditions
and operation, the possibility of failure will be reduced, but not eliminated. It is necessary to minimize transient operation, particularly temperature swings, which can induce high thermal stresses in the blades.
In the event of damage to the rotating blades of the types shown
in Figures 8.10.2 and Figures 4.3.6, 4.4.5, and 4.4.10 in chapter 4,
if the owner/operators makes a decision to dress these blades and
return the unit to service, there are several actions that should be
taken and considered. These actions are in addition to the ordering
of replacement blades, and include:
•
An initial nondestructive examination. This should be undertaken using dye penetrant or magnetic particle techniques.
Any cracks found should be dressed out. If a crack becomes
too deep, a detailed analysis should be made, and the risks
assessed
•
The inlet edge should be dressed removing any excess material that would block or obstruct flow of the steam into the
blade passage. A minimum of material should be removed.
However, some small amount of rounding and blending of
the inlet edges should be undertaken
279
Turbine Steam Path Maintenance and Repair—Volume Two
•
After dressing, a final dye or magnetic particle check should
be made of the remaining material, and any cracks should be
removed
•
If the coverbands have sustained damage, they should be
checked by hardness testing. It may be necessary to replace
these or at least trim them to remove hard spots. The extent
of permissible dressing can only be established in terms of its
effects on coverband stresses. No general rules can be provided for coverband trimming, and each case must be evaluated separately for the level of damage and risk involved
Note: The decision to make temporary repairs to blades having
suffered some level of damage and then return them to service is one
that is made only when considerable pressures exist to generate
power for the system. This is rarely a justified decision, because
there are considerable efficiency losses associated with this, in addition to the expense associated with re-opening the unit after a relatively short service period. In each instance the engineer and management should review carefully the “return-to-service” decision for
its overall costs, and possible consequences.
Fig. 8.10.2—Massive particle damage to a rotating blade row.
280
Refurbishment Techniques for Rotating Blades
CORROSION EFFECTS
When corrosion damage is discovered on a blade, it often occurs
in areas that are not washed or influenced to some extent by the
flowing steam. The occurrence of corrosion damage is normally a
reason for concern among operators, and the condition found to be
present must be evaluated. Such damage has the potential to force a
unit from service, and depending upon the nature of the damage can
be a cause for expensive repairs and/or corrective actions.
The regions most susceptible to damage
While the vane normally presents the most visible signs of damage and deterioration, other regions of the blade suffer, and will usually be more significant in terms of their potential to force the unit
from service. These include the following:
The root fastening area. The root fastening is an area where corrosive ions can wash and accumulate in sufficient strength that as conditions change, the ions become active and initiate corrosion damage.
A major concern with this type of accumulation is that the presence of corrosion is difficult to detect after the unit is removed from
service. If it is suspected that corrosive action is occurring, there is
no easy selection of the correct action. To remove the blades will
require an expensive rebuild of the tenons, and not to remove them
could place the unit in a condition where it is operating at risk. This
becomes a decision and judgment call on the part of plant staff. Figure 8.11.1 shows the condition of a rotating blade row, as removed
from service, showing the corrosive attack that is present on the
lower portion of the rotating blades, and will almost certainly have
penetrated into the root portion of the root attachment.
In chapter 6 Figure 6.2.1 shows a rotating blade row with heavy
deposits on the blade root area. This deposition occurred on a unit
281
Turbine Steam Path Maintenance and Repair—Volume Two
with seawater cooling, and had suffered a condenser tube leak.
When such a condition is found, determine the constituents of the
deposits, and assess if they are of a corrosive nature or could form
corrosive compounds. The composition will assist in the decision.
Fig. 8.11.1—Corrosive pits on the lower portion of a rotating blade row,
where corrosion may also exist in the load transfer portions of the root
attachment.
The tie wire hole in the vane. The tie wire hole represents a
region where corrosive ions can accumulate. At these holes the wire
can be loose [see Fig. 8.11.2(a)], or have a braze connection
between the wire and vane as shown in Figure 8.11.2(b). Both
arrangements are subject to corrosive attack.
The braze connection can be weakened by cracking at the interface between the wire and vane, often caused by poor wire/hole
alignment, and the introduction of high bending stresses, which
282
Refurbishment Techniques for Rotating Blades
cause separation between the wire and vane as the unit goes into
operation. These small separations can be sufficient to allow the collection of corrosive products and then the initiation of cracks at the
point where residual stress exists.
Fig. 8.11.2(a) and (b)—Tie wire attachment to the blade vane, provides regions where
corrosive ions can collect. In (a) is shown the loose connection with gaps between the
wire and hole providing easy access for corrodents. In (b) is shown the braze connection, with excess braze; this is an area which will also collect corrodents.
The coverband at the tenons. With tenon-attached coverbands,
there always exist gaps between the tenon, at its fillet radius, the
blade tip platform, and the underside of the coverband. The voids
that are produced in this region provide a convenient location for the
collection of corrodents, and the possibility of corrosive action in
this region is high. Unfortunately, the early stages of any cracking are
difficult to detect, as the coverband will not have begun to lift, visual inspection is almost impossible, and nondestructive examination
from the outer surface of the tenon is unlikely to indicate damage.
283
Turbine Steam Path Maintenance and Repair—Volume Two
Figure 8.11.3 shows a section through a vane and cover, indicating the voids at the underside where the corrosive ions will collect.
They can become chemically aggressive when the environment exists
to support corrosive attack on the coverband and blade material.
Fig. 8.11.3—The voids formed between a tenon head and cover band. These voids can
act as collection points for corrosive ions.
In the event any of these forms of damage exists, cannot be seen,
but are suspected from the observed accumulation of corrosion compounds on the visible portions of the blade, as shown in Figure 6.2.1
in chapter 6, the blade should be checked by visual and nondestructive means as far as possible to determine if cracks are present.
In the event examination indicated cracks are present, it is often
necessary to provide replacement blades. However, tie wire hole
cracks can be repaired by various methods. Coverbands can be
removed, the tenons rebuilt, and the coverband (new or the original)
reattached. Unfortunately, root cracks cannot be easily corrected on
either the root or wheel portion of the fastening, and more radical
correction is required.
284
Refurbishment Techniques for Rotating Blades
Should corrosion pits, such as those shown in Figure 8.11.4, be
found on the blade vane, efforts should be made to identify the
source of the corrosive ingress into the unit. No efforts should be
made to dress these pits. A typical section through such a pit is
shown in Figure 8.11.5. It can be seen there are no “lips” of
deformed material that could be dressed to restore the profile. These
pits can represent sources of stress concentration, which could lead
to cracking and ultimately blade failure. If after examination no
cracks are found in the region of the pits, the unit can normally be
returned to service. However, for rotating blades, it must be recognized the possibility of failure exists, and consideration should be
given to the provision of replacement blades that can be installed at
some convenient outage. The need for replacement is dependent
upon the local steam conditions and the extent of damage. Corrosion pits will also introduce turbulence at the surface and therefore
induce energy losses.
Fig. 8.11.4—A blade row having suffered extensive corrosion pitting.
285
Turbine Steam Path Maintenance and Repair—Volume Two
Fig. 8.11.5—A corrosion pit in the root of a 12% chrome blade root.
The most obvious methods of preventing, or minimizing corrosion damage, are the use of materials that are better able to resist
corrosive attack, and secondly to eliminate or minimize the ingress
into and the formation of the corrosive compounds within the steam
path. To achieve this, it is necessary for the operator to ensure that
as far as possible, corrodents are excluded from the unit. This is done
by improving steam purity. Having stated this, it is also known that
such action is practically impossible in a normal operating unit, and
the most successful method of eliminating significant damage and
forced outage, is vigilant examination when a unit is removed from
service and available for inspection.
The rotating blade path is particularly susceptible to corrosive
attack, because once corrosive ions are present, hideouts always
286
Refurbishment Techniques for Rotating Blades
exist as a consequence of general stage geometry and blade design.
Also the rotating components are almost always subject to continual tensile stress during operation and often the stage temperatures are
high. Therefore, if a corrosive compound is present, then corrosive
attack will occur. The operator is responsible for determining the
presence of conditions that will allow failure to occur, and for minimizing the possibility of their occurrence.
The forms of rotating blade corrosion
The various forms of corrosion damage are covered in chapter 6,
but the significant forms of corrosive damage affecting rotating
blades are summarized as follows:
Stress corrosion cracking. Turbine blades in the low-pressure
section operate alternately in the superheated and saturated condition as the load changes; they are the most likely to be affected by
stress corrosion cracking. This is because during the drying process
corrosive ions concentrate in the solution and become chemically
aggressive. Materials such as AISI 403 and 410 are susceptible when
the material has been hardened above its normal value (BHN 260245), and AISI 422 is susceptible due to its higher hardness.
In the event cracks are found, and from other observations it is
anticipated these are due to stress corrosion, then it is necessary to
take corrective action. Weld repair methods allow tie wire holes to
be corrected, depending upon the extent of damage, and tenons can
be rebuilt. Cracking in the roots require a totally different approach
to correct the condition. These are described in chapter 9.
Corrosion pitting. Corrosion pitting is probably the most common form of attack found on rotating blades. Fortunately it does not
always represent a serious situation, and many rows of blades are in
operation with extensive pitting of the type shown in Figure 8.11.1.
While such pitting is not a desirable situation, this form of damage
287
Turbine Steam Path Maintenance and Repair—Volume Two
can be tolerated, unless and until there is evidence of cracks emanating from the pits, due probably to stress concentration. The greatest difficulty in deciding to return pitted blades to service is the fact
that the initiation of cracks cannot be predicted.
After determining that the condition of crack initiation and probable propagation is present on a blade, then corrective action
becomes essential.
Fig. 8.11.6—Pitting corrosion at the closing window on a rotor.
The most serious consequence of corrosion pitting occurs when
pits form in the root of the blade. At these locations stress levels tend
to be higher, and the consequences of stress concentration are more
severe. Figure 8.11.6 shows a wheel rim at the closing window
where there have been high levels of corrodent ingress and corrosive
attack. This damage was found as a consequence of removing the
blades to refurbish tenons that had been damaged by corrosion-
288
Refurbishment Techniques for Rotating Blades
induced cracking. When corrosion is found to exist, it is unfortunate
that the rotor material is normally more susceptible to damage than
the blade material (see Fig. 6.8.19 in chapter 6), and this condition
poses a far more significant challenge to the correction of the damage, if correction is necessary.
Corrosion fatigue. These are failures attributable to corrosion
fatigue occurring in the blade, and other susceptible regions that are
subject to corrosive attack, with a level of alternating stress sufficient
to initiate cracking. The regions most susceptible to cracking are
those where there is some stress concentration. If this condition
occurs, depending upon the location, it may be possible to undertake some refurbishment, however the possibility of this is remote, as
the damage is normally found after rupture has occurred.
Coating protection against corrosion
Although controlling station water chemistry, and preventing the
ingress and formation of corrosive compounds still represent the
optimum solution to preventing corrosion, such measures cannot be
guaranteed. Relatively minor leaks into the water/steam system still
represent the potential for significant damage under the influence of
mechanisms for the concentration of ions that occur within the
power plant.
The most serious forms of corrosion tend to be those that are
generally referred to as “aqueous corrosion,” and that occur most
commonly in the wet region of the expansion. However, such damage can also initiate during periods of shutdown, when steam will
condense and lie in the lower portion of a unit, or even during construction when the unit components in all portions of the steam path
can be exposed to humid atmospheric conditions.
Damage during periods of extended shutdown can be reduced,
or entirely eliminated, by using “nitrogen blanketing.” If a unit is
289
Turbine Steam Path Maintenance and Repair—Volume Two
removed from service and “layed-up” with the intention of being
returned to service at some future date, then blanketing, while incurring an additional cost, represents a justified investment. The costs
associated with blanketing can be more than justified both in terms
of reducing the amount of work necessary to return the unit to a serviceable condition, and even in the retained efficiency level from the
retention of the original surface conditions. (This assumes the surfaces were in an acceptable condition when the unit was shutdown,
or had been cleaned and, as necessary, dressed.)
During construction, components of the steam path are exposed
for considerable periods to local atmospheres. Such atmospheres
can be both humid and also contain airborne corrodents. This is particularly the case with construction at coastal locations. While turbine manufacturers protect their components with suitable coating
materials prior to shipment, there is little to prevent either the accidental removal of these coatings during the erection phase, or the
conscious removal to facilitate assembly and alignment. With the
protective coating removed, the components can be subject to various forms of corrosive attack. Many components produced from susceptible materials can be damaged in the high-temperature, highpressure region of the steam path.
Because of the potentially serious consequences of corrosive
action within the steam path, there has been continuous research
into coatings that can be applied to the blades to protect them
against such damage. Various materials have produced encouraging
results, and concerns for this form of damage can be reduced with
their application. Research continues and undoubtedly more (and
possibly improved) coatings will be developed. At this time encouraging results have been obtained using:
290
•
cadmium electroplate per AMS 2416 (there may be limitation to
the use of this compound and process in some jurisdictions)
•
sulfamate electroplate per AMS 2424
Refurbishment Techniques for Rotating Blades
•
aluminide diffusion alloy per TMT 2813L
•
ion vapor deposited aluminum
These coatings have been developed, and have proven successful for use on the blades. What is not certain is their adequacy on the
wheel portion of the fastening, and the effect of mechanical deformation of tenons.
ROTATING BLADE
REFURBISHMENT
The rotating blade is subject to various operating phenomena,
and damaging mechanisms that will degrade its condition. Some of
this damage, such as the deposition of compounds from the steam,
will not be significant unless the compounds are corrosive. The
blades can be cleaned by blasting, and the surface restored to a satisfactory condition. Other situations require corrective action to prevent a deteriorating condition from worsening, or to return the blade
to an efficient condition. Some of the more common and proven
techniques applied to the blade will be reviewed.
Erosion damage, repairs, and control
When examining erosion damage on a blade, it is necessary to
consider its magnitude in relation to the number of hours the blade
has been in service. For example, if damage occurs that is beyond
level (3) (as defined in Table 3.8.1 in chapter 3) during the first year
(or 7,000 hours) of operation, then it is possible this blade will be
subject to excessive damage unless remedial action is taken. In the
event of suspected or obvious secondary erosion, it is advisable to
look for the cause of the damage. The damage could be caused by
291
Turbine Steam Path Maintenance and Repair—Volume Two
a blocked water collection belt and/or drainage system, or could be
a collection point existing within the steam path. In the case of
excessive secondary erosion, it will be necessary to evaluate means
of correcting this situation.
For new units. It is necessary to ensure the turbine drainage system is connected and operational. Failure to connect the drain system at assembly could cause water to collect at some point where it
is drawn over into the steam path. Damage caused by a failure to
drain normally occurs in the outer regions of the blade.
Note: In the case of a blocked drain, where water has been collecting, any sudden drop in unit load will cause water (in large drops or
slugs) to enter the steam path, probably causing extensive blade damage.
If primary erosion is severe, examine the adequacy of the drain systems one or two rows upstream of the affected rotating blade row. If collection drains just ahead of a rotating blade row are not working as
intended, they will cause localized or secondary erosion. Upstream
drains have the ability to disperse the moisture in a greater radial direction, and therefore will tend to affect a greater radial distance of the blade
inlet edge. This in effect, appears to increase the primary erosion.
For older units. When older units are examined, and before any
decisions on repairs are taken, it is necessary to examine the extent
of erosion damage the turbine has sustained in terms of its operating
hours. If it is determined repairs are necessary and justified, several
options are available. Field repairs can be undertaken for both
brazed and welded erosion shields. Brazed shields can be removed
and replaced. However, these repairs must be performed under controlled conditions; but they can be completed without the need to
remove the blades from the rotor.
The methods of weld attaching an erosion shield were described previously in this chapter. The method of making a temporary weld repair by
the deposition of raw Stellite 6B® was described earlier in this chapter.
292
Refurbishment Techniques for Rotating Blades
Vane weld repairs at the tie wires holes
Cracking that originates at tie wire holes can affect the integrity
of the blade vane. These cracks are induced by the high combined
stresses often produced as a consequence of the stress concentration
associated with dynamic loading within the region of the hole itself.
Corrodents, which migrate into the gap between the tie wire and the
hole, also affect the cracks and cause corrosion fatigue cracks to initiate. It is possible to introduce stress corrosion cracking where
brazed tie wires have been used, if the brazing process has overheated either the vanes or wires. These effects, together with the
residual stresses caused by poor wire hole alignment (which results
from deflecting the wire and vane during assembly) often initiate
cracks, which will then propagate by high cycle fatigue (HCF).
Fig. 8.12.1—A crack initiating at a tie wire hole and
progressing across the blade vane.
293
Turbine Steam Path Maintenance and Repair—Volume Two
The combined effect of the normal operating stresses due to centrifugal loading and dynamic effects, together with the residual
stresses, act to produce cracks that initiate at the wire hole corner
and run across the chord width of the blade vane. Such a crack is
shown in Figure 8.12.1. Here the crack has initiated under the braze
material and is extending along the chord, and across the blade. This
type of damage is, depending upon the length and depth of the
crack, capable of being weld repaired. It is necessary first to excavate the crack to establish its extent. Such excavation is shown in
Figure 8.12.2. The following are three distinct methods of repairing
these cracks:
Fig. 8.12.2—Grinding away a crack in a blade vane prior to weld
repair.
294
Refurbishment Techniques for Rotating Blades
By weld repairing in-situ. This process is acceptable when the
depth of crack does not exceed more than 30% of the vane thickness
at the location of the crack. (These cracks are normally present in the
low-pressure end of turbines, on those blade rows that employ tie
wires.) Such a repair is undertaken using an Inconel 82 rod.
The repair is affected after complete excavation of the crack,
which must be undertaken by successive grindings and nondestructive examination to be sure the crack is completely removed. After
cleaning the crack, remove all traces of the braze material if the wire
is braze connected to the vane. Not to do so would contaminate the
weld. A suitable means of removing the braze material is a “soak and
wash” process using a 50% hydrochloric acid solution. After removing the braze material, the surfaces must be thoroughly cleaned
before the weld process is begun.
When the surface is clean the weld process builds the blade
material back to its original form. To undertake this repair it is necessary to preheat the blade material, and then apply “post weld” heat
treatment at a temperature that is about 50°F below the austenetizing temperature. The period should be about 30 minutes.
It is also common when undertaking this repair to weld connect
the wire to the vane.
One advantage of this weld repair technique for last stage blades is
that it can be undertaken inside the exhaust hood by gaining access
through the low pressure (LP) manways, assuming the crack is visible and
accessible. Therefore, it can be undertaken during a weekend outage.
Removal of the blade from the unit, or to have unlimited access
to it. If the blades are removed from the rotor for other purposes, and
cracks are known to exist or are found to be present, the crack can
be removed as described earlier, and the vane repaired. The requirements of crack excavation, pre- and post-weld heat treatment are the
same as those described above. The steps in the process require the
295
Turbine Steam Path Maintenance and Repair—Volume Two
crack be excavated after careful removal of any original braze material that might be present. The vane is then pre-heated, the weld
deposited, and the normal stress relief undertaken. The hole is then
reamed after any drilling, which might be required. The hole placement requirements are defined in Figure 8.12.3.
G
G
+
+
A
-
h
h
n
n
-
R
G
G
Dr
Fig. 8.12.3—The tolerances defining the
position of a tie wire hole in the vane in
both the chordal and radial directions.
Hole plugging and re-drilling. Another repair method requires
the hole in the vane be filled with weld deposit, the vane is then
stress relieved, and the hole redrilled. This method will normally be
completed with the blades removed from the rotor or wheel, and the
hole is filled with a weld metal compatible with the blade material.
The requirements of stress relief after welding the hole are the
same as described for the first option.
296
Refurbishment Techniques for Rotating Blades
The first method of repair is considered temporary, suitable for a
short period of operation, and the blade should ultimately be
replaced. The second and third methods are more complex in terms
of pre-repair preparation, but the quality of the repair is considerably
higher and it is considered permanent.
At completion of any of these three methods of repair, the hole
position requirements are the same as those required for a new
blade. These are shown in Figure 8.12.3.
To pass through the blade vane, an access hole must be provided. Producing such a hole introduces a high stress concentrating
effect within the vane, and will increase local stress levels. There are
therefore special requirements associated with hole production that
must be observed. The chordal and radial location and internal hole
finish are selected and produced with careful attention being paid to
the main dimensions for its location. Suggested hole location tolerances are shown in Figure 8.12.3.
s
k
k
t
Fig. 8.12.4—A blade vane showing the tie
wire hole, with reinforcement to replace
the material removed to permit the wire
assembly.
297
Turbine Steam Path Maintenance and Repair—Volume Two
Here the diametral location is “2R,” with a tolerance of
+/-0.010". In the axial direction the hole should be located at some
distance “A” from a radial line such as “G-G” that passes through
the vane center of gravity. This location should be within +/-0.015".
Small deviations outside these axial locations can be accepted, but
only if the entire blade row has the same level of error. It should be
noted that the blade vane will have a bending stress induced in it
equivalent to the product of the centrifugal mass of the wire, and the
hole displacement. The theoretical alignment of the wire to the hole
is shown in Figure 8.12.4
If the production of the access hole induces unacceptably high
stresses in the vane, local reinforcement can be used. Such reinforcement is shown in Figure 8.12.5, where a vane having a thickness “s” at the hole center, is reinforced locally by the addition of
areas “E1” and “E2,” shown in Figure 8.12.5, which approximate the
area removed for the wire. This reduced area is equal to “S.(D+2k),”
where “D” and “k” are the wire diameter and radial clearance.
E1
S
k
D
k
E2
Fig. 8.12.5—A tie wire hole
showing the dimensional and
clearance requirements for a
satisfactory admission hole.
Engineering dimensional and surface finish requirements of the
hole must be observed, to prevent local stresses reaching unaccept-
298
Refurbishment Techniques for Rotating Blades
ably high levels. In producing the hole, as shown in Figure 8.12.4,
the wire has a diameter “D,” and the vane hole a diameter “D+2k,”
where “k” is the mean radial clearance from the wire to the hole surface. After assembly there should be no vane deflection on the larger radial height elements, and no wire distortion on any shorter
vanes. The hole must also have either a smooth radius at the entry
and exit point from the vane (detail “s”), or a smooth chamfer (detail
“t”). The surface finish of the hole should be 32-64 √µ-inches, preferably produced by reaming.
Weld repair of integral snubbers. Under high frequency loading,
and possibly aggravated by high degrees of blade untwist, it is possible for integral snubbers to fail. This is not a common occurrence,
but the blade cannot be reused, should it occur. To replace a blade
is expensive. However, techniques exist to permit the weld rebuild
of the snubber. After this weld rebuild, the vane must be stressrelieved (temperatures depending upon the base blade material).
Vane tip cracks
If the vane tip has integral tenons, situations can arise in which
high amplitude vibratory loads on the tenon can induce failure. In
certain circumstances, the resulting crack can extend into the vane
tip. Most often the crack initiates in the fillet radius at the base of the
tenon, and is caused by either:
•
A poor finish of the fillet radius, the radius not blending correctly with the tenon vertical face and the blade tip platform
•
Interference between the underside of the coverband chamfer and the fillet radius
Figure 8.12.6 shows such a failure, with the tenon and a small
portion of the blade tip missing. Here the crack at the tip has progressed into the vane to the extent the tenon and a small portion of
the blade vane have detached.
299
Turbine Steam Path Maintenance and Repair—Volume Two
Fig. 8.12.6—A tip crack showing the tenon having failed as a consequence
of high cycle fatigue.
With failures of this type, it is possible to rebuild the tenon and any
vane material that has been removed. Figures 8.12.7(a) through (d) show
the steps in completing such a repair. In (a) is shown the blade as removed
from the turbine, with a crack extending under the entire chord of the
tenon. It will also be seen that the tenon and coverband are missing. The
only reason there had not been complete separation from the vane was
that adjacent blades had not allowed further outward movement of the
coverband and tenon. In fact, the tenon had failed at the fillet radius starting on the inlet edge.
In (b) the tenon has been removed, exposing the failure surface.
At (c) the tenon has been rebuilt and stress relieved. Before rebuilding, grind away the failure surface to provide a clean area that has
no oxides or other contaminants on the surface, which would pro-
300
Chromalloy HIT Division
Refurbishment Techniques for Rotating Blades
Chromalloy HIT Division
Fig. 8.12.7(a)—The cracked blades as
removed from the rotor.
Fig. 8.12.7(b)—The blade after removal of the
tenon region.
301
Chromalloy HIT Division
Turbine Steam Path Maintenance and Repair—Volume Two
Chromalloy HIT Division
Fig. 8.12.7(c)—The weld rebuilt tenon, and
vane tip material.
Fig. 8.12.7(d)—The weld rebuilt tenon
after finish machining.
302
Refurbishment Techniques for Rotating Blades
duce a less than optimum weld attachment. In (d) the tenon has been
reformed to the original dimensions.
The rebuilding of tenons is discussed in greater detail in chapter 9.
When undertaking this type of repair in the region of an erosion
shield, it is necessary to check the braze integrity of the shield at
completion of the weld, if it had not been removed prior to the weld
rebuild.
Forming new tenons for coverband attachment
A common region for damage to occur is in the tenons, which
attach a coverband to the tip section of the blade vane. This coverband is required principally to form one surface of the expansion
passage. The forms of damage suffered by tenons are discussed in
detail in chapter 9.
WATER INDUCTION
In terms of the damage caused to rotating blades by water
ingress, it is likely that a “slug” of water being ingested into the steam
path will cause unrepairable blade damage. The blades will most
likely be bent, and therefore destroyed. However, if the water reenters as a steady flow in small quantities, as occurs with a blocked
drain, it will cause a concentrated level of erosion, which while
removing material from the inlet edge, will not cause mechanical
deformation of the vane.
Under these circumstances of localized damage, the condition
may be acceptable, and will not require refurbishment to allow it to
be returned to service and operate in a satisfactory manner. Howev-
303
Turbine Steam Path Maintenance and Repair—Volume Two
er, if this type of damage is found to be the reason for the water concentration, it should be investigated and corrected if possible. Such
damage is often the consequence of a blocked drain or a closed
extraction valve. It is possible this situation could deteriorate and a
“slug” be generated, which would destroy the blades.
FRETTING CORROSION
Fretting corrosion occurs in the blade rows, most often at the
interface between the roots and between integral coverbands. It has
also been observed at tie wire holes where there is a tight fit, and by
contact between a deformed wire and vane. The possibility of refurbishment or correction will depend upon the form and extent of corrosion. The conditions must be evaluated independently.
304
Refurbishment Techniques for Rotating Blades
REFERENCES
1. Cotton, K.C., and J. Angelo. Observed Effects of Deposits on
Steam Turbine Efficiency, ASME Paper 57-A-116
2. Fraser, M.J. Weld Repair Procedure for Refurbishment of
Steam Turbine Blade Elements, Turbine Blading (USA), Inc.
3. Beaudry, R.J., and K.S. McLeod. The Development and
Application of Welded Cobalt-Free Erosion Shields for LowPressure Steam Turbine Blades, IJPGC, Atlanta, Georgia,
1992
4. ASME Radiographic Inspection Standard Section VIII, Para.
UW-51
5. Swetz, S.D., M.J. Fraser, and G.J. Russell. Major Weld Repair
to Tuned L-0 Turbine Blades, EPRI Steam Turbine Blade Reliability Workshop, Los Angeles California, March, 1986
6. EPRI C5085 Project 1408.2 Proceedings: EPRI Steam Turbine
Blade Reliability Workshop, Los Angeles, California, March,
1986
7. Bollman, P., R. Tewes, and H. Viertmann. On-Site Weld
Repair, without Disassembly to Low Pressure, Last Stage
Blading of a 300MW Condensing Unit, presented at VGB
Conference, Maintenance Within Power Plants, February,
1992
8. Jordan, S., and M.J. Fraser. Design Modifications and Repairs
to Existing Steam Path Components to Improve their Existing
In-service Performance, Turbomachinery Congress, Berlin,
October, 1991
9. General Electric Company Publication GEA-11850: There
are Solutions to Solid Particle Erosion Damage
305
Turbine Steam Path Maintenance and Repair—Volume Two
10. Tran, M.H., and J.R. Kadambi. Characteristics of New Control Stage Section with Contoured Endwall, AMSE Publication PWR-Vol. 7, Latest Advances in Steam Turbine Design,
Blading Repairs, Condition Assessment, and Condenser
Interaction, papers presented at the JPGC, Dallas, Texas,
October, 1989
11. EPRI Research Project Report TR-107021: State-of-the-Art
Weld Repair Technology for Rotating Components, Volume
2, December, 1997
12. Protective Coatings for Steam Turbine Components,
Sermatch International, Inc., Publication
13. Kramer, L.D., J.I. Qureshi, R.A. Rousseau, and R.J. Ortolano.
Improvement of Steam Turbine Hard Particle Eroded Nozzles
using Metallurgical Coatings, ASME Publication, PWR-29
14. Power: Coating Technology may help Fight Steam Turbine
Corrosion, February, 1982
15. EPRI Project Report CS-5415: Erosion Resistant Coatings for
Steam Turbines, September 1987
16. EPRI Report CS-5415: Protective Coatings, prepared by the
General Electric Company, Schenectady, New York, September, 1987
17. EPRI Proceedings Project CS-4683: Solid Particle Erosion of
Utility Steam Turbines, Chattanooga, Tennessee, 1985 Workshop, August, 1986
18. EPRI Proceedings Project GS-6535: Solid Particle Erosion of
Steam Turbine Components, New Orleans, Louisiana, 1989
Workshop, September, 1989
19. Solution to Solid-Particle Erosion, EPRI Journal October/
November, 1990
306
Refurbishment Techniques for Rotating Blades
20. Specialized Coatings, Chromalloy Research and Technology
Publication CRT 06913500
21. EPRI Project Report CS-2932: Corrosion Fatigue of Steam
Turbine Blading Alloys in Operational Environments, September, 1984
22. Ortolano, R.J. Users Guide for the Use of Corrosion Resistant
Coatings on Steam Turbine Blades, EPRI Report, December,
1986
23. Sanders, W.P. Moisture Damage in the Turbine Steam Path
and its Impact on Life Extension, Turbomachinery International, Vol. 33, No. 1, January/February, 1992
307
Chapter
9
Damage Mechanisms
Arising from
Operation and
Refurbishment
Techniques for
Rotating Components
INTRODUCTION
When a rotating component has been damaged, the maintenance
engineer will review the situation and reach a decision concerning
the most appropriate action to be taken to allow the condition to be
corrected, so that the unit can be returned to a structurally safe and
serviceable condition, in the shortest time possible. These corrective
actions must be consistent with retaining component and stage reliability. There are instances where utilities and other owners have spare
309
Turbine Steam Path Troubleshooting and Repair—Volume Two
rotors or replacement components, which allow necessary replacements to be made. However, this is not a common practice; it is normally limited to installations where duplicate units exist, or in some
instances where the financial consequences of a forced or extended
outage are considerably greater than the cost of the rotor.
The concern with rotating components is that they are normally
subject to high levels of centrifugal stress due to their own weight.
The rotating components may also be subjected to other forms of
stress, which in a damaged condition may be aggravated, thereby
increasing total stress levels and reducing the “factor of safety” (FofS)
below an acceptable level.
When a damage condition has been determined to exist, two
actions are required of the engineering group responsible for correcting the situation:
•
First, a review should be made of the condition to determine
the cause of the damage. This review should include, as far
as possible, identifying the initiating and driving mechanisms
involved (see chapter 1). In many situations this is a relatively simple process, and can be resolved by visual and nondestructive examination. However, there are other situations
that can only be established as the result of careful evaluation of both the component, and the manner in which it has
been operated
Often it is not sufficient to identify the mechanisms of failure or
damage. This is because these mechanisms, particularly those initiating the condition, are the consequence of some abnormality in the
unit components or mode of operation. The maintenance engineer
must identify the cause rather than the effects. This evaluation is an
essential step, as it may be possible to take corrective action for future
operation, which could prevent a reoccurrence of the damage.
310
The Repair of Rotating Components
•
Secondly, the most appropriate corrective action must be
determined, from an analysis and comparison of the available alternatives, their cost, and time to completion
The damaging mechanisms leading to failure in a rotating component are essentially those involved in damaging the stationary
components. There are also the effects of complex stresses introduced as a consequence of rotation. It is necessary to consider these
effects on the stress levels of the components, and the possible consequence if corrective or remedial action is not taken to return them
to an acceptable condition, and then continue to operate.
This chapter considers the damaging mechanisms that can affect
the rotating components (other than the rotating blades, which are
discussed in the chapter 7). It also discusses some of the refurbishment techniques that are currently in use.
THE ROTATING COMPONENTS
In considering the repair/refurbishment of the rotating components, in addition to the blades, it is necessary to consider their possible forms. The most commonly damaged components, and those
most often in need of repair are the rotor, coverbands, and tie wires.
Many stages of the steam turbine utilize either coverbands or tie
wires. These two groups of components have similar, but functionally different purposes within the stage. The tie wires transmit vibratory loads developed in one element to adjacent members. By
accomplishing this they dampen the magnitude of vibration of the
blade, and therefore lower the stress levels to which the blades are
subjected. Coverbands have a primary function of limiting the steam
flow around and through the tip section of the blade. But they are
311
Turbine Steam Path Troubleshooting and Repair—Volume Two
also used to fulfill a secondary function, as the tie wire, of helping
dampen the amplitude of vibration experienced by the blade vane.
These two groups of components, while they perform necessary
functions within the stage, add some measure of complexity both to
its design and means of construction. They do this by changing the
natural frequencies, and by adding additional centrifugal load,
which must be supported by the vane and root. When these two
components are used, they require a defined method of attachment
to the vanes, again possibly adding some level of design and assembly complication complexity.
These components are normally batched into discrete lengths
within the row, with a specific number of blades in each. In recent
designs there have been attempts to increase the length of these
batches to the extent of making them span the complete 360 degrees
of the blade arc, i.e., the ties through the blades are continuous.
There are distinct advantages to this, but design complexity is often
added, because it is still necessary for the wires and coverband to
grow in the tangential direction during operation as the result of both
radial extension of the blades, as a consequence of stress, and as they
expand with temperature as the steam is admitted. Because the tie
wires and coverbands will heat and cool at different rates from the
blades and rotors, they will change dimensionally at different rates.
These effects must be considered and accounted for in the design.
The turbine rotors are the major rotating component of the turbine, to which are attached the rotating blades, designed to extract
energy from the steam as it expands through the steam path. This
energy as a force is then transmitted to the turbine rotor, which
drives the generator, causing it to rotate against a magnetic field and
produce electrical power at the generator terminals. Alternatively,
this rotor can be attached to, and drive some mechanical device.
Rotors may be manufactured to one of several forms, these
forms depending upon the manufacturer and the expertise that has
312
The Repair of Rotating Components
been developed and utilized in their particular manufacturing
processes. The details of the rotor construction are also influenced
by the proportions and dimensional requirements of size, and the
number of blades it must carry. The form of the rotor is also influenced by the environmental condition within which the blades and
rotor must operate.
In sizing a rotor, the torque it is expected to transmit during full
load conditions, and the centrifugal stresses induced in it at both
normal and emergency overspeeds are perhaps the most critical
considerations. In addition to these principal considerations, are the
influences of stresses due to blade loading, thermal transients, and
the effects of shrink-fit stresses in regions where such methods of
construction are used. Also, there are bending stresses induced in
the rotor due to its own weight supported between the bearings.
The rotor, like all high speed rotating components, requires very
careful dynamic balancing. The rotor must be carefully analyzed for
critical speeds to ensure these do not occur at values close to, or at
multiples of the operating speed.
Functions of the coverband
The coverband can be used to provide more than one functional
advantage to the stage. These can be summarized as the following:
Principal function. To provide the fourth side of an expansion
passage, and in so doing prevent the steam from expanding through,
or being centrifuged through the tip section of the vanes. For rotating
blades, the coverband ensures the flow is through the blade expansion passage, and there is no excess leakage due to centrifugal action.
Secondary functions. The coverband, however, provides other
secondary functions:
313
Turbine Steam Path Troubleshooting and Repair—Volume Two
•
The coverbands on large radial height blades, which produce a large discharge area, provide a means of controlling
that area at the design value, by restraining the vanes from
tangential and axial distortion
•
The coverband can be designed to incorporate part of a
steam leakage control system. The coverband can provide
either a seal strip, or a platform against which a seal strip,
located within the stationary portion of the stage, will seal
•
The coverband is normally designed to attach to the blade
inner (stationary) or outer (rotating) extremities, and it helps
to guide the steam through the vanes with a minimum of
aerodynamic loss
•
By being firmly attached to the blade ends, the coverband
acts to tie the blades together and dampen the magnitude of
vibrations in the same manner as the tie wires
Forms of the coverband
There are three forms or combinations of coverband design that
are used in the stationary and rotating blade rows of the steam turbine, which should be considered:
Attached coverbands. These coverbands are produced separately from the blade in multi-pitch lengths and attached to the blades
through tenons. The accepted practice is to produce the tenon integral with the vane and to pass this tenon through specially prepared
holes in the coverband. The tenon head is then formed.
Integral coverbands. The integral type coverband is generally
more robust than the attached coverband, and is produced as an
integral part of the manufactured blade vane.
314
The Repair of Rotating Components
Multi-layer coverbands. There are multi-layered coverbands, in
which more than one coverband is used. This form of coverband (in
two or three layers) is used for mechanical/structural reasons. The
inner coverband is normally produced integrally with the vane, with
outer layers being attached through tenons, the tenons being produced with the vane/inner coverband machining.
To simplify considerations of the coverbands they can be placed
into six groups, depending upon their form and secondary function.
These groups are defined, according to the manner in which they
form a seal, or provide a concentric platform against which seals can
be formed with another steam path components. These groups are:
Type
A – Coverbands that do not provide any seal or seal platform
Type
B – Coverbands that provide an axial seal
Type
C – Coverbands that provide single or multiple radial seals
Type
D – Coverbands that provide radial seal platforms only
Type
E – Coverbands that provide an axial seal and a radial seal
platform
Type
F – Special design coverbands
In considering the possible shapes of the coverbands, several
modifying or influencing factors that affect the overall performance
of the stage should be noted:
•
Ideally, the coverband will act to guide the steam from the
stationary blade row or diaphragm into the rotating blade
rows through which the steam is expanding. Figure 9.2.1
shows how the coverband helps deflect, or guide the steam
into the rotating blade row. It does this by having an overhang, as shown in Figure 9.2.1(a). In some cases a radial seal
is provided at the inlet edge, as shown in Figure 9.2.1(b), but
315
Turbine Steam Path Troubleshooting and Repair—Volume Two
with adequate radius on the underside to help divert most of
the steam down and through the rotating blade row
•
On the discharge side, the inner edge should be sharp
enough to have a minimal directional influence on the steam
exhausting from the rotating blade. A comparison between a
sharp and rounded discharge edge is shown in Figure 9.2.2
•
The blade passage can have either parallel or tapered outer
walls, as shown in Figure 9.2.3(a) and Figure 9.2.3(b). The
final shape is influenced by stress consideration, and the possible need to produce an adequate seal surface
•
The coverband can have considerable axial movement during operation, due to differential expansion between the
rotating and stationary portions of the unit. This will help
determine whether it is possible to use an axial seal, or
whether radial seals would be preferable
These modifying factors should be recognized as secondary, but
valuable considerations in the coverband design.
Casing
Diaphragm
Outer
Ring
Stationary
Vane
Rotating
Blade
Vane
Casing
Diaphragm
Outer
Ring
Rotating
Blade
Vane
Stationary
Vane
(a)
(b)
Figure
9.2.1
Fig. 9.2.1—coverbands and their ability
to help
deflect the steam discharging from the
Cover
and
their row.
ability to help deflect the steam discharging from the
nozzle into
the bands
rotating
blade
nozzle into the rotating blde row.
316
The Repair of Rotating Components
Effect of rounded
corner on the
underside
Rotating Blade
Fig. 9.2.2—The effect of a rounded under
surface.
Casing
Casing
Stationary
Blade
Rotating
Blade
Rotating
Blade
Axial running
clearance
Radial
clearance
Radial clearance
above blade.
(a)
(b)
9.2.3
Fig. 9.2.3—Showing a cylindrical outerFigure
sidewall
in (a), and an outward tapered design
in (b). Showing a cylindrical outer sidewall in (a), and an outward tapered design in (b).
Functions of the tie wires
Tie wires are included in many stages where the rotating blades
have a large radial height vane. Their functions follow:
Principal function. The tie wire has a single function within the
stage, which is to mechanically link the blades. This must be done
with sufficient rigidity, because the tangential vibrations developed
in one blade are transmitted to the other connected elements.
317
Turbine Steam Path Troubleshooting and Repair—Volume Two
However, during start-up, shutdown, and transient operation,
there are periods when the tie wires will be at temperatures that are
different to other components within a row. During these times the
wires should have sufficient tangential flexibility, or be connected in
such a pattern that the temperature differentials will not induce
excessive stresses in either the blades or tie wires.
Secondary function. The tie wires have a secondary function of
helping dampen both axial and torsional modes of vibration. Their
ability to do this will be a function of their form. Those tie wires that
pass through holes produced in the vane and are not brazed to it,
will have little influence on torsional modes, but possibly some
effect on axial modes. If the tie wires are brazed to the vane, there
will be some dampening to both modes, the actual damping being
dependent upon the strength of the braze. However, a brazed joint
is not particularly strong against vibratory motion.
Therefore, tie wires and coverbands, while having certain similarities in terms of their relationship to the blade, do have different
functions within the stage.
Forms of the tie wires
Because of the energy and therefore efficiency losses associated
with the use of tie wires, the designer will avoid their inclusion as
much as possible. However, the structural value of these wires and
their ability to dampen vibrations has been well documented and the
majority of manufacturers continue to use them. These wires not
only dampen the magnitude of vibration, they can also modify the
vibratory characteristics of an entire stage.
Note: Some manufacturers continue to produce long latter stage
blades without tie wires, preferring to rely upon the accuracy of the
manufacturing process used to produce blades of known vibratory
318
The Repair of Rotating Components
characteristics, and by tuning, adjust the blades to avoid dangerous
levels and harmonics of vibratory motion.
A blade row may contain one, two, or three wires, which can be
constructed as a continuous band or arranged in a segmental pattern. The number of wires selected, their diameter, and the pattern in
which they are arranged depends upon the designer, who has available experimental data indicating which arrangement will provide
the most effective source of damping. However, the final arrangement of any stage will also be influenced by manufacturing considerations and the ability to assemble the wires in the row in the factory and field.
Note: The use of three wires has now been discontinued, but
many units are still in service with such an arrangement, and these
will require maintenance until they are decommissioned.
Irrespective of the form of the tie wires used in any row, they will
need to be mechanically attached to the blade by some suitable
means. This is necessary so they can transmit vibratory stimuli
induced in one blade to those to which they are connected. In addition to these requirements, the form of the wire is selected to help
satisfy certain other requirements of the stage:
•
That the aerodynamic losses induced by the wires are minimized, and therefore stage output is maximized
•
To help to ensure the wire can be manufactured and assembled in the row by economical means. For fieldwork this
may require considerations than those required during initial
manufacture. The ability to undertake work in the field without the need to disassemble too much of the stage, or to
make in-situ repairs is important to the operator, when trying
to minimize down time
319
Turbine Steam Path Troubleshooting and Repair—Volume Two
•
To ensure the design is sufficiently robust, the wire will be
able to operate within the stage environment without introducing operational problems
•
To ensure the stresses induced in the wire due to its own
mass, and any pieces it carries, such as ferrules or spacer
washers, will not exceed a value that will force the unit from
service
There are two basic forms of wire. First is the integral type, in
which the wire, or a stub of the wire is formed integral with the vane
of the blade, and then mechanically connected by some bridging
piece. The second type of wire is the continuous wire, which passes
through a hole produced in the vane. Both forms of wire fulfill the
same function. However, there can be differences in their method of
connection, and certain complexities are often introduced into the
stage by their form.
Wire cross sections
The wire can be produced to any cross section that adequately
ties the blades together, can be assembled to the row, and can carry
the centrifugal load induced in it by its own weight. The most common form of wire in the steam turbine is the circular cross section
type. Such a wire can be either solid or hollow. The hollow wire is
often used to reduce the bending stresses induced in it by its own
weight. This high stress can occur with course pitched blades. Under
these circumstances a hollow wire is used, which produces a high
section bending modulus, and induces a minimum centrifugal force.
If the designer determines, from considerations of required
damping, that a tie wire must be used in a blade row, the elliptical
shape is preferable to the circular form (from efficiency considerations). It has been demonstrated that generated turbulence, and
therefore losses associated with a circular wire are about three times
320
The Repair of Rotating Components
greater than those of an elliptical wire, whose minor axis is of the
same dimension as the diameter of the circular wire, and whose
major axis is four times the minor. Such wires are shown in Figure
9.2.4(a), and a hollow wire in Figure 9.2.4(b). In fact it is difficult or
impossible in a steam turbine to use an elliptical wire with a major
to minor diameter ratio of 4.0 with the continuous form of wire, as
this wire must pass through a hole produced in the blade vane.
However, this may be achieved or approximated in the integral wire.
The circular wire is also used in the hollow format, as shown in
Figures 9.2.4(c) and (d). This configuration is particularly useful
when local stresses in the vane are high, and it is necessary to reduce
wire centrifugal loading without reducing wire section modulus by
a significant amount. It is also possible to employ wires produced
from titanium, this being a metal of lower density and higher
mechanical strength. However, this does add to unit cost. When
replacing wires, a solid wire should not be substituted for a hollow
wire without a careful analysis of the implications for both the wire
and blade vane.
While the true elliptical wire may not be practical in many applications, the turbine manufacturers have approximated this form to a
degree sufficient to eliminate some of the losses that are induced by
the generation of turbulence behind the wires.
Dx is the major elliptical diameter.
Dn is the minor elliptical diameter.
Dx
Dn
(a)
(b)
(c)
Figurewire
9.2.4cross
(a), (b)
and (c)
Fig. 9.2.4 (a), (b) and (c)—Alternate
sections.
Alternate wire cross sections.
321
Turbine Steam Path Troubleshooting and Repair—Volume Two
Dx is the major elliptical diameter.
Dn is the minor elliptical diameter.
Dx
Dn
(a)
(b)
(c)
Fig. 9.2.4(d)—A hollow outer wire of the form shown in figure 9.4.2(c)
brazed into position on the blade vane.
Some common wire forms approaching the elliptical are shown
in Figure 9.2.5. In Figure 9.2.5(a) is a wire section formed by the
straight portion, length “L,” and two rounded ends of radius “R.” In
Figure 9.2.5(b) the form is similar, but the radii are not equal; they
compromise two radii “R1” and “R2” between a length “L” between
centers. These forms of wire shown in Figure 9.2.5 are suited to the
integral form of wire only.
L
R
(a)
L
R
R1
R2
(b)
Fig. 9.2.5—Various wire sections approximating the elliptical.
The steam in the latter stages of low-pressure sections can have
a large radial flow component at some radial locations, in addition
to the axial component of velocity. Such a flow pattern is shown in
Figure 9.2.6, where the streamlines at the tie wire positions are
inclined to the axial or horizontal direction by “α.” Because of this
322
The Repair of Rotating Components
radial flow effect, if an elliptical wire is used, it is advisable to adjust
its inclination so that there is no large degree of steam incidence at
the wires, and the steam will approach the elliptical form at an angle
consistent with steam direction. It is difficult to predict exactly the
angle of the streamline at the wire position, and the effects of flow
patterns at part load, when the steam is flowing at a reduced rate,
and possibly in some stages with a modified volumetric flow compound this difficulty. However, there are advantages to this elliptical
(or semi-elliptical) wire, and manufacturers use it to good advantage
in their units.
α1
α2
Fig. 9.2.6—Stream line flow pattern around an elliptical
tie wire.
Batching of coverbands and tie wires
Blades are batched (connected in groups) by both tie wires
and coverbands. This is done so that each batch forms a continu-
323
Turbine Steam Path Troubleshooting and Repair—Volume Two
ous, although complex, mechanical structure. This interconnection of the blades is capable of dampening the amplitude of vibration induced in individual blades, by vibratory forces from one
blade on all those to which it is connected. When mechanically
connected, the blades within each group act together to form a
total damping arrangement.
The batching patterns used by different manufacturers, and used
at different times in their technology development to establish the
pattern they will use to batch the blades together can be placed into
two broad categories:
•
Connecting the blades of a row into separate and discrete
batches. Such a discrete batching is shown in Figure 9.2.7(a).
Normally each group contains an equal number of blades,
which for larger blades are selected after tests have been
conducted to ensure these will operate within a safe frequency range
For smaller radial height blades, this batching may be only the
grouping of coverbands. If blades are to be replaced the same pattern must be achieved in the replacement blades, with particular
attention being paid to the position of any closing blade or root
block. While the number of blades in each batch should be about
the same, the number of batches and any differences in blade count
from one batch to the next must be observed.
In this batching pattern all ties (wires and covers) connect the
blades together. These batches are free to vibrate as separate assemblies without any influence being transferred from adjacent batches
within the row.
•
324
Arranging the blades into a staggered (or random) pattern, as
shown in Figure 9.2.7(b). In this case an attempt is made to
tie the blades in such a manner as to transmit the stimuli
from any blade throughout the entire row. This is done in
The Repair of Rotating Components
such a way that it can maintain sufficient tangential flexibility. Blades can accept temperature changes without excessive distortion.
(a) Groups of six blades in discrete groups
(b) Random batching
Fig. 9.2.7—Alternate batching of tied blades. In (a) is
shown discrete batching with elements in groups of six
elements, tied by a coverband and two tie wires. In (b)
is shown a random batching with a coverband and two
tie wires.
For multiple ties (2, 3, or 4 connections), the pattern should be
selected so that at no point are there less than two ties transmitting
the motion between any two batches. It is also necessary for the
manufacturer to specify the minimum number of blades that must be
present between adjacent batch ends.
The staggered pattern is usually specified by the designer, and
must be adhered to in the manufacturing phase and assembly. If such
325
Turbine Steam Path Troubleshooting and Repair—Volume Two
blades are disassembled for maintenance, details of the batching
pattern should be recorded and repeated on reassembly.
Single tie connections
Manufacturers still disagree as to the most suitable pattern to batch
the blades together when a single tie connection is used (a coverband
or single tie wire). However, there are certain basic requirements that
should be considered and evaluated for the assembly:
•
An odd, preferably prime, number of batches should be used
•
The manufacturer can elect to use a varying or constant
number of blades per batch. The actual pattern must also be
influenced by the number of blades in the row, and the stage
temperature (because of equalizing expansion and end thrust
with temperature changes)
•
Batch length for the blades should exceed three stationary
blade pitch lengths for the row from which they are receiving steam. This is considered necessary to prevent a swinging mode of vibration developing in a tied batch. Such a
swing would be due to the impulse received by the blade
vane as the batch moves past the nozzle element, which
does not have an even pressure distribution across its discharge pitch
Some blades have their mechanical design improved by the use
of long arc coverbands (shrouding), which ties the blade with sufficient constraint, considerably reducing the amplitude of vibration
and therefore vibratory stress.
326
The Repair of Rotating Components
Functions of the rotor
As the primary rotating component of the steam turbine, the
rotor has a number of functions that it must fulfill in order to allow
the turbine to operate successfully. The main functions are:
•
The rotor must be capable of withstanding the centrifugal
loading developed on it due to its own weight. In the case of
monobloc and welded designs, the rotor must also be able to
withstand the centrifugal loading from components such as
blades, tie wires, and coverbands that it carries
•
In operation, torque is developed on the rotor due to expansion of steam and the work done by the blades during this
expansion. The rotor must be capable of transmitting this
torque to the generator. It must also be capable of transmitting torque developed on other rotors further from the generator, or driven equipment
With multiple flows, low-pressure rotors, and when more than
one similar unit is used in any station or system, it is often advantageous to produce these rotors as dimensionally close to each other
as possible so they may be interchanged between both units and stations. This often requires the couplings and journals on low-pressure
rotors to be sized to meet the total torque transmission and the load
requirements of the final rotor in the multiple rotor train.
•
The rotor must be sized and manufactured from materials
that are capable of withstanding high temperatures and pressures, and be able to operate for long periods in these conditions. Such rotors must possess high resistance to both
creep and rupture
•
During operation, it is possible rubs will occur between the
rotating and the stationary parts of the turbine. These rubs
can generate very high local temperatures. The rotor must be
327
Turbine Steam Path Troubleshooting and Repair—Volume Two
able to operate to the greatest extent possible without sustaining undue damage from this occurrence
•
Many rotors operate in an environment having relatively
high moisture content. The rotor material must offer resistance to both impact erosion and washing erosion
•
Although station chemistry limits, as far as possible, the
ingress of aggressive chemical compounds to the steam path,
complete freedom from chemical attack cannot be guaranteed. It is therefore necessary for rotors to be manufactured
from a material having as great a resistance as possible to
chemical reaction with any compounds that may be introduced with the steam
The rotors must be thermally stable and resist any tendency to
bow or sag as a consequence of temperature or temperature
changes. The rotors must exhibit good journal characteristics and
be capable of producing a high quality surface finish in the vicinity of journals.
Rotor construction
The form of rotor selected for any particular turbine section or
application depends upon several factors:
•
328
The manufacturing techniques developed and proven by
the manufacturer. It is reasonable to anticipate that different manufacturers should develop different techniques
based upon an extrapolation of their design philosophy,
sources of raw material, and availability. Such final selection will also be influenced by their in-house manufacturing capabilities
The Repair of Rotating Components
•
The stress levels produced by the operating torques, centrifugal loads, and bending stresses. The stresses a rotor experiences must be considered relative to the temperature at
which the components will operate, and also the temperature transients they are likely to experience during operation.
A rotor construction unsuited to high temperature operation
may well be acceptable, and structurally preferable at lower
temperatures, where requirements have changed or are significantly different
•
The rotor’s operating environment can also dictate that
certain forms of construction and materials would be
unsuitable
•
Raw material sources, purity, and integrity are a significant
factor in establishing rotor form. Recently there has been
a trend towards increasing the diameters of rotors. This
increase is normally limited by the forge master’s capability of producing larger rotor forgings of sufficient mechanical strength
•
Shipping limitations have not yet caused a limitation to rotor
size. It has, however, meant many low-pressure rotors, particularly those driving four-pole generators (1,800 and 1,500
rpm), are shipped without final assembly of the long, last
stage blades. For the majority of manufacturers, this has
required shop assembly of the long blades for testing and
balancing, and then disassembly at the manufacturing plant
before shipment. Final reassembly is made at site
There are three basic forms of rotor constructions used in modern turbines, and certain combinations of these forms in older units,
many of which are still in use. The three basic forms are:
329
Turbine Steam Path Troubleshooting and Repair—Volume Two
Monobloc. The monobloc rotor is produced as a solid forging.
After forging, the rotor may be gashed to form individual discs for
each stage or group of stages. These discs will then be machined at
their outer diameters to allow blades to be attached. The decision to
gash normally depends upon the pitch, or the ability to admit a
diaphragm in the axial spacing between the rotating blades rows.
This form of construction using a diaphragm enables steam leakage
sealing to be affected at a smaller diameter, and therefore, reduces
the leakage area. In general, because the impulse type turbine has
fewer stages than the reaction turbine, and there is a greater pitch
between them, it is possible to arrange access for the diaphragm, and
provide sealing at a smaller diameter. Figure 9.2.8 shows the outlines of a typical gashed monobloc rotor.
Inspection bore
hole
Local chambering to remove
impurities and inclusions
Fig. 9.2.8—The “gashed” monobloc rotor forging, used for the wheel and
diaphragm type unit.
Figure 9.2.9 shows the outline form of a monobloc rotor more
characteristic of the reaction design, which requires more stages to
expand the steam between the same pressure limits. This arrangement does not therefore have sufficient axial space to permit the use
of a diaphragm. Because of its greater central diameter, this rotor has
certain advantages associated with its greater stiffness and load carrying capability.
330
The Repair of Rotating Components
Typical root form
machined into
rotor surface
Inspection bore
hole
Fig. 9.2.9—The “drum” type monobloc rotor forging, as used in reaction type
rotors.
With this form of construction, the blades are attached into root
form grooves machined directly into the outer surface of the rotor
body. Because of the lower pitching, these individual stages are closer together than those associated with the impulse design.
Built-up. When the diameter required at the blade root exceeds
what the forge master can produce, and ensuring adequate material
properties, the rotors are often produced by a shrink assembly of
individual wheels onto a central forged spindle. These wheels are, in
addition to their shrink fit, keyed to the spindle. This keying is
intended to prevent any movement of the wheels during transient
operation particularly during “emergency” overspeed conditions
when the shrink fit could become loose. If the wheel does lose its
shrink fit, it would be able to rotate and/or move in an axial direction. Therefore, the key is not required to transmit any torque produced on the blading by the expanding steam, but is a simple locating device intended to limit any movement of the disc during conditions that would eliminate the shrink fit. The basic geometry of
such a built up rotor is shown as Figure 9.2.10.
331
Turbine Steam Path Troubleshooting and Repair—Volume Two
Fig. 9.2.10—The “built-up” rotor. This Figure
design 9.2.10
has individual discs shrunk onto a central
forgedThe
spindle.
This
design
is required
when rotor
diameter
"built-up"
rotor.
This design
has individual
discs shrunk
ontorequirements
a central forgedexceed
spindle. the
ability of the
to produce
larger diameter.
This forge
designmaster
is required
when rotorforgings
diameterof
requirements
exceed the ability of the
forgemaster to produce forgings of larger diameter.
Because there can be problems associated with stress concentration and the accumulation of corrosive ions at the key way and
shrink fit regions, there has been continuing pressure to develop a
means of producing the larger diameter rotors from a monobloc forging. This can now be achieved to increasing diameters. However,
there are a large number of rotors in operation of built-up construction, and they are subject to the problems associated with this form
of construction.
Welded. The welded rotor consists of a series of individually
forged discs located relative to each other in both the axial and radial direction. Each disc is located by positioning grooves and spigots,
or other devices that are used to ensure axial and radial alignment is
sufficient to provide concentricity, and help ensure no significant
“out-of-balance” forces will exist at completion of the welding
process. The individual forgings must be accurately machined on
332
The Repair of Rotating Components
their inner surfaces before welding, as access is not possible once
welding is complete. Therefore, all surface machining, evaluation,
and fit requirements must be complete and acceptable on the inner
surfaces before “vertical stacking” to undertake the welding process.
Figure 9.2.11(a) shows a rotor constructed from individually
welded discs for a double flow low-pressure section. It contains six
individual forgings, of which two have shaft end stubs integral with
the last stage disc. Figure 9.2.11(b) shows a similar rotor for a smaller unit used in a high-pressure section. In this high-pressure rotor,
shown in section (a) can be seen the four individual forging details.
In (b) are the individual forged discs before welding, and in (c) is the
final rotor after machining the root details. These four discs are welded together, forming a rotor containing both an impulse (control
stage), and reaction stages.
In both Figures 9.2.11(a) and 9.2.11(b) the individual discs are
welded together to form a continuous rotor. The preparation of the
joint for both location and welding is critical to the success of this
form of construction, because once welding is complete, there is no
access to make any correction to the inner surface. Weld defects, if
they are found to exist, must be removed and the weld remade. A
considerable advantage to this form of construction is that there is no
requirement for a central inspection borehole, and the welds are
completed in regions of low stress, where the weld and rotor components are not subject to the same high levels of stress normally
developed in an operating unit.
To complete this welding, the individual discs are stacked vertically, and then a root pass weld is applied to each joint, as shown in
Figure 9.2.12, with four welding heads placed at 90-degree positions. At completion of this root pass, the rotor is turned to the horizontal position and the welds are completed. Typical weld preparations and locating spigots are shown in Figure 9.2.13.
333
TAPPS
Turbine Steam Path Troubleshooting and Repair—Volume Two
Fig. 9.2.11(a)—A welded rotor comprising six individual forgings joined by welding.
(b)
(d)
Fig. 9.2.11(b), (c) and (d)—Welded rotor details for a small rated unit.
334
TAPPS
(c)
TAPPS
The Repair of Rotating Components
Fig. 9.2.12—Individual forgings ‘stacked’ and root pass welds being applied.
To join the individual forgings, they are stacked vertically, the
stack being checked for axial alignment, and the root pass welding
is completed using four welding heads, which are located at 90degree locations around the circumference. This weld root pass is
335
Turbine Steam Path Troubleshooting and Repair—Volume Two
deposited onto the four circumferential locations simultaneously.
This is done to eliminate uneven local heating. The root pass is
undertaken using TIG methods. After building an adequate root pass,
the rotor is turned to the horizontal position, placed in a lathe, and
the welding is completed using the submerged arc process.
w
w
w w
w w
w
w
Fig. 9.2.13—Various weld preps, in each design the locating
diameter is shown as ‘W-W’.
The hybrid rotors. It is often convenient for the designer to
employ rotors that are composites or hybrids of the monobloc, and
built-up rotor forms, as described earlier. It is also possible for some
rotors to be of a bolted construction in which interference, rather
than shrink fits, are used to locate the individual components relative to each other. The use of these other forms is evaluated and
applied where they offer advantages both in terms of performance,
or the cost and delivery of material.
336
The Repair of Rotating Components
In fact, these forms tend to have been used on older design units,
where materials of the size and quality required were not available
to support design requirements. However, certain applications are
still used to advantage, in terms of costs, delivery, and quality.
Rotor forgings
A major characteristic of turbine rotors is that the principal components are produced by forging. For the monoblocs and central
spindles of the built-up assembled rotors, the forgings are large,
require a homogeneous material, and must be produced to meet the
requirements for reliable operation for many years.
Basic production of the forging. After pouring the melt for a
rotor forging, it begins to solidify at its outer surfaces where heat is
lost through the walls of the vessel containing the liquid metal. Soon
after the melt is poured, it is common to take a ladle sample for
chemical analysis as a check on chemical composition. This small
sample may also be checked for mechanical properties.
As the melt cools from the outside impurities, nonmetallic inclusion and nitrogen bubbles are forced towards the center of the cooling ingot. When solidified the core will be the region where such
impurities are concentrated.
Note: Modern practice involves vacuum degassing, a process
by which the space above the melt is partially evacuated to assist in
removing many of the dissolved gasses, which have the probability
of producing “blow holes” if they are not effectively removed.
After cooling to a suitable temperature, the ingot is removed, and
the forging process is undertaken. For rotor forgings, the initial size
of the ingot can be taken to have an overall length of “L” and a diameter “D.” The upper surface of the ingot will have deformed to a
concave form, caused by the volumetric reduction of the metal as
the melt cools, with the cooling having occurred first at the outer
337
Turbine Steam Path Troubleshooting and Repair—Volume Two
diameters. Also, as the cooling occurs and the grains of alloy are
formed, any impurities will have migrated, or have been driven, to
the center portion of the ingot.
The production of the basic forging ingot is made in one piece, normally using basic electric vacuum degassed steel. The ingot should be
cropped to remove piping and any evidence of segregation, which could
exist at its ends. The ingot is then placed in a press, either vertically or
horizontally, and worked over its entire cross-section to a diameter ratio
of about two to one. Care must be exercised in this forging process to
ensure the axial center of the ingot remains common with the axial center of the final rotor. This is necessary to ensure impurities and other undesirable inclusions, which tend to migrate towards the center of the melt
as it cools from the outer surfaces, remain central to the rotor and can be
removed by boring and chambering. Or if unbored, this ensures that the
impurities are in a radial position where the least adverse effects to the
stresses are induced in the rotor.
Before being shipped from the forge master’s plant, the rotor can be
given a central bore, intended to remove any remaining impurities in the
material. This central bore can also include the provision of local chambering to an engineering specified maximum diameter to remove any
local impurity or blow hole concentration.
Inspection of rotor forgings during manufacture. The in-process and
final inspection phases are an integral part of the total manufacturing
process of a turbine rotor forging. These inspections are conducted at the
forge master’s works, which are performed after pouring, and also during
and after rough machining. Extensive examination is also undertaken at
the turbine manufacturer’s plant, where considerable examination and
evaluation is completed during the metal turning.
The most significant tests and examinations relative to rotor material
integrity are those performed by nondestructive techniques. It is a general requirement of most turbine manufacturers that a magnetic particle
examination is made of all exterior surfaces, and if the rotor has an inter-
338
The Repair of Rotating Components
nal borehole, these internal surfaces should be inspected visually to the
greatest extent possible.
If magnetic particle inspection indicates the presence of cracks or linear defects beyond an agreed limiting value, the turbine manufacturer
can reject the forging. However, such defects should be examined in
terms of their axial position, local stresses, and the material that will be
removed during final machining, as some defects could be removed in
this manner. Also, the forging position can often be optimized so that significant indications are removed.
General Electric Co.
Ultrasonic examination, together with magnetic particle examination, must show the forging to be free of cracks, discontinuities, flakes, fissures, seams, and laps. It is normal for the turbine manufacturer to reserve
the right, as defined in the material inspection, to reject any forging showing indications that cannot be removed without modifying the mechanical integrity of the forging. Figure 9.2.14 shows a rough machined
monobloc rotor being ultrasonically examined.
Fig. 9.2.14—The ultrasonic examination of a rotor forging.
339
Turbine Steam Path Troubleshooting and Repair—Volume Two
An ultrasonic examination is made of the rough machined forging from all possible and available surfaces. This is done before final
heat treatment and the removal of test pieces. At completion of
rough machining, final heat treatment, and the removal of test
pieces, the rotor is normally given a final ultrasonic examination by
the forge master before delivery.
If a forging is found to have a recordable defect, it is normally
referred to the purchaser (also to the turbine manufacturer, and possibly the user), who has the final responsibility for acceptance, and
must make an accept/reject decision based on the predicted duty of
the unit, the rotor stress levels, and defect locations. It is difficult to
state absolute levels of acceptability for any defect. Acceptance levels are specified by the designer/manufacturer, usually to cover both
single and cloud defect clusters.
Typically, a turbine manufacturer’s material specification would
list acceptable defects as being an isolated defect in a critical area,
and should not have its major diameter exceeding 0.04-0.08". In less
critical areas, the major diameter should not exceed about 0.150.20 ", and cloud or clustered defects should have no indications
whose major diameter is in excess of 0.05". Critical areas of a rotor
are considered those adjacent to a blade root fastening, those within 2.0" of the bore, and for monobloc rotors with integral coupling
flanges, those occurring within the coupling flange area.
Central inspection boreholes. The pouring, solidification, and
forging of a turbine rotor all tend towards a concentration of impurities and non-metallic inclusions at, or close to the axial center of the
finished forging. In setting up the forging to be machined in the lathe,
care is taken to ensure the axial center of the forging will coincide
as close as practical with the axial center of the finished rotor.
The reason for this effort is to place those inclusions and undesirable constituents in a physical position where they can be
removed by boring. Many monobloc rotors and spindles, which are
340
The Repair of Rotating Components
intended to have wheels or discs mounted or produced on them, are
bored along their entire axial length. This boring is undertaken
before final machining to ensure the bored hole remains central during subsequent machining and metal removal, and does nothing to
adversely affect the dynamic balance, which can be achieved with
the rotor. Normally, during the boring process, portions of the core
are trepanned to remove suitable test pieces, which can be analyzed
for both chemical composition and mechanical properties.
With the introduction of vacuum degassing, and the general
improvement in material production technology, some manufacturers now place sufficient confidence in their rotor forgings that they
do not bore. However, these manufacturers will undertake extensive
NDT to ensure their specifications concerning material integrity are
satisfied. Defects can go undetected, as illustrated by the failure (during unit start-up) of a rotor forging in Figure 9.2.15, where a cloud
defect has precipitated a massive rupture of a high-pressure forging.
Fig. 9.2.15—Showing a rotor forging which has failed, having a defect near its center
location. This defect has initiated a major rotor failure.
341
Turbine Steam Path Troubleshooting and Repair—Volume Two
Borasonic examinations. The borasonic examination provides a
means of determining the condition of rotors, especially those that
have been in service for a period of time, at high temperatures, and
in which there is some level of concern regarding the possibility of
cracks having developed in the rotor material.
Many rotors have been placed in service with known “cluster”
or “single” void defects close to their center, and even nonmetallic
inclusions. These defects met the engineering specifications in place
at the time the rotors were produced, and represented good engineering judgment in terms of the manufacturing capabilities of the
forge master. While these defects were acceptable at that time, it is
necessary for the owner to monitor material condition to help ensure
continued satisfactory service.
Turbine rotor discs
In many low-pressure sections, because of the increase in steam
volumetric flow, the blades must have an increased radial height and
must also be carried on rotors at a significantly increased diameter.
It has for many years been impossible for the forge master to produce
a suitable rotor of monobloc form to meet these requirements. One
alternative construction discussed previously is shown as Figure
9.2.10. This form of construction was used to increase the effective
stage diameter. With current forging technology, monobloc rotors
can be produced to meet many of these larger diameter requirements. However, there are currently many units in service with this
older construction, and these units must be maintained and must
continue to operate for many more years.
For the built-up form of construction, many of the considerations
concerning rotor design are also applicable to discs. However, a further duty imposed on a disc is the stresses induced in the disc by the
shrink fit between the spindle and disc.
342
The Repair of Rotating Components
Functions of the disc. The discs form part of the rotor, and are
attached to it through a shrink fit. This fit is sufficient to locate the
disc both axially and radially, and hold it in intimate contact with the
central spindle. The functions and requirements of the disc can be
summarized as follows:
•
To carry the blades of those rows attached to it, and to transmit the force developed on them to the central spindle. The
torque developed on the discs is transmitted to the central
spindle by the frictional fit of the disc on the spindle. Locating keys are not designed, or intended, for this torque transmission function
•
To be capable of withstanding the stresses induced in it by its
own weight, and the weight of the components it carries. The
disc must also be able to withstand the shrink fit stresses
•
To be capable of withstanding the temperature and pressure
gradients developed across it, the material having a high
resistance to both rupture and creep
•
Because discs are used in the low-pressure, low-temperature
regions of the unit with large volumetric flows, the steam conditions in this region are normally operating in a wet steam
environment in their latter stages. Therefore, it is necessary
for the material from which the discs are produced to be
capable of resisting water erosion and also, as far as possible,
resist any corrosive action associated with the corrodents that
come out of solution in the wet region. Figure 9.2.16 shows
the face of a disc in the moisture region of a geothermal unit,
which has suffered material loss from a combination of moisture washing, erosion, and corrosive attack
Forms of the discs. The disc is assembled onto the central spindle through a shrink fit, and is held in the correct axial and tangential location by keys, plugs, and/or retaining rings. These restraining
343
Turbine Steam Path Troubleshooting and Repair—Volume Two
Fig. 9.2.16—The face of a wheel from a geothermal unit that has suffered material loss
due to water being centrifuged radially out along the wheel face.
devices make no contribution towards transmitting the torque developed on the disc to the central spindle. The torque transfer is
achieved through the frictional shrink fit at the disc/spindle interface.
The disc shape is influenced, and to a degree, defined by the
loads developed on the stage and the general design requirements of
the unit. In its simplest form, the disc connects the blades to the spindle, shown in Figure 9.2.17(a), with a shrink fit existing at the surface “aa.” In this figure the shrink fit is shown to exist along the complete axial width of the disc. There are, however, designs having a
shrink fit only at the outer edges of the interface, as shown in Figure
9.2.17(b), with a central relief of the disc bore.
For multi stage rotors it is necessary to provide stationary blade
rows between the rotating rows. For this reason, it is necessary to
design these stationary rows, which have a pressure drop across them,
344
The Repair of Rotating Components
center
relief
a
a
c
a
a
a
Shaft seal
(a)
Center of shaft
(b)
Fig. 9.2.17—Characteristics of “shrunk-on”wheels.
Figure 9.2.17 In (a) is a single wheel, and in (b) a
“pair”
with a diaphragm
containing
a shaft
seal
between
them. and in (b) a "pair" with
Characteristics
of "shrunk-on"
wheels.
In (a)
is a
single wheel,
a diaphragm containing a shaft seal between them.
to provide a steam sealing system between the rotating rows or discs.
These seals are intended principally to minimize leakage, and to a
lesser extent to guide the steam and ensure it follows the designed
steam passage. The location of a typical seal system attached to a stationary blade row is shown in Figure 9.2.17(b) and 9.2.18.
Interstage Seal
Systems
Figure
.2.18seals.
Fig. 9.2.18—The
interstage
The interstage seals.
345
Turbine Steam Path Troubleshooting and Repair—Volume Two
Due to the stress levels in the discs of modern, large rated units,
it is rarely acceptable to use a straight or parallel-sided disc, as
shown in Figure 9.2.17(a). For these larger output units, the disc profiles are shaped to minimize stress levels.
For multi disc designs, a step change in spindle diameter normally exists between discs, as shown in Figure 9.2.18. This step provides a vertical shoulder, which aids in locating the disc in the correct axial position. This radial step also reduces the distance over
which the disc must travel when being assembled, when contact
between the disc inner surface and spindle diameter (at interferencefit diameter) could cause “chilling,” and cause the disc to “grip” the
wheel, making the completion of assembly impossible without first
removing the disc and reheating. This is a long, expensive, and difficult process, and one that can cause damage to the shrink surfaces.
There are other considerations of disc geometry, which need to
be considered:
•
Interstage seals—To permit a satisfactory seal to be provided
between the rotating blade rows, the discs are designed so a
seal surface can be provided. This seal can be produced at
the hub of the discs, as shown in Figure 9.2.19, or on an
axial projection from the wheel, as shown in Figure 9.2.20.
The sealing device used depends upon the design details of
the stage. However, the most effective seal design will
employ a large number of seal strips, and have the seal
formed at the smallest diameter possible
In selecting the seal form to be used, the design engineer will
evaluate the alternatives, and select the one that provides the best
return on manufacturing costs against leakage loss. As shown in Figure 9.2.20, one possible advantage of using a blade root seal on
either the inlet or discharge side, is that it could be possible to
reduce the “rotor span,” by the elimination of a diaphragm inner
346
The Repair of Rotating Components
Diaphragm Inner
Web
Disc 1
Interstage
Seals
Disc 2
Shaft
Shrunk on
Locking Ring
Fig. 9.2.19—The interstage seal system. In this design
forming a seal on the cylindrical faces of adjacent discs.
Fixed Blade
Row
Moving
Blade
Row
Radial
Clearance
Fig. 9.2.20—The basic seal formed on an axial
projection from the disc. This design is used
in certain low pressure applications.
347
Turbine Steam Path Troubleshooting and Repair—Volume Two
web. However, this can only be done when the disc stresses permit
a reduction of the axial span of the discs.
•
The disc rim—At the rim of the disc, provision is made by the
production of a root-fastening slot to enable the rotating
blades to be securely attached to it. Most discs carry only
one blade row. However, in some designs, the earlier stages
can have more than one blade row on a single disc, with the
stationary blades and their seal located between the rotating
blades. Such an arrangement is shown as Figure 9.2.21. This
is a suitable construction and permits several rotating blade
rows to be carried in a shorter axial span
Axial
Clearance
Axial
Clearance
Radial
Clearance
Fig. 9.2.21—Two rotating blade rows carried on a
single disc.
This arrangement helps reduce the overall length of the rotor,
therefore reducing its bending stress. The disadvantages with this
form of construction are the steam leakage seal between the blade
rows is made at a larger diameter, thus providing greater leakage
area, and the pressure and temperature differentials across the disc
348
The Repair of Rotating Components
are increased from a “1 stage condition” to a “2 stage condition.”
While these factors represent disadvantages, they introduce no significant problems if geometries and stage arrangements are chosen
and fully evaluated.
If the discs are in an axial plane that is used for rotor balancing,
and are to have balance weights attached to them, such weights are
normally attached at an outer radius close to the blade root.
•
Pressure balance holes—The discs, if they are to be produced with pressure balance holes, the hole geometry must
meet the same requirements for finish and location as for a
monobloc rotor. They must also be produced without discontinuities or surface gouges in the holes
One advantage for the individual discs, however, is that there are
no constraints as to the diameter at which the holes are produced,
since these holes can be produced before the discs are mounted to
the central spindle.
Many built-up assembled discs do not have pressure balance
holes. These are omitted for two reasons:
–
The first being that these are often on double flow, lowpressure sections, where the axial thrust is balanced, and
sudden changes in pressure would not significantly modify the axial thrust from one end to the other. The exception to this is that should mechanical damage to the flow
on one end occur, there would be a thrust unbalance, but
the thrust block would not be sized to cope with a single
incident
–
The second consideration is that on many low-pressure
stages, the stage is designed with a higher degree of reaction in them than is normally experienced on a rotating
blade row. Therefore, including a balance hole would
promote the flow of a considerable amount of steam
349
Turbine Steam Path Troubleshooting and Repair—Volume Two
through the hole, causing it to bypass the rotating blades
and generate no power
•
Disc to hub fillet radii—Certain discs carrying large centrifugally heavy blades have high stresses developed in them as
a consequence of their size, and the blades they carry. For
this reason the disc profile conformance is critical in establishing the stress distribution
Machining for shrink fits. The disc forging is received from the
forge master in a rough machined condition, and will have been
examined by nondestructive methods to ensure its mechanical
integrity. The turbine manufacturer has to finish machining the disc
and the spindle to ensure the correct shrink fit is obtained.
The disc is bored to the design diameter, and care is taken to
ensure the surface finish is at design specification. It is necessary to
match individual disc bores to spindle at diameters, ensuring the
bore is circular. This is done by measuring at least two, and probably four diametral positions. Readings are taken from the machined
bore to ensure it is both concentric and perpendicular to the disc
wall faces. These requirements are important to ensure the final
assembly will be concentric, disc to spindle, and the disc will be at
right angles to the spindle diameter, as shown in Figure 9.2.22. An
alternate method is to heat the disc, and then to lower the spindle
into the disc. Both methods are successful.
Machining the shrink fit is normally an individual machining operation. The disc bore is machined, measured to establish a mean bore
diameter, then the spindle is finished to achieve the required shrink fit.
Discs with either cracks initiating at the “keyway,” or discs requiring a
modification to the key and keyway geometry to avoid cracking, have
commonly been repaired by removing the disc, mounting a collar
with a small shrink fit sufficient to transmit the steam induced forces in
the blades and disc, and then remounting the re-bored disc over the
collar. This arrangement is shown in Figure 9.2.23. Such a design
350
The Repair of Rotating Components
90°
°
Hot
Clearance
Fig. 9.2.22—Lowering a disc onto
its central spindle. The disc must
be maintained at a 90° position to
permit assembly.
change cannot be initiated without an investigation of the stresses
induced in the re-bored wheel, and if material is removed from it, an
investigation of the spindle. However, this does offer an opportunity to
correct what could become a significant stress and high-risk situation.
Disc shrink fits. The shrink fit used for any stage is selected to
achieve certain design requirements:
•
The stresses must not exceed the allowable values of the
material at its operating temperature
351
Turbine Steam Path Troubleshooting and Repair—Volume Two
•
The shrink fit must be sufficient to maintain contact at all
loads that will be developed in the blade row. The shrink fit
reduces as speed increases
•
The shrink fit must be sufficient to secure the disc onto the
spindle, and transmit torque without any tangential slipping
•
The shrink fit must be maintained at all rotor speeds up to
emergency overspeed (118-120%) before the fit is lost
Disc
Central
spindle
Shrink
collar
Keys
9.2.23 to allow a disc with center bore
Fig. 9.2.23—Showing the use of a Figure
“shrink collar”
cracking to beShowing
reused. the use of a "shrink collar" to allow a disc with
center bore cracking to be reused.
Vertical assembly of the disc onto the spindle. To assemble the
discs onto the central spindle requires a shrinking operation. This
process consists of expanding the discs sufficiently, by heating, to
overcome the interference fit (diametral difference) between the
spindle diameter and the disc bore, then allowing the disc to settle
over the spindle in a correct axial alignment.
In order to achieve the correct alignment of the disc to the spindle, it is necessary to arrange the spindle in a vertical plane, for long
352
The Repair of Rotating Components
double flow rotors in a pit, with sufficient headroom above, for the
crane to lower the disc without any form of interference. This general arrangement for such an operation is shown in Figure 9.2.24.
(An alternative method, but one rarely used now, is to lower the
spindle vertically into the heated disc where the spindle weight
ensures contact between the disc hub face and the shoulder
machined onto the spindle.) The assembly then cools.
Crane
hook.
90°
Clearance between
spindle and disc
Disc heated to achieve
an acceptable clearance
from disc to spindle
Central spindle
held vertically
Fig. 9.2.24—The assembly requirements for lowering a
disc onto the central spindle after temperature requirements have been met.
353
Turbine Steam Path Troubleshooting and Repair—Volume Two
To achieve an acceptable assembly, it is necessary to arrange the
spindle vertically. Any attempt to shrink in the horizontal plane
would eventually mean contact between the cold spindle and the
hot disc at one tangential position, and almost certainly before the
disc was in the correct axial position. This would cool the disc locally over the contacting surface, due to heat conduction, which would
cause uneven cooling, local shrinkage, and distortion of the disc.
Any building that is used for this disc assembly process, whether
a pit is used or not, should be designed to be relatively free from
drafts, as these could cause uneven cooling and cause the disc to
“cock” on cooling.
The disc is preferably heated in an oven under controlled temperature conditions. The disc should be raised in temperature slowly, in an attempt to keep the temperature of the whole disc fairly
even. Under no circumstances should the disc be heated above a
specified maximum value. This value is normally about 750°F, the
exact value being determined by design requirements.
In the event this is the reassembly of an existing disc, with blades
already fitted to the disc, and with brazed coverbands, tie wire, or
erosion shields, these components should be shielded by suitable
insulating baffles, and no direct impingement of hot air on them
should be permitted. The disc is heated until the desired bore expansion is achieved. At completion of this heat soak, there should be no
temperature differential from one part of the disc to another in excess
of about 45°F.
When the desired expansion of the disc has been achieved, it is
removed from the oven, and connected by previously selected slings
to ensure it can be adjusted to a level condition. The crane then raises the disc above the spindle. Some manufacturers will oil the spindle surface on the area to have the disc assembled. This is done to
facilitate final positioning. If oil is used, care must be taken to ensure
354
The Repair of Rotating Components
the use of mineral or vegetable oil contains no corrosive compounds,
or other elements that could degrade or modify to form them.
The disc is located above, and then lowered slowly into place,
avoiding any tilting, ensuring the disc is at 90 degrees to the axis of
the spindle. The crane positioning is critical in this operation, and
should any tilting occur, the disc should not be allowed to touch the
spindle surface for any significant time, otherwise it will quench
locally. If significant contact occurs because of poor crane position,
the disc must be raised, the tilt corrected, and lowered again. When
the disc hub seats onto the shaft shoulder or retaining ring, it should
be determined that the spindle diameter and disc bore are concentric, alignment is correct, and any locating keys or pins that are not
accessible from the upper face are in place before the final lowering
of the disc.
As soon as the disc is in its correct position, the underface can
be cooled. This cooling should not be undertaken too rapidly and
any air used must cause even cooling around the entire circumference. By cooling from the underside, the lower edge of the disc hub
grips the spindle first. This cooling, if undertaken, should commence
within 15 to 20 minutes of the disc seating on the spindle. This will
allow sufficient time to adjust the hot disc to a concentric position,
and make any necessary shim adjustment. To aid in locating the disc
and maintaining a correct axial gap between discs, axial shims can
be used to ensure correct alignment. These shims are preferably
heated to within 100°F of the disc temperature, and should not project down beyond the mid portion of the hub. A typical arrangement
for shim placement is shown in Figure 9.2.25. These shims must be
removed at completion of shrinking. As the assembly cools the shims
will normally become loose.
When the disc has cooled sufficiently, it should be checked to
ensure squareness from hub to spindle faces.
355
Turbine Steam Path Troubleshooting and Repair—Volume Two
Spacer
shims
Fig. 9.2.25—Assembling discs to the central
spindle and using ‘shims’ to locate the disc.
Removal of discs. There are certain situations where discs previously assembled to spindles will have to be removed to allow some
form of corrective action to be taken. These situations include:
•
If, after the initial assembly, the disc does not meet engineering limitations for squareness, an attempt to resettle the disc
can be made. This can require the removal of the disc, and
its repositioning
To resettle a disc, it is necessary to heat it until the shrink fit is
released. The disc should be heated uniformly with gas ring burners.
The number and location of gas rings depends upon the proportions
of the disc. Figure 9.2.26 shows two typical arrangements. During
the heating process, the maximum temperature differential from one
356
The Repair of Rotating Components
part of the disc to another should not be in excess of 300ºF to 600ºF.
After re-settling, the disc should again be checked for squareness. If
on resettling, the disc does not meet squareness specifications, it
must be removed.
Note: There are limits to the amount of time a disc can be heated before temperature embrittlement of the disc occurs; the higher
the temperature, the shorter the exposure time. The time and temperature selected will depend upon the disc material specification.
Gas
heaters
Gas
heaters
Fig. 9.2.26—Arrangements for heating ‘shrunk-on’ discs prior to
removal.
•
In addition to the requirement for the removal of initially
assembled “cocked” discs, there are also instances where,
after some years of operation, it may be necessary to remove
and replace, or refurbish a disc. This occurs when the disc of
a unit that has been in service needs to be removed to allow
inspection or re-boring to employ a collar, as shown in Figure 9.2.23. Unfortunately, it is occasionally necessary to
remove several unaffected discs in order to gain access the
disc in need of repair or inspection
357
Turbine Steam Path Troubleshooting and Repair—Volume Two
For those units that have been in operation for some years, it is
necessary first to clean the rotor of any chemical deposits, dirt, or
grease that may have accumulated on it and could interfere with the
removal of the disc. This interference can occur because it makes the
task of heating the disc, without excessive heat being transferred to
the central spindle that would make it expand, more difficult. In addition, any shrunk on details, such as gland sleeves or couplings that
would interfere with the removal of the disc, must be removed.
There are situations where a tight shrink is present, or large quantities of scale or other deposits exist, and it may be necessary to cool
the central spindle of the rotor in order to release the shrink fit. This
is particularly so for rotors that have been in operation for some time,
and on which chemical deposits may act to conduct heat from the
disc to the spindle, making it difficult to achieve the desired temperature differential between the two components.
In such a situation it might be necessary to remove one plug from
the end of a bored spindle, and connect clean water inlet and outlet
pipes, as shown in Figure 9.2.27. When the disc has been removed,
the spindle must be drained and dried with suitable materials and the
plugs replaced. Occasionally, for tight interference fits, or in the case
of excessive scale, a liquefied gas, such as nitrogen, has been used to
cool the spindle. However, this practice is not recommended and
should be avoided if at all possible.
If the central spindle does not contain a borehole, the cooling of
the spindles becomes difficult, or impossible, in the case of discs that
have only a small axial clearance between them.
Disc keyways and securing. The discs transmit the force as torque
from the blades to the central spindle by means of the shrink fit friction. No means are required to maintain the disc in its original assembled location. However, a key (or other tangential locating device) is
required to locate the disc on the spindle, and to provide positive tangential location of the disc during an overspeed transient when the
shrink fit could be lost, and when the driving torque will be removed.
358
The Repair of Rotating Components
Cooling
Medium
Gas
Vent
Coupling
Removed
Gas
Burners
Bore
Plug
Support
Brackets
Fig. 9.2.27—Method for removing ‘tight’
discs by cooling the central spindle.
Most discs are designed so that up to about 118-120% overspeed they will expand elastically, and will maintain their shrink fit.
Above this speed, the shrink can be lost. Therefore at this speed, a
key or locating device must continue to hold the disc in a correct
359
Turbine Steam Path Troubleshooting and Repair—Volume Two
axial alignment, and also prevent tangential migration, which could
affect the balance of the rotor. At speeds considerably in excess of
the loss of shrink fit, the disc becomes eccentric, causing a large
rotating imbalance, which could cause rubs, and eventually disc
and rotor failure.
With early designs, the method used for locating the disc to the
shaft was a keyway design, such as shown in Figure 9.2.28. The portion in the disc being square, as shown in Figure 9.2.28(a), although
the key could have a small chamfer. Figure 9.2.28(b) shows another
form where the keyway has radiused corners, and the key is chamfered. Figure 9.2.28(c) shows the final form, using the same form of
keyway within the spindle, which was not at risk, and a semi-circular form in the disc. Unfortunately, these forms of keyway have high
stress concentration at their corners, and have led to numerous failures. These keyway constructions are also a common “hideout” for
corrosive products, which are capable of causing stress corrosion
cracks to initiate at the corners of the disc keyway. In modern units
the tendency is now away from the simple keyseat, and towards
replacing systems using low stress areas for the point of attachment.
The button type locator (shown in Fig. 9.2.29) uses a circular pin
to locate the disc to a central collar produced integral with the spindle. Subsequent, downstream discs are located to the inner ones.
The locking screw (shown in Fig. 9.2.30) attaches discs together. A
key or button would be used with the inner disc.
The surface finish and fillet radii of the keyseat are critical. Attention must be paid to ensure all fillet radii are smooth and continuous. Any machining marks produced on the disc bore surfaces must
be dressed smooth. The internal finish of buttonholes is not nearly so
critical. However, any burrs, tool marks, or other surface indications
should be dressed.
To prevent axial movement, some manufacturers use a retaining
ring to locate the disc. This ring, shown in Figure 9.2.31, is either
360
The Repair of Rotating Components
Key
Shaft
(a)
Key
Shaft
(b)
Key
Shaft
(c)
Key
Shaft
Fig. 9.2.28—Forms of the shaft to disc
key.
Screw
Attachment
Connecting
plug
disc 1
disc 2
Disc 1
Disc 2
Spindle
Fig. 9.2.29—Button connection from
the disk to a central collar.
Fig. 9.2.30—A locking screw
attaching disc 1 to disc 2.
361
Turbine Steam Path Troubleshooting and Repair—Volume Two
Seal
System
Clearance
The retaining ring
Fig. 9.2.31—A retaining ring preventing axial migration of
the disc.
shrunk in place or, as is often done, the ring is of a split type, which
is held in radial position by the disc. This retaining ring is positioned
as the disc is shrunk onto the central spindle.
COVERBAND DAMAGE,
REPAIR, AND
REFURBISHMENT METHODS
During operation, the coverband is subject to a variety of stresses due to its own weight and the forces imposed on it by the attempts
of the blade to vibrate. Many of the damage phenomena that affect
the blade have a similar influence on the tenons, and tenon hole
regions. The most common forms of coverbands and tenon damage
will be considered.
362
The Repair of Rotating Components
Impact damage
The coverbands and fastenings formed from the tenon, which
are produced from the blade vane material, can be subject to
impact damage from solid-particle debris, and erosion by water or
oxide scale. These various impacts occur as the particles are transported over the coverbands by the steam. Such impacts can remove
material from the tenon heads, and this material loss can continue
until, eventually, it weakens the clamping effect of the tenon, and
the head has insufficient material to restrain the coverbands in place
against the shear forces introduced by the centrifugal force of the
coverbands itself.
Solid-particle impact damage. Impact type damage results from
particles generated either within the steam path itself from detached
Fig. 9.3.1—Impact damage on the tenons and coverband.
363
Turbine Steam Path Troubleshooting and Repair—Volume Two
components, or as a consequence of debris carried over from the
boiler and steam leads. Damage of this type is shown in Figure 9.3.1,
and appears as a removal of tenon material at the front edge of the
head with material deformation, affecting the leading edge of the
tenon. Had a foxholed tenon been used in this instance, the damage
would have been of less significance.
Water impact damage. Water carried over from the preceding stationary blade row, as relatively large droplets will impact with the
blade vane inlet edge, and also the coverband. This water tends to be
centrifuged out, and carried between the coverband and casing inner
surface. Water above the coverband may also rebound between the
casing inner surface and the coverbands if the radial clearance is
small.
The water in the radial gap can remove material from both the
tenon and coverband. Normally coverband’s material loss, as shown in
Figure 9.3.2, is not significant. However, the loss of tenon material can
have serious consequences. Figure 9.3.2 shows a tenon with material
removed by water-impact erosion. This tenon head of Figure 9.3.2 is
Fig. 9.3.2—Moisture impact erosion on the coverband and tenons. The material loss is
becoming severe, and could result in the loss of a coverband segment.
364
The Repair of Rotating Components
Fig. 9.3.3—Tenon material loss due to moisture
impact erosion.
not recessed; therefore, the loss of head material can become serious,
possibly leading to the loss of a coverband segment. Shown in chapter
3, Figure 3.8.14 shows a similar loss pattern, but in this case with a foxholed tenon. Here the material loss is less significant, but can become
serious if there is a loss sufficient to weaken the clamping effect of the
tenon head of Figure 3.8.14, or if there has been enough material loss
that the vertical surface of the tenon hole is beginning to be uncovered.
The same loss in a fox holed tenon is shown in Figure 9.3.3.
Solid-particle erosion damage. The considerations relating to the
loss of material due to solid-particle erosion (SPE) of the blade vane
apply to the coverband also. Figure 4.8.12 of chapter 4 shows the
mechanism by which scale migrates to, and is carried over the
coverband, and therefore is able to remove from the tenon by SPE.
The loss of tenon material and exposure of the tenon hole vertical surface will reduce the clamping effect of the tenon, and therefore weaken the strength of the attachment. The frictional fit between
the tenon material and the tenon hole vertical surface is expected to
be sufficient to hold the coverbands in place when the unit is new.
365
Turbine Steam Path Troubleshooting and Repair—Volume Two
But as the unit ages, if the blades have been subjected to high alternating loads, this can weaken the attachment. The attaching frictional force reduces, and clamping is then maintained by the shear
strength of the tenon head material. Therefore, material loss of the
tenons can reduce the clamping capability of the tenon.
As material loss from the tenons continues, and the clamping effect
is lost, the most common observation of this damage is for the coverbands to begin to lift on the leading blades in each group. This is
because most often the bending moments of the coverbands are a
maximum at this location, and the material loss is normally at the leading surface of the tenons. The gap between the blade platform and the
underside of the coverbands should be checked.
One method of minimizing these types of material losses, and one
used by several manufacturers, is to employ a recessed or “foxholed”
rivet, as shown in Figures 9.3.4 and 9.3.5. In this arrangement, the
coverband has the rivet head formed internal to the foxhole, which is
below the outer surface of the coverbands. Although the outer portion
of the tenon head can still be eroded, it will take a longer period of
operation to remove enough material so that the integrity of the attachment will be influenced to the extent that corrective action is required.
Figure 9.3.5 shows the geometry of such a foxholed tenon, in which
the tenon has been protected by the thicker coverband.
If material loss has occurred to the extent the integrity of the
attachment is in jeopardy, a new design that employs the recessed (foxholed) head can often be introduced. This new attachment can be
achieved in one of two forms:
Coverbands of the same thickness. This is achieved by using the
coverbands that have been removed, without sustaining damage, or
the use of new coverbands of the same thickness, and producing the
recess or foxhole in the existing thickness. This is shown in Figure
9.3.5(a). If this is intended, the engineer should make certain checks,
366
The Repair of Rotating Components
Cl
R
H
Fig. 9.3.4—The recessed coverband. ‘H’ is the thickness, and ‘R’ the depth of the recess.
Lip
s
s
t
s
t
To
Hn
s
Ho
t
Vane
t
(a)
Mb
Tn
Mb
(b)
Fig. 9.3.5—Showing the geometry of the ‘fox-holed’ tenon for both normal and thickened coverbands.
as the effective thickness of the load bearing portion of the coverbands
has been reduced from “To” to “Ho.” These checks include:
•
The stresses in the reduced thickness section “Ho” of the
coverband should be evaluated. This will include the shear
stress on the coverband at the inner overhang, across “t-t,” and
the bending stress due to “Mb” on the reduced thickness
•
It will also be necessary to consider the shear stress in the
tenon, on section “s-s,” if it is to be formed from the original
material, without any form of weld rebuild
367
Turbine Steam Path Troubleshooting and Repair—Volume Two
Note: Because of the difficulty in defining the tenon head thickness “s-s,” some manufacturers prefer to use a loading per inch of
tenon perimeter. This is calculated by determining the centrifugal force
of one pitch of coverband, and dividing by the tenon perimeter.
•
If the blade material contains tungsten, it may be necessary to
undertake the re-peening in the hot condition
Coverbands of increased thickness. If the original coverbands
aren’t thick enough to allow a recess to be formed, it may be necessary to use a thicker coverband. This increased thickness is shown
in Figure 9.3.5(b), where the thickness has been increased to “Tn”
with an inner ledge thickness of “Hn.” In this case the thickness has
been increased so that “Hn” is equal to the original thickness “To,”
but is of a sufficient depth “Hn” that a protected head can be
formed. The stresses in the coverband and tenon should be checked
for acceptability.
Other forms of coverband damage
In addition to those forms of damage resulting in the loss of
tenon head material, other forms exist. These include:
Excessive overspeed. Should a turbine experience an excessive
overspeed transient for an extensive period, or the tenon head has
deteriorated, it is possible that the coverband could detach. Shown as
Figure 9.3.6 are tenons where the cover has detached. In this case,
because of the form of the remaining tenon head material, it is possible the original clearance between the tenon and coverband hole was
excessive, preventing a full head, and filling the hole that was formed.
Heavy radial rubs. Figure 9.3.7 shows a coverband that has sustained a heavy rub at the end pitch. When such a rub occurs, the
material is immediately heated by the frictional force that is developed, and just as quickly “quenched.” This heating/quenching cycle
causes the material to become brittle and easily cracked. Chapter 4,
368
The Repair of Rotating Components
Fig. 9.3.6—Blades in which the coverband has detached. The centrifugal
action of the cover has overcome the clamping effect of the tenon head.
Fig. 9.3.7—A coverband that has sustained a heavy rub on its thin
inlet edge.
369
Turbine Steam Path Troubleshooting and Repair—Volume Two
Figure 4.11.6 shows the same row as shown in Figure 9.3.7, where
a portion of the coverband inlet edge has detached.
Axial rubs. Axial rubs are a common occurrence, and are most
often present in the axial gap between the stationary blade discharge
and the rotating blade inlet. At this location the coverband can form
a knife-edge seal (Type B coverband), and when a rub occurs only
the knife-edge is destroyed. This causes only minor damage to the
cover, which is unlikely to lead to mechanical rupture. Such a rub is
shown in chapter 4, Figure 4.11.5.
Cracks initiating at the coverband hole. A relatively common
form of failure is the development of a crack from a coverband hole.
There are a number of possible causes, which include a poor finish
to the coverband hole, a poorly designed tenon form, requiring the
Fig. 9.3.8—A crack initiating at the tenon hole, and progressing
across the coverband. This coverband will eventually detach if
not replaced.
370
The Repair of Rotating Components
coverband hole be too sharp, causing high stress concentration, or
even over peening when the tenon head was formed. If such a condition is found when a unit is inspected, the coverband should be
replaced. Figure 9.3.8 shows such a crack. While it cannot be determined from this figure where the crack originated, it is suspected it
formed at the acute angled long tenon, and then propagated by highcycle fatigue. Examination of the fracture surface would confirm this.
The reforming of tenons
If the tenons have sustained damage, or have eroded to the extent
the integrity of the coverbands’ attachment is suspect, there are obvious cost advantages to refurbishing the tenon material so the blades
can be reused. There is an even greater advantage if this can be done
without the cost of removing the blades from the rotor. Such refurbishment can be undertaken in certain circumstances, and the tenon
can be returned to an acceptable condition. To undertake this refurbishment it is necessary to remove the existing coverband, which is
done by making a series of axial or semi-axial cuts, and removing the
material from under the rivet head. This is an operation requiring considerable care, and must be undertaken using hacksaws or cutting
discs to cut through the thickness of the coverbands, without cutting
either the tenons or the blade at its outer section. There are then available several options for reforming an adequate tenon:
Reworking the original material
By reworking the tenon material, it can be reformed so a new
coverband can be passed over it. This is shown in Figure 9.3.9. This
tenon material is then used to reform a rivet head. After removing the
coverbands, the tenon can be reformed to achieve a total height of
“Ht,” with a height above the coverbands of “Hs.” It is then necessary to evaluate the stresses and general geometry to determine if
foxholing is required.
371
Turbine Steam Path Troubleshooting and Repair—Volume Two
Drawn out tenon material used
to reform a new rivet head
Hs
Fig. 9.3.9—Form of the tenon after the coverband
has been removed.
Dependent upon the blade material, it may be necessary to heat
treat the tenon material to ensure it can be reworked without cracking. It may also be necessary to remove sharp edges to minimize the
possible effect of crack formation in the “thin” edges. In this situation it may be necessary to preheat the tenon before peening.
Weld deposit on the existing tenons
With modern welding technology, it is possible to rebuild the
tenon as shown in Figure 9.3.10. This process has been a most successful method of extending blade life, and can be undertaken without degrading the integrity of the attachment. This weld repair method
can be undertaken using either a material that is compatible with the
blade material, or one of the Inconel family of materials. The Inconel
materials do not require a stress relief operation. However, there is
normally a need to relieve the stresses from the blade material.
The weld material chosen in any particular situation depends
principally upon the operating temperature of the stage. Inconel
repair can be undertaken with the blades in-situ. A steel weld buildup often requires the blades be removed from the unit to allow the
level of pre- and post-heat treatment control required to achieve
material properties and stresses relief. However, with care even a
372
The Repair of Rotating Components
Excess
mold
thickness
Finished
form
of tenon
Weld
deposit
Copper
mold
(a)
Copper
mold.
Weld
deposit
Lip
Design
tip platform
diameter
(b)
Copper
mold
ToFinished
Ho
form
of tenon.
Weld deposit extends into the blade vane, which is
then dressed to achieve the design tip diameter.
This method removes the HAZ into the vane material.
Fig. 9.3.10—Forms of the weld rebuild. In (a) the tenon has been rebuilt.
In (b) it was necessary to remove blade tip platform material also to
achieve a satisfactory tenon.
steel rebuild can be stress relieved with the blades still mounted to
the rotor.
The weld rebuild method for tenons
The weld rebuilding of tenons is a mature procedure, and can be
used with considerable success in many applications. However,
373
Turbine Steam Path Troubleshooting and Repair—Volume Two
although the techniques are known, there is still considerable skill
required to complete this process and produce tenons that meet the
requirements of the original design, in terms of material properties,
dimensions, and therefore stress levels. There can be a considerable
degree of complexity depending upon the weld material used in any
application, and if the process is not adequately controlled, there
can be some level of distortion, twisting, or lean of the blades.
The details of the weld repair for any blade should be established
in terms of the stage material, the row operating temperature, and
the form or geometry of the tenons themselves. It is possible the weld
can be undertaken using the shielded metal arc, or gas tungsten arc
process. The general steps include the following:
•
Removal of the existing tenon material sufficient to allow the
tenon to be rebuilt
Cl
Original
tenon
outline
Cover
band
δ
Sub height ' δ' required to remove
the HAZ from region of deformation,
and prevent cracks forming upon
riveting
Fig. 9.3.11—The material to be removed from
the existing tenon, down to a height “δ”.
374
The Repair of Rotating Components
The tenon may have lost only a small amount of material. However,
after removal of the coverbands, it is considered necessary to remove
material down to a sub-height of “δ,” as shown in Figure 9.3.11.
This metal removal allows the HAZ (from welding) to be formed
low on the neck of the tenon, and in an area that will not be
deformed, or at least suffer a minimal amount of deformation from the
riveting process. There are instances when it could be advisable to
remove material from the blade platform as shown in Figure 9.3.10(b).
This is something that should be determined in terms of the tenon
shape and the thickness of the coverband.
•
After grinding the tenon down to the sub-height “δ,” the
remaining tenon material and blade outer tip surface should
be cleaned. This can conveniently be completed using a grit
blast or polishing
•
The tenon and blade tip region area should then be examined
by NDT methods to ensure no cracks are present in the blade
vane
•
Establish the amount of material to be deposited to reform the
new tenon. This requires a definition of both the cross section
and radial height of the new tenon. The tenon cross section
can be determined from the existing material, after grinding
down to the sub-height “δ.” The radial height must be established, and be sufficient that a suitable head can be reformed
by riveting. The defined requirements are shown in Figure
9.3.12. This height normally provides 0.090" to 0.11" above
the coverband
•
Prepare a copper mold that will be placed over the existing
tenon stub and guide the weld deposit process. The molds
should be designed to have sufficient clearance around the
old stub, Figure 9.3.13, that the reformed tenon can be shaped
after the mold is removed. These molds should be thicker than
375
Turbine Steam Path Troubleshooting and Repair—Volume Two
Tenon material for
forming rivet head
Cover
thickness
Tenon
height
Clearance tenon
to cover
Fig. 9.3.12—Details of the material to be left to
form a rivet head.
Tenon/Mold
clearance
Required final
form
of tenon
Mold
thickness
Form of
original tenon
Fig. 9.3.13—Details of the ‘copper mold’
which is required to form the weld rebuilt
tenon.
376
The Repair of Rotating Components
the required final height of the tenon by about 0.110" to
0.150". The molds can take several forms, either forming the
complete shape of the tenon to be rebuilt with a location to an
adjacent blade tip, or simply the tenon form with no location
The most applicable form in any stage is dependent upon the
process to be used, the stage location, and to a degree the welder
preference and experience.
•
Before welding commences, the blade must be preheated to
ensure an acceptable weld. The amount of preheat temperature to be achieved is dependent upon the blade vane and
the weld material. It is necessary for this preheat to be maintained throughout the weld deposit process. The blade vane
should be preheated for a length (at least 2.0") of the vane
that will ensure no cracks form as the vane cools
•
In making the weld deposit, it might be necessary to employ
chill blocks. This can be determined from the vane geometry
and mass. In making the deposit, the molds should be placed
over the tenon stub at its center, and the weld deposit made
around the entire tenon. The procedure for weld deposit will
depend in part on the weld process, “shielded metal arc”
(stick), or “tig.” It might also be necessary to remove the
mold and clean off the slag from the deposited weld metal
before proceeding. This is a procedure that can be established before work proceeds
•
At completion of the weld deposit, remove the mold, deslag,
wrap the blade tip in an insulating material, and allow it to
cool to room temperature
•
The tenon may, depending upon the weld metal, require a
stress relief process, of which there are several acceptable
methods. These include the use of a torch, in which the flame
is kept in constant motion over the tip section, the temperature
377
Turbine Steam Path Troubleshooting and Repair—Volume Two
being monitored by means of “tempil sticks.” This is normally
only considered suitable for short time operations up to about
20 minutes
Electric resistance heating is also very suitable, and can be used when
a precise temperature is required for a longer period.
At completion of the heat treatment, the blades should be covered
with an insulating material and allowed to cool to room temperature.
•
Reshaping the tenons requires care, and is relatively complex, particularly if the blades are still mounted in the rotor.
If the blades have been removed, it is possible to use
machine tools. But on the rotor handwork is normally necessary to establish the final form, and the clearances between
the tenons and coverband holes. It is normal to remove the
coverbands intact (to the greatest extent possible) so they
may possibly be used as a template in producing the holes in
the new coverbands. Prior to producing the tenons and new
coverbands, it is necessary to establish the tolerances that are
required for the coverband to tenon clearances
It is important that fillet radii at the base of the tenon are formed
to remove any sharp sections or discontinuities of form, which could
induce stress-concentrating regions. Also, the surface finish must be
fine, and the radius not so large there will be any interference
between the fillet radius and coverband underside chamfer.
•
At completion of the tenon rebuild and reforming process, it is
recommended a final NDE be completed of the tenons before riveting commences. Figure 9.3.14 shows a rebuilt tenon, where the
weld has not fused at one radial location. This tenon will need to
be ground away and rebuilt
When tenons are weld rebuilt by hand methods, it is typical for the
blades to remain assembled to the rotor. If the blades have been
removed, and tenons require rebuilding, another method is to weld
378
The Repair of Rotating Components
Chromalloy
Fig. 9.3.14—A weld rebuilt tenon where the weld layers have
not fused completely This must be rebuilt.
Fig. 9.3.15—The robotic weld rebuild of a tenon.
379
Chromalloy
Turbine Steam Path Troubleshooting and Repair—Volume Two
Chromalloy
Fig. 9.3.16—The raw deposited weld from
robotic rebuild.
Fig. 9.3.17—The tenons of Figure 9.3.16 after
finish machining.
380
The Repair of Rotating Components
restore the tenon robotically. This process is shown in Figure 9.3.15,
where the welding head is depositing a compatible material over a
vane chord length sufficient to form the deposit shown as Figure 9.3.16.
This weld deposit is then formed into the required tenons by machining (Fig. 9.3.17).
A concern with the refurbishment of some critical stages is the
inability to establish, by NDT methods, the condition of the weld,
particularly in the HAZ after the tenons have been riveted. A method
that can be used in this case, for vane sections that have a small turning angle, and where the vane section is relatively flat, is to weld in
a complete tip section and then to reform the airfoil. This method
has been used successfully in applications where the blades have a
long radial height, and the end user wishes confirmation that material integrity has been maintained after tenon formation. Typical
geometry of the tip rebuild is shown in Figure 9.3.18.
Weld attached
coupon
Cut off
Outer section
Weld Gap
"g"
Weld
preparation
(a) Original
form of
blade tip
Heat
affected
zone
(b)
Final
form of blade
after dressing
(c)
Fig. 9.3.18—The repair of a blade vane by removal of the outer section and the reattachment of a new tip which is then formed into suitable tenons.
(a) Is the original design.
(b) Is the weld attached coupon.
(c) Is the final material form with the HAZ removed from the tenon material.
381
Turbine Steam Path Troubleshooting and Repair—Volume Two
Screw attachment of the coverbands
In the event only a small number of tenons are involved in failure, it is possible to employ a screw arrangement as shown in Figure
9.3.19. With this method a hole is produced in the blade vane, Figure 9.3.19(a) and (b), through the center of the failed tenon. This
hole passes down through the blade vane, providing a means of
screwed attachment. After coverband attachment by means of the
screw, it is necessary to “stake” the screw to prevent its rotation during operation. Figure 9.3.20 shows an attachment of the screw type.
To ensure a suitable attachment, it is advisable to use a screw produced from titanium, and a better attachment is achieved if there is
local thickening at the tip section of the blade. Each situation must
be evaluated separately, in terms of the local stage geometry. Figure
9.3.19(c) shows another method of attachment, in which a washer is
placed over the failed tenon, and welded to the remaining material.
Note: Various forms of screw can be used; the most suitable in
any situation should be selected in terms of the geometry at the
blade vane tip. This form of attachment should only be used when
steam temperatures permit the use of titanium.
Titanium
screw
Coverband
Blade
vane
(a)
Coverband
Blade
vane
(b)
Coverband Weld
Washer
Blade
vane
Tenon
(c)
Fig. 9.3.19—Methods of securing the coverband to the blade vane outer surface
when tenon integrity has been reduced.
382
The Repair of Rotating Components
Fig. 9.3.20—The titanium screw attached cover, of the type shown in Figure
9.3.19(a).
Original vane and
tenon outline
Modified vane tip
diameter
Original vane tip
diameter
Radial position
of tenon crack
dH
Fig. 9.3.21—A blade vane shortened by an
amount “dH”. This will provide sufficient
material that a new tenon can be formed.
383
Turbine Steam Path Troubleshooting and Repair—Volume Two
Blade vane shortening
Depending upon the form of the tenon, the blade can be shortened by the small amount “dH” of Figure 9.3.21; this method is
often termed “tipping,” with enough material removed to allow the
tenon length to be increased without welding. This has the disadvantage that the height of the blade vane is reduced, which modifies
the outer “lap,” see Figure 2.12.2(a) and (b) of chapter 2. Also, for
those stages in which the coverband provides a radial seal (or seal
platform), the radial clearance and leakage will be increased, unless
a non-standard sealing device is used, or the effective depth of the
platform is increased to the original diameter. Also, if the coverband
forms an axial seal, the shortening of the blade could move the seal
point into a position where it will be unable to form an acceptable
barrier to steam leakage.
This method does, however, ensure the tenons can be reformed and
the coverband reattached. There are mechanical disadvantages to this
procedure: if the coverband segments covers a sufficient number of
blades, they may be too long, at the new reduced diameter, and will
therefore require new coverbands segments. However, for those stages
where this type of repair can be considered, this may not be significant.
It will also mean the discharge area from the blade, and therefore the
pressure will be modified. It does, however, often allow a unit to be
returned to service quickly. Another consideration, which it is difficult for
the maintenance engineer to evaluate in a short period of time, is the possible effect on blade natural frequency, and coincidence with the “nozzle
passing frequency.” This will only be of concern if the frequency margin
of the original design was insufficient.
With sufficient care this blade vane machining can be undertaken
without the need to remove the blades from the wheel to shape the
tenons. This has an obvious cost advantage.
Note: With the successful development of weld repair techniques,
and the ability to undertake repairs to even the highest temperature stages
384
The Repair of Rotating Components
on the rotor without the need to remove the blades, “tipping” should no
longer be necessary.
Rubs of the outer surface of the coverband
Instances occur where there are radial rubs between the coverband
and stationary components of the unit, or even between debris trapped
between the coverband and stationary components. These rubs are in
most instances light, between the coverband and seal strips placed above
the coverband to limit tip leakage. Such a light rub, which is relatively
common, is shown in Figure 9.3.22. This rub, while almost certainly
increasing the leakage steam flow, will have little detrimental effect on the
structural integrity of the seal strip or coverband.
Fig. 9.3.22—A coverband with light rubs on the inlet edge.
Fig. 9.3.23—A coverband with heavy rubs on the inlet edge.
385
Turbine Steam Path Troubleshooting and Repair—Volume Two
Other forms of rub have different levels of severity. Figure 9.3.23
shows a coverband that has sustained a damaging rub, causing relatively severe damage to the coverband inlet side, including damage to
the tenons. The principle concern with a rub of this nature is the hardening, which could have been caused to the coverband and more particularly the tenons, making them more brittle, and therefore subject to
failure. In such a case, hardness checks should be made to verify the
adequacy of the material. Local hardening in excess of 15-20 Brinell
points is unacceptable.
There are, in fact, some major causes of coverband outer surface
rubs. The most significant of these are:
•
debris located above the coverband
•
excessive overspeed, causing the rotor to grow radially
•
high levels of rotor vibration, particularly at start-up and
shutdown, when the unit passes through critical speeds
•
distortion of the stationary components of the unit to which
a seal strip is attached
When a rub has occurred between the coverband/tenon seal surface and the sealing strips, it should be noted what material the seals
strips are produced from. In the case of high temperature stages, the
strips are normally made from a hard alloy steel with enough
mechanical strength to resist the bending stresses induced in it by the
pressure drop across it, at a high stage temperature. In this instance
the coverbands will have suffered some damage, the extent depending upon the magnitude or severity of the rub. In the case of low temperature stages the seals are often made from a softer copper base
alloy, and the seals will in general have been damaged far more significantly than the coverbands.
Figure 9.3.24 shows the damage sustained by coverbands and
tenons when a seal strip, produced from alloy steel, detached
386
Next Page
The Repair of Rotating Components
between the coverband and casing, causing a heavy rub on the
blade row outer surface. Here there has been significant hardening
of the coverband and tenon material requiring a complete rebuild.
Similar damage situation is shown in chapter 4, Figure 4.11.7.
If the rotor has a large radial clearance above the coverband, there
is, on excessive overspeed, a tendency for the coverband to “curl.”
This curling effect can occur on both the inlet and discharge side of
the coverband, or at the segment end overhangs. Curling can occur to
the extent the rivet head is deformed outward, then on resumption of
normal operation, the integrity of the tenon will have been reduced,
Fig. 9.3.24—Coverband damage sustained when metallic debris
has been trapped between the coverband outer surface and the
casing.
387
Chapter
10
Seals, Glands, and
Sealing Systems
INTRODUCTION
The steam turbine requires that seals be provided at a number of
locations to minimize the leakage of steam between and past stationary and rotating components. These seals can be arranged, or
classified, in one of three main groups:
•
On those stationary components that enclose, and possibly
carry steam path components. These include the casings,
both inner and outer, and the diaphragms or stationary
blades. Such components are used for both high and low
(including sub-atmospheric) pressure application
•
On the rotating components, which compromise basically
the rotor, which carries and locates the rotating blades,
together with the stage hardware associated with these rows.
It can also include some “shrunk-on” components, such as
wheels or discs, coupling flanges, and thrust collars
549
Turbine Steam Path Troubleshooting and Repair—Volume Two
•
The other, basically stationary equipment, which is included
to support, connect, and control the unit operation
The steam path is defined as those components included in the
first two groups, which expand the steam and extract work from it.
The components of these two groups operate in close proximity, but
must maintain a running clearance sufficient to prevent hard contact.
In operation, steam leakage occurs between the stationary and rotating components of the unit. This leakage steam represents a bypass of
the blade system, and therefore is a waste of the energy that is available from the working fluid. To limit such leakage, sealing arrangements are made within the unit to prevent or minimize this loss.
There are three major locations within the steam path where
seals are employed:
•
Where the rotor passes through the casing, to be supported
on the journal bearings. This includes designs utilizing one
rotor to carry two expansions of the total steam path. At shaft
ends, these seals can be used to minimize the outward leakage of steam and the inward leakage of air
This location can also be considered to include the leakage,
which would occur along the shaft from one expansion portion to another at a lower pressure
550
•
At the stationary blade inner diameters to the rotor. These are
seals that, in stages employing a diaphragm construction, are
normally carried at the inner diameter of the web inner surface. In stages employing blades located directly in the casing, these seals are arranged on the inner surface of the blade
coverband
•
Those seals located above the rotating blade rows. These
seals are used to prevent steam that is discharging from the
stationary blade row by passing the rotating blades, and
therefore doing no work in the rotating blade row
Seals, Glands, and Sealing Systems
At those positions within the unit, where the stationary and rotating
components are adjacent, and require sealing, the seals can be located
in either the stationary or rotating components. The configuration chosen for any location is dependent upon a number of factors, including
the experience and preference of the design engineer. Another consideration when selecting the seal form and location is that those seals
mounted on the rotor are subjected to the centrifugal forces of rotation.
Therefore, if those seal strips normally used to produce the flow constriction are attached to the rotating components by mechanical means,
this attachment must be able to resist those forces.
STEAM PATH SEALS
During operation, many parts of the steam turbine contain highpressure, high-temperature steam, and other portions are subject to
vacuum. Also, because it is not possible to locate bearings within
either of these steam environments, and it is necessary to ensure
shaft continuity, provisions must be made for the shaft ends of the
unit to project through the casings. The casings contain these highenergy gases, so they can locate on bearings at atmospheric conditions, and also be coupled to other rotors of the system to form a
continuously coupled rotor system.
At those points where the rotor passes through the casings, and
where significant pressure/temperature differentials exist from internal to atmospheric conditions, it is necessary to provide a sealing
system between the two portions to prevent excessive “leak in” of air
at the sub-atmospheric or vacuum end, and “leak out” at the above
atmospheric locations. Such sealing is usually achieved by the use
of strips, which provide a labyrinth seal of small radial and/or axial
clearance. These seal points are also constructed to cause a minimum of damage if rubs should occur.
551
Turbine Steam Path Troubleshooting and Repair—Volume Two
In the low-pressure sections of the unit, the rotor passes through
what is normally a fabricated casing in a region where a vacuum
exists. In this case, there is a positive pressure (atmospheric to vacuum) trying to induce an ingress of air to the casing. This air, which is
not condensable at system temperatures and pressures present, would
tend to degrade the vacuum. In so doing, the air would degrade the
energy range of the steam, and therefore reduce the efficiency of the
cycle. Also, the oxygen contained within the air could react within the
unit to introduce or accelerate corrosive action.
At unit “start-up,” it is normal to first pull vacuum in the condenser, whose action produces sub-atmospheric pressures throughout the steam path. This vacuum would be sufficient to induce an
inward flow of air into the system, making it difficult to bring the unit
to speed, and to begin generating power. For these reasons it is necessary to provide the unit with a steam sealing system, which at all
internal to external seal locations is able to provide steam to the
seals, and regulate or limit the inward flows of air.
To accomplish this sealing, a system must be provided that is able
to minimize both air leakage inwards at “start-up” and at all loads
from the low-pressure sections; and outwards from the high-pressure
locations at “start-up” and during normal operation. This system must
be effective under all conditions of load and steam conditions. Therefore, it is necessary for such a system to employ some regulatory
devices that controls and limits these flows to an acceptable level.
At no load and light loads, a vacuum exists throughout the turbine unit, extending as far up the steam path as the high-pressure
section, i.e., at “start-up” the entire unit will be under vacuum.
Under these conditions, the sealing system must be capable of providing a sealing capability sufficient to provide positive atmospheric packing, at all sealing locations, and sufficient to prevent the
ingress of air, while allowing the vacuum to be maintained in the
condenser.
552
Seals, Glands, and Sealing Systems
Before the development of the current designs of sophisticated
sealing systems, it was normal to provide a positive seal at the lowpressure ends, and an atmospheric leak-off, which vented steam from
the gland housings to the atmosphere. With the present design of systems, the “atmospheric leak-off” is maintained at a pressure marginally below atmospheric, and is vented to some suitable location within the system. The lowest pressure leak-off in the system is often taken
to separate condensers, defined as “gland steam condensers,” which
are maintained at pressures marginally above (e.g., 0.5" Hga), the
main condenser pressure. In these condensers the steam is condensed, and any air present is expelled from the system to prevent its
promoting corrosion within the system, mainly in the boiler.
The constricting seal is normally obtained by the using fine,
metallic strips, which are arranged to provide a small clearance (axially or radially), to minimize the leakage area available, and therefore minimize the quantity of steam that is able to flow past them.
Such a steam flow, because no work is done, is a “throttling,” or constant enthalpy expansion. These seal strips are normally arranged to
provide a series of throttling constrictions. In doing this, there is a
continual reduction of steam pressure in successive steps from one
section, or from one chamber, to another. Depending upon the differential expansion present at the location of the seal, the strips will
be arranged as a labyrinth, in which the steam is continually forced
to change direction because the seals are being formed at alternating “high” and “low” diameters. When there is differential expansion
between stationary and rotating parts to the extent that too large a
distance would be required between the seal strips, a “straightthrough” type of construction is used, i.e., all constrictions are made
at the same sealing diameter.
Many older units still employ water-sealing glands, but their number
is reducing, as these units are retired. However, if such a unit is to be kept
in service, there are considerable advantages to replacing the water seal
with the now more conventional, and effective steam throttling system.
553
Turbine Steam Path Troubleshooting and Repair—Volume Two
The steam sealing system (general requirements)
Figure 10.2.1 shows a typical steam sealing system, capable of fulfilling the thermodynamic requirements for efficient operation. To perform these functions adequately, the system requires the incorporation
of certain special purpose devices and the availability of an adequate
steam supply. It is common practice in many power cycles to use the
boiler steam, this being the highest pressure available in the cycle and
therefore capable of sealing all the shaft glands, under all those conditions of load and pressure it experienced during operation.
Fig. 10.2.1—A typical steam sealing system, capable of
regulating sealing steam flow to both the high and low
pressure glands of a unit.
554
Seals, Glands, and Sealing Systems
Under normal operating conditions, steam is supplied to the sealing system from within the cycle, as required. The pressure of this
sealing steam is adjusted by some suitable regulator to a pressure of
from 30-45 psi for the 2,400-3,500 psi steam from the boiler, the
actual pressure depending upon the cycle design. Figure 10.2.2
shows schematically the essentials of this system regulating high and
low-pressure glands on the ends of a shaft. The steam flow “into” and
“out of” the glands is shown at the high-pressure end on the left of
this diagram. Under normal operating conditions the steam pressure
“Pa” is at a relatively high pressure and temperature and will flow out
past gland sections “1,” “2,” and “3.” It is possible for steam to be
extracted from intermediate pockets at pressure “Pc” or “Pd.”
In Figure 10.2.2, steam is shown being extracted to an intermediate point at pressure “Pc.” At position “D” steam at pressure “Pd” is
extracted and supplied to the regulator of the system. Its pressure is
adjusted to the regulator pressure of about 30-45 psi. This steam is
eventually used to seal the low-pressure systems. At position “E” the
pressure is maintained at a slightly sub-atmospheric value by use of an
Fig. 10.2.2—Schematic arrangement of the steam flow to the high and low pressure
glands under normal operating conditions.
555
Turbine Steam Path Troubleshooting and Repair—Volume Two
auxiliary gland steam condenser or alternatively a steam jet air ejector.
This leakoff takes overflow from the position “D” extraction point and
also air that is drawn in from position “F” past constriction “5.”
At the low-pressure end steam is supplied to position “H.” This
steam is at a pressure controlled by the regulator “R,” and leaks past
the seal constriction “6” into the low-pressure section, whose pressure “Pg” is at vacuum corresponding to the turbine exhaust pressure. A portion of the steam leaks past constriction “7,” where it
mixes with air at pressure “Pj,” and is taken to the auxiliary steam
condenser or steam jet air ejector.
Figure 10.2.3 shows the flow conditions at light loads. In this condition, it is assumed the high pressure point “Pa” is sub-atmospheric.
In this case, steam is taken from the external source to the regulator,
adjusted to a regulator pressure of 30-45 psi, and supplied to both
points “D” and “H,” where the steam is at the high-pressure end of the
unit. There may or may not be extraction of steam at “Pc” to a point
in the cycle. At the low-pressure end the steam will enter at “H” and
flow outward to “G” and “J.” At “J” this steam is mixed with air and
taken to the steam jet ejector or auxiliary gland steam condenser.
Fig. 10.2.3—Schematic of the gland system of Figure 10.2.1 when the unit is operating under light load conditions, and the pressure at “a” is sub-atmospheric.
556
Seals, Glands, and Sealing Systems
The steam sealing system (nuclear units)
Some nuclear cycles employ “hot steam,” i.e., the working fluid
in the turbine has been into the main reactor vessel, and therefore
has had the opportunity to have picked up some degree of radioactive matter and become contaminated. In this type of system it is
considered prudent to provide for a separate source of clean sealing
steam that is free from the possibility of contamination, so that
should outward leakage occur, it will not be a health concern.
Such a system will use a special boiler or evaporator. This steam
raising equipment will use a feedwater quality water source, which
produces steam of a quantity sufficient to seal the glands. The heat
used to generate this clean steam can be supplied from an external
source, or it can employ main steam extracted from some convenient point in the main power cycle, and use this in a heat exchanger vessel internal to the evaporator.
FUNCTIONS OF THE
STEAM SEALING SYSTEM
The steam sealing devices are part of an integrated subsystem
within the steam cycle. This subsystem is intended to provide sealing at shaft end points to prevent both the “egress” of high-pressure,
high-energy steam, and also the “ingress” of air, at sub-atmospheric
pressure locations. Ingressing air, being non-condensable, will
reduce vacuum, and therefore degrade the efficiency of the unit.
Steam, if required to provide a sealing source for use in these
devices, is extracted from some convenient point in the cycle, used,
then returned to the main steam cycle when it has performed its
duty, and been purged of air. The functions of this subsystem can be
considered to be:
557
Turbine Steam Path Troubleshooting and Repair—Volume Two
•
to maintain an essentially constant pressure steam supply to
the sealing system header to supply the shaft end seals. This
steam supply must be effective with the sealing strips in
either a new or worn condition
•
the removal of the non-condensable gases, which gain
access to the system through the shaft glands
•
to perform this sealing function over a wide variation of
steam conditions and of load requirements placed on the turbine generator set
•
the gland housings must by capable of adjustment to achieve
an optimum alignment to the rotor. These gland housings
and seals must be capable of withstanding high internal pressure and temperature, and suited to withstand or resist both
erosive, and corrosive attack from within the unit
•
the system must be capable of accepting steam blowdown,
and during operation prevent or minimize the ingress of
water to the turbine from its internals
•
the piping of the system must be capable of supplying adequate quantities of steam to the gland housings, without
exceeding acceptable velocities or incurring excessive pressure drops. This piping must be sufficiently flexible to permit
thermal expansion, and ensure end point movements do not
result in excessive stresses or reactions on the gland housing.
Such piping must also be connected so that it does not provide regions where water can collect in large quantities during “shutdown” periods
At locations internal to the various portions of the steam path,
the seal strips have a single function of minimizing the quantity of
steam, bypassing the stationary and rotating blade rows. Such leakage flow represents steam bypassing the blades and therefore doing
no work on them. This steam will also, after bypassing the blade row,
558
Seals, Glands, and Sealing Systems
re-enter the main steam flow both at high incidence angles and
velocities, which will tend to destroy locally the orderly flow of the
steam from blade row to blade row.
There are various arrangements or configurations for these sealing strips, and depending upon their individual geometry, they can
provide some of the following requirements:
•
The ability to be changed when worn
•
The capability to be spring loaded, so they are able to “back
away” from contacts, or rubs, and return to their sealing position when the contact or rub condition is removed or corrected. During such rubs these seal strips could have suffered
some “wear”
•
The ability to have a sharp or “knife” edge produced on them
at the sealing position
•
The ability to withstand the pressure differential that exists across
them, and to withstand the stresses these pressures induce
STEAM LEAKAGE
THROUGH LABYRINTH SEALS
In order to determine the leakage quantities that flow past a
labyrinth seal, it is necessary to employ the basic theory of flow. The
relationship relating leakage quantity to the physical properties or
characteristics of the steam, and the geometric arrangements of the
gland, and its dimensions is derived by the relationship of Martin.
559
Turbine Steam Path Troubleshooting and Repair—Volume Two
Consider the single seal constriction, shown as Figure 10.4.1.
From the equation for steam quantity “m” flowing through any area
“Ae”), the flow can be found from:
where:
R=
Inlet pressure/discharge pressure = P1/P2
Ae = Discharge area = 2π R ε
This equation can be represented in the more recognizable form
of “Martin’s equation,” for labyrinth seals:
where:
X=
P1 =
P2 =
Vs1 =
N=
560
The
The
The
The
The
pressure ratio across the seals P1/P2
inlet pressure
pressure at discharge
specific volume corresponding to pressure P1
number of series constrictions
Seals, Glands, and Sealing Systems
P1
Ae
=
2 π R ε
dp
P2
Steam flow
R
Fig. 10.4.1—A single sealing strip showing
the principal dimensions controlling leakage
flow.
To determine the value of constant “k,” consider the dimensions
of the equation for flow, in foot/pound/second imperial units.
This equation assumes the flow coefficient “ψ” is equal to 1.0.
Also, “lbf” is pounds force, and “lbm” is pounds mass.
For a series of “N” strips, the Martin equation assumes the pressure ratio “x” is constant across each of the series constrictions of the
total seal arrangement, and the conditions are in either the superheat
or saturated regions for the entire throttling expansion. The leakage
flow will be a maximum when the value of the pressure ratio “x”
reaches the critical value. Because the flow cannot exceed that associated with the critical pressure ratio, if the value of “x” exceeds the
critical value, the critical value from Figure 10.4.2 should be used.
In many portions of the steam turbine, labyrinth seals are used in
series groups, the pressure falling successively through each expansion (or throttling). At each throttling constriction a portion of the
total heat drop across the series gland arrangement is converted to
kinetic energy, which is subsequently destroyed in the steam chamber formed between the strips. This kinetic energy is partially
561
Turbine Steam Path Troubleshooting and Repair—Volume Two
reconverted to pressure energy as its velocity reduces in the chamber. Most of the remainder of the energy that is converted at each
constriction will be converted to heat.
Fig. 10.4.2—The critical pressure ratio across a series arrangement of labyrinth
seals.
Consider this throttling effect of the four series throttling strips in
a single gland, shown as Figure 10.4.3. The steam has an initial pressure “Pi” at entry to the gland. This condition is represented on the
Mollier diagram (Figure 10.4.4) by point “Ai.” After expanding past
this first constriction, the steam will have been reduced to condition
“Ao,” pressure “Pa.” In the chamber formed between the first and
second seal strips, the kinetic energy of the steam is destroyed, and
reconverted at constant pressure “Pa,” to condition “Bi.” From point
“Bi,” there is then a further expansion of the steam past the second
constriction, with the pressure falling to “Pb,” condition “Bo.” The
kinetic energy is again reconverted in the chamber between the second and third seal strips, raising the thermal energy level from condition “Bo” to “Ci,” at constant pressure “Pc.” This process of expansion and kinetic energy reconversion is continued throughout the
series of seal strips, until the final expansion takes the steam to condition “Do” at pressure “Pd.” The locus of the points “Ao.......Do” is
562
Seals, Glands, and Sealing Systems
called the “Fanno curve.” Note that at exit from the final strip, the
steam condition is represented by “Do,” the steam having kinetic
energy, at a reduced enthalpy “Hdo.”
ε
Pi
Pb
Pa
Pc
Pd
Steam flow
Fig. 10.4.3—A series arrangement of four seals with inlet and
outlet steam pressures of ‘Pi’ and “pd’.
Pi
H
Pa
Ai
Bi
Ao
Hdo
Pc
Pb
Ci
Di
Pd
Ei
Bo
Co
Do
s
Fig. 10.4.4—The expansion of steam through the four seals of
Figure 10.4.3 shown on the Mollier diagram.
563
Turbine Steam Path Troubleshooting and Repair—Volume Two
Any group of labyrinth seals has pressures before and after them
determined by either the cycle parameters, or the internal arrangement of the steam path parts. There are two types of flow, or pressure distribution that should be considered:
•
Those groups in which the number of constrictions is sufficiently large, the pressure ratio across each, including the
last, is less than critical
•
Those groups in which there are insufficient constrictions for
the pressure ratio, that flow through the last has a pressure
ratio “x,” which exceeds the critical value
When this critical pressure ratio has been exceeded, the flow
through a group of constrictions can be determined using the following equation:
where:
“λ” is a function of the labyrinth pressure ratio “x”
where:
N=
X=
Pi =
Pd =
Number of effective series constrictions
Pressure ratio across seals = Pi/Pd
Inlet pressure to the seals
Outlet pressure from the seals
When the critical value of “x” has been exceeded, this expression can be used, using the critical pressure ratio in place of “x” from
Figure 10.4.2.
564
Seals, Glands, and Sealing Systems
Calculation of leakage flow past blade
constrictions (small number of series seals)
As an alternative method of determining the leakage at a stage
sealing position where the pressure drop is relatively small, and only
one or two constrictions are effective, the following method can be
used, with reasonable accuracy, to determine the leakage quantity.
The steam velocity “C” through an opening with a pressure drop
“dp” across it can be determined by:
C2 =
2g 144 Vmean . dP
where:
V mean is the mean specific volume over the seals
dP is in psi
For openings 1 and 2 in series:
Consider the rotating blade tip sealing arrangement shown in Figure
10.4.5(a). Because of differential expansion, only two of the three seal
strips are effective at any one time. Here these two strips form an arrangement of seals in series. It can be assumed the pressure at inlet to the first
seal is “Pi,” and at discharge from the second seal strip is reduced to “Pe.”
It should be noted, these pressures are set not by the seal, but by the rotating blade pressure drop at the tip section, as determined by the thermodynamic design calculations for the stage. The pressure drop across the
blade tip section is “dP.” The velocity at discharge from the series seals is
“C.”
The specific volumes corresponding to pressures “Pi” and “Pe” are
“Vsi” and “Vse.” For such a small total enthalpy drop it can be assumed
“Vsi” = “Vse,” or the enthalpy drop across the first seal constriction is the
same as that across the second. This makes the velocity of discharge from
both strips the same and constant at “C,” on the reasonable assumption
that the discharge area through the two clearances is the same.
565
Turbine Steam Path Troubleshooting and Repair—Volume Two
Pi
ε
Pm
∆ Had
Pe
(a)
(b)
R
Fig. 10.4.5—In (a) is shown a radial seal arrangement, where two seals
are effective at any axial position of the moving blade. In (b) is shown the
conditions on the Mollier diagram.
The steam conditions around the tip are those shown on Figure
10.4.5(b). In this case the total enthalpy drop across the seal strips is
“∆had.”
For a design such as that shown in Figure 10.4.5(a), it is general
for “Ae1” to equal “Ae2,” which is equal to “A.” If “N” is the number of constrictions with the same leakage area “A” and coefficient
of discharge “ψ.”
However AE = 2. π .R .ε
Also the velocity “C” due to an isentropic enthalpy drop through
one throttling constriction is:
where:
“N” is the number of effective constrictions
566
Seals, Glands, and Sealing Systems
In Figure 10.4.5(a), this is two at any one operating condition,
and “∆Had” is the enthalpy drop across the seal.
Inserting equations 10.4.9 and 10.4.10 into equation 10.4.8 for
the mass flow through the two series constriction gives:
Experiment has shown a mean value of “ψ” can be taken as 0.82,
and a mean value of “Vs” between the inlet and outlet conditions
yields an accurate result.
QUANTIFYING
LABYRINTH LEAKAGE
(APPLYING THE
METHOD OF MARTIN)
The following analysis, undertaken for various portions of the
steam path, is made using the method of Martin. It is recognized that
this equation derived by Martin is theoretical, and must be modified
by specified values of the “flow coefficient.” However, this expression
has been found to yield good results in practice and therefore provides an acceptable method for operators to estimate leakage quantities for various parts of a unit. In this analysis, leakages are calculated
for both an impulse and reaction unit, to demonstrate the level of
leakage that occurs in both designs. The actual values for any unit can
be determined from information derived from the heat balance, and
567
Turbine Steam Path Troubleshooting and Repair—Volume Two
clearance measurements. These values are calculated using the heat
balance, and knowledge of the unit cold setting clearances.
Shaft end glands
Consider the high-pressure end of the turbine steam path, shown
in Figure 10.5.1, (which can represent shaft glands of either an
impulse or reaction design). In this case, the steam that leaks from
the steam path is controlled by the first section of 35 constrictions,
and the pressure ratio across this portion of the gland housing. The
sections of 20 and 15 constrictions, simply determine the quantity
that flows to other, lower-pressure points of the cycle, with the
remaining portion flowing to some cycle position at a lower pressure. Therefore, for calculating the leakage flow quantities and the
associated losses, attention must be focused on the 35-constriction
gland, in terms of controlling the steam leakage quantity, which
degrades efficiency. In fact, the points at which the leakage steam is
returned to the unit does ultimately effect the output and efficiency,
but their influence can be considered as secondary in terms of current considerations, and unit performance.
The effective seal diameter at the 35-constriction gland is
20.7125", with a radial clearance of 0.020". The upstream pressure
P1=1045 psia, with a specific volume Vs1 of 0.687 cu ft/#. The conClearance Cl1
P1,Vs1
P2,Vs2
Qa-Qb
P3,Vs3
Qb-Qc
N=35
P4
Qc+Qd
N=20
N=15
Q1
N=4
Q2
Q3
Shaft End
Dr
De1 = Effective
seal diameter
De2/Cl2
De3/Cl3
Fig. 10.5.1—The shaft end gland arrangement of a high pressure section.
568
Qa
Pa,Vsa
Seals, Glands, and Sealing Systems
ditions at the first leak off are P2=237 psia, at a specific volume
Vs2=3.00 cu ft/#. The shaft gland is then arranged to provide a 23constriction gland at a seal diameter of 18.5", again with a radial
clearance of 0.020". After this second seal system, there is a second
leak off to a pressure P3=21.00 psi with a specific volume of 34.5 cu
ft/#. There is then a third group of seals, comprising a 17-constriction
gland at a diameter of 15.5", with a radial clearance of 0.020". This
leak off is connected to a gland steam condenser at a pressure of 14.5
psia. A fourth seal section is provided to allow the ingress of air,
which goes with the leakage steam to the gland steam condenser.
Determining the leakage quantity “Q1” past the initial 35 seal
strips is:
For leakage “Q2” past the second seal system containing 23
strips,
which is overcritical, therefore use “x” = 6.12 from Figure
10.4.2. This indicates the latter constrictions will have an “over critical” pressure drop.
569
Turbine Steam Path Troubleshooting and Repair—Volume Two
Leakage past the third seal system “Q3” is found from:
The first leak off quantity = Q1 - Q2 = 13,936 - 3,53 = 10,483 #/hr
The second leak off quantity = Q2 - Q3 = 3,453 - 299 = 3,154 #/hr
The third leak off quantity = Q3 + Qa = 0.0683 + Qa
In terms of the degradation of unit output, only the first leakage,
past the 35 seal strips need to be considered. If the enthalpy at position “1” is 1,413.5 Btu/#, and at no portion is the steam returned to
the steam path to generate power, then the output loss is equal to its
initial enthalpy, minus that at the condenser, say 956.4 Btu/#.
This represents at per mil loss of 1,866.9/20 = 93.35 kW/0.001"
of radial clearance.
Note: While the loss rate is indicated at 93.35 kW/0.001" clearance, this is misleading since the radial clearance will decrease as
the unit goes into operation, and there is adjustment due to radial
growth. However, this number is of considerable value when comparing measured clearances at an outage with the design values.
This loss rate is anticipated by design and allowed for in determining unit output. It is the opening of clearances above these values that represent a real loss to the unit.
570
Seals, Glands, and Sealing Systems
60
Shaft End Leakage
Loss - 'Q1'
9,000
55
8,000
50
7,000
45
6,000
40
5,000
35
4,000
30
3,000
25
2,000
20
1,000
15
0
0
20
30
40
50
60
70
Radial Clearance in 0.001"
80
Leakage frow in 1000#/hr
Loss of Output in kilowatts
10,000
10
90
Fig. 10.5.2—The leakage and power losses for
the shaft end gland shown as figure 11.5.1
The leakage flow and output loss, as a function of radial clearance, is shown in Figure 10.5.2. These calculations assume the flow
coefficient is unchanged as any “rubbing” occurs, and the seals wear
uniformly along their length. These shaft-end packings represent a
large loss in any unit, and it should be standard practice during
maintenance outages to check this region, to determine the need to
install new strips or gland rings.
The dummy pistons of reaction turbines
In reaction-designed units, because of the axial thrust developed
on the rotating blade rows by the pressure drop that exists across
them, it is often necessary (when reversed flow sections cannot be
used) to employ a dummy, or pressure-balance piston to help counter a portion of this axial thrust, and therefore lower the load that
must be carried by the thrust block.
571
Turbine Steam Path Troubleshooting and Repair—Volume Two
T1 =
2
2
π
Dd Dr xP1
4
Dummy Piston
Seals (see details)
Pb
T2 =
P1
P2
π
2
2
Dd - Ds xP2
4
Shaft End
Packing
Blade Thrust Tb
Rotating portion
of unit
(a)
Total Thrust = Tb + T2 = T1
Dr
Dd
Ds
Dt
Dq
Fig. 10.5.3—The effect of the ‘dummy piston’ on total piston thrust.
A dummy piston is shown diagrammatically in Figure 10.5.3,
which has on it an axial thrust in both directions. The magnitude of
these thrusts is determined by the shaft diameters and pressure
intensities on the two sides of the piston. The pressures and shaft
diameters are selected to minimize, or reduce to an acceptable level
the resultant axial thrust.
Because there is a net pressure drop across the dummy piston,
and there is a rotating/stationary interface at the outer diameter, this
system requires a labyrinth seal be produced at a relatively large
diameter to limit the amount of leakage steam. The steam leaking
over the dummy then passes to the shaft end packings, or another
suitable point within the steam path.
P
k
Clo
Cli
Clo
Dummy Piston
Cli
Dummy Piston
(a)
Fig. 10.5.4—Various arrangements of dummy pistons.
572
P
P
(b)
Dummy Piston
(c)
Seals, Glands, and Sealing Systems
Figure 10.5.4 shows the basic arrangements of dummy pistons
having “N” seal strips. As an example, dimensional and steam characteristics around such a dummy as shown in Table 10.5.1:
P1 = 2050psia VS1 = 0.352cu.ft/#
x = 2050/612 = 3.350
Dimensional Characteristics: Dd = 25.00"
Cl = 0.025"
Steam Characteristics:
PS = 612psia
N = 60
Table 10.5.1—Dummy Piston Steam and Physical Characteristics.
From Martin’s equation (equation 10.5.3) the leakage quantity
“m” can be found as:
If the steam has an initial enthalpy of 1432.4 Btu/#, and a final
enthalpy at 612 psia of 1,310.2 Btu/#, the lost kilowatts at 0.025"
radial clearance is:
Therefore, the loss rate is 1,114.7/25 = 44.6 kW/0.001" of radial
clearance, in excess of design-specified clearances. Figure 10.5.5(a)
shows the losses at the dummy piston as a function of the radial
clearance between stationary and rotating components.
The dummy piston is normally located close to the thrust block. This
is done so that there are relatively small amounts of differential expansion
at this axial position, and a greater number of seal strips can be incorporated into the design. Shown in detail (a) of Figure 10.5.4 is the design
where alternate seal strips are fitted into the stationary and rotating components, and in details (b) and (c) where strips are assembled to the rotating and stationary components respectively.
573
Turbine Steam Path Troubleshooting and Repair—Volume Two
2000
60
Dummy piston
leakage loss
1600
50
1400
1200
40
1000
800
30
600
Leakage frow in 1000#/hr
Loss of Output in kilowatts
1800
400
20
200
0
0
20
30
40
60
70
50
Radial Clearance in 0.001"
80
90
Fig. 10.5.5(a)—Shown are the basic dummy piston losses in
terms of leakage flow and kilowatts.
2000
1900
Loss of output in kilowatts
1800
1700
1600
1500
1400
1300
1200
1100
1000
10
20
30
40
50
Number of Series Constrictions "N"
60
Fig. 10.5.5(b)—The effect of reducing the number of effective
seals in the dummy piston.
574
Seals, Glands, and Sealing Systems
Figure 10.5.5(b) shows, for the same dummy piston, the effect of
reducing the number of constrictions. If it is assumed the unit has a
design specification of 60 effective strips, with a radial clearance of
0.025", this will give a design leakage loss equivalent to 1,114.7 kW.
Unfortunately, these strips are normally located in a “hot” environment, where they can become brittle after a period of operation.
While designs exist for spring loading these seals, they do tend to
suffer damage due to rubs, and the effects of exposure to high temperatures. It is not uncommon to find strips, or portions of strips,
missing when such a dummy is opened for inspection. Figure 10.5.6
shows that in some older design of smaller rated reaction units, the
“dummy piston” was arranged to be stepped. This allowed the thrust
to be adjusted with greater ease.
Outer
Dummy
Inner
Dummy
Fig. 10.5.6—The reaction unit with inner and outer dummies.
Steam path seals
Within the steam path (the stationary and rotating blade system),
it is necessary to minimize the leakage past both sets of blade rows.
The number of seal strips that can be accommodated at any position
is influenced by the seal form and location, the axial width, and
arrangement of the stage. These are factors established during the
early design phase.
To illustrate the effectiveness of these various seals, portions of two
high-pressure expansions will be considered. One is an impulse design
section having five stages, the other an equivalent reaction section with
575
Turbine Steam Path Troubleshooting and Repair—Volume Two
0.80
0.70
Vs
2200
P
0.60
Steam Pressure psia
2000
1800
1600
1000
T
0.50
1400
1200
900
0.40
600
1300
850
800
1000
800
950
Temperature °F
2400
Specific Volume Cu Ft/#
nine stages. These stages, because of the difference in the design velocity ratios, cover approximately the same enthalpy range, and therefore are
comparable. In each of the two expansions, the stages have the same
vane root diameter, and a velocity ratio consistent with design philosophies. The total stages of the two expansions follow current design practice and have a rotor span of equal axial length. The steam conditions
throughout the expansion are shown in Figure 10.5.7, as a function of
expansion line enthalpy. It has been assumed in the following calculations, that both expansions have the same state line efficiency (which may
not be a fully justified assumption, but will be sufficient to allow a basic
comparison of the two design philosophies).
0.30
1320 1340 1360 1380 1400 1420 1440 1460 1480
Enthalpy BTU/#
750
700
Fig. 10.5.7—The variation of steam path conditions in a high
pressure section. These conditions relate the state line properties
as a function of enthalpy.
576
Seals, Glands, and Sealing Systems
P=2002
V=0.386
P=2240
V=0.365
P=1945
V=0.394
P=1725
V=0.426
P=1662
V=0.437
P=1465
V=0.480
P=1240
V=0.552
P=1403
V=0.497
P=1050
V=0.636
P=1180
V=0.574
P=1003
V=0.665
1
1
1
1
1
S1
S2
R1
13
R2
6
P=1971
V=0.391
S4
R3
S3
6
P=1690
V=0.432
S5
R4
4
5
P=1430
V=0.491
P=1200
V=0.566
P = Pressure in psia, V = Specific Volume in cu ft/#,
S = Stationary Blade Row, R = Rotating Blade Row.
N
R5
P=1019
V=0.654
= Number of Effective Seals
Fig. 10.5.8—A five stage impulse design steam path, and the steam conditions at
various locations that affect the quantity of steam by-passing the steam path
blade rows.
P=2168 P=2014 P=1864 P=1715
P=1568
P=1425
P=1297
P=1075
P=1175
V=0.366 V=0.381 V=0.414 V=0.429 V=0.457
V=0.492
V=0.530
V=0.618
V=0.574
P=2240
P=2080 P=1925 P=1772
P=1623
P=1480
P=1343
P=1220
P=1109
P=1020
V=0.365 V=0.372 V=0.394 V=0.418 V=0.445
V=0.477
V=0.515
V=0.537
V=0.602
V=0.648
2
2
2
2
2
2
2
2
2
S1
R1
S2 R2
2
2
P=2151
P=1991
S3
R3
2
P=1832
S4
R4
2
P=1683
S5
R5
2
P=1539
P = Pressure in psia, V = Specific Volume in cu ft/#,
S = Stationary Blade Row, R = Rotating Blade Row.
S6
R6
S7
R7
2
2
P=1264
P=1398
N
S8
R8
S9
2
2
P=1148
R9
P=1050
= Number of Effective Seals
Fi
10 5 9
Fig. 10.5.9—A nine stage reaction design
steam
path, and the steam conditions at
various locations that affect the quantity of steam bypassing blade rows of the
steam path.
577
Turbine Steam Path Troubleshooting and Repair—Volume Two
Figures 10.5.8 and 10.5.9 show the basic arrangement of the two
steam paths. Figure 10.5.8 shows the impulse unit, and Figure 10.5.9
shows the reaction design. The steam conditions at the various stage
points are shown on these figures. The vortex, or radial flow action of the
steam will cause a radial pressure gradient at discharge from the stationary blades; but no such gradient exists at discharge from the rotating
blades. This is a valid assumption for blades with a small ratio of blade
height to mean diameter.
The number of effective seals at any location is shown as “N.” These
seals are considered to be radial. However, they could be replaced by an
axial arrangement; no effort has been made to differentiate between the
effectiveness of the two. The clearance with an axial seal is a function of
the “hot running clearance,” which varies from stage to stage, and is
dependent upon the rotor differential expansion. It has been assumed,
with justification in this analysis that cold axial clearances are chosen by
design so that both “radial” and “axial” seals are equally effective during
normal operation.
For an operating unit with rubs on the axial seals, their effectiveness
will have deteriorated, and will be difficult (or impossible) to repair if the
seal is formed as an integral part of the rotating blade. Therefore, it can be
assumed axial effectiveness is variable, and possibly incapable of remedial action once damage has occurred. For this reason, no considerable
error will be introduced by assuming radial and axial are equally effective.
Impulse unit. The impulse unit, because of the larger axial space
available per stage, can be arranged to employ a greater number of seals
in series, in an effort to reduce leakage between the diaphragm and rotor,
which is where the majority of the pressure drop occurs. In this present
example, the number of seals selected is consistent with the type of unit
arrangement. (The number of seals used on any diaphragm will depend
upon their pitching, and whether castellations or “high-low” formats can
be used). The number and arrangement is influenced by diverse factors
such as the rotor differential expansion, diaphragm creep, and deflection
(see chapter 2).
578
Seals, Glands, and Sealing Systems
In this unit, the blade tip sealing has been assumed to employ
only one effective seal. This can often be increased to two or more
for larger rated units by the use of a combined axial and radial
arrangement, the use of a radial seal on both the inlet and outlet of
the coverband, or a two strip inserted segment. The axial seal is normally formed as part of the shroud or coverband. When seals are
formed as part of the rotating component of the unit, they are difficult to repair once they have suffered material loss due to rubs.
Under these conditions, to repair could require replacement of an
entire blade row. When rotors are double flow construction, it is
often difficult to produce an effective axial seal on the expansions,
which move away from the thrust block in operation. If axial seals
are produced on both flows of a double flow rotor, it is questionable
as to their effectiveness in one flow, and more reliance is normally
placed on the use of a radial seal configuration.
Shown as Figure 10.5.10 are the calculated losses due to leakage between the diaphragm and rotor. The upper curve shows the
total losses for all five stages, and the lower curves show the individual stage losses. These are plotted for radial clearance from
0.020"- 0.060". Using the mean slope of these curves, it can be calculated that the loss associated with a high-pressure section
diaphragm is about 10.5 kW/0.001"/stage (kW/mil).
The losses associated with blade tip leakage are shown in Figure
10.5.11. In this case, the loss due to increased leakage is at about the
4.0 kW/mil/stage level. If there were two effective seals at each
stage, the level would reduce to about 2.8 kW/mil/stage.
An estimate of the deterioration in cylinder output for clearances
beyond design values can be determined based on the output loss
rates in the diaphragms and blade tips. Using these two loss rates, it
is possible, when examining the high-pressure section of any unit to
be able to assess the total loss associated with rubbed clearances. At
that time the most appropriate remedial action can be selected.
579
Turbine Steam Path Troubleshooting and Repair—Volume Two
2200
Total Kilowatt Loss in 5 Stages
2000
1800
1600
1400
Total Loss for
5 impulse stages
1200
1000
800
0.020
0.030
0.040
0.050
0.060
Radial Clearance - inches
F2, N= 6
F3, N= 6
Stage Loss - Kilowatts
500
Individual Stage Losses
400
F1, N=13
300
200
100
Fig. 10.5.10—The diaphragm leakage in the impulse stage.
580
F4, N= 5
F5, N= 4
Seals, Glands, and Sealing Systems
1800
Output Loss - Kilowatts
1600
Tip Leakage Loss for 5 Stages
(Assumes 1 Effective
Constriction/stage)
1400
1200
1000
800
600
0.030
0.040
0.050
0.060
0.070
0.080
Radial Clearance - inches
Fig. 10.5.11—Blade tip leakage for the five impulse stges, with one effective seal
strip on each row.
Similar calculations can be made for an intermediate pressure
cylinder, and would yield values of about 3.0 kW/mil/stage for the
blade tip leakage, with one seal strip, and 2.0 kW/mil/stage for the
diaphragm leakage. For the low-pressure cylinder the corresponding
values would be about 1.5 kW/mil/stage, for both the blade tip and
diaphragm leakage.
Reaction unit. The reaction unit contains a greater number of
stages carried by the rotor in the same (or slightly longer) axial span.
Therefore, there is not enough axial space to provide a large number
of seal constrictions at each stationary blade row as used in the
impulse unit. However, as the same enthalpy range and pressure
drop occurs across a greater number of rows, the per stage pressure
drop is smaller. In addition, in the 50% reaction unit, the pressure
drop also occurs in two distinct and equal steps, across both the stationary and rotating blade elements. Therefore, the strip seal duty at
each sealing point is somewhat reduced.
581
Turbine Steam Path Troubleshooting and Repair—Volume Two
In order to determine the effectiveness of the sealing system, it is
necessary to calculate leakage flows, and determine their value on a
similar basis to those determined for the impulse unit. In this manner both the loss rate, and the potential for deterioration and refurbishment can be established.
The curve in Figure 10.5.12 shows the loss of output for both the
stationary and rotating blades for the nine stages. Based on these
curves, it can be estimated that steam path losses are at about the 4.0
kW/mil/stage level for both sets of blades. By analysis, a similar set
of values can be determined for an intermediate pressure cylinder at
about the 2.5 kW/mil/stage level, and a similar value of about 1.5
kW/mil/stage for a low-pressure cylinder. Both values for the intermediate and low-pressure cylinders apply to both the stationary and
rotating blades. The values calculated for the high-pressure section
assumes two effective seals at each of the stationary and rotating
blade coverbands.
Some designs of low-pressure units, both impulse and reaction,
are arranged to carry more than one row of rotating blades on a single wheel. In such a design, this can modify these values of per stage
losses, because the interstage seals are effective at a larger diameter.
This fact influences the effective leakage area. In such a case, it
would be necessary to make a separate evaluation of the design.
From the potential gains, operators might be tempted to close
seal clearances below their design value, being prepared to recognize these clearances will open during operation, probably during
the first “run-up” after return to service. While this procedure would
produce a minimum of clearance, and be consistent with what the
unit can tolerate, it must be recognized that clearances could open
further during transient operation. It must also be recognized that
rubs generate heat, and excessive amounts of heat can cause metal
embrittlement, possibly causing the failure of the entire sealing strip.
Such a decision to rub clearances must depend upon the amount of
material that must be removed by the rub. One serious disadvantage
582
Seals, Glands, and Sealing Systems
Rotating & Stationary Blade Losses in Kilowatts
2400
2200
Total Blade Leakage Losses
for 9 Stage
2000
1800
Rotating Blade
Losses
1600
1400
Stationary Blade
Losses
1200
1000
800
0.020
0.030
0.040
0.050
0.060
Stationary and rotating blade radial clearances - inches
Fig. 10.5.12—The leakage on the stationary and rotating blades of the reaction unit with two seals effective on each row.
of rubbing clearances is that the “coefficient of discharge” of the
constriction can be increased when produced by a rub. This can
effectively increase the losses to values in excess of what might have
been obtained by initially installing seals at the design values. Typically, a “mushroom” edge will have a flow coefficient greater than a
knife-edge.
The calculated losses at the blade tips in the above examples
were determined using the equation of Martin. It is perhaps more
correct to use the expression developed as equation 10.5.11. Figure
10.5.13(a) shows a comparison for a typical stage, where the Martin
equation and the equation 10.5.11 values are compared for the
blade tip section, shown as Figure 10.5.13(b).
583
Turbine Steam Path Troubleshooting and Repair—Volume Two
In these calculations it has been assumed that the number of tip
seals can be increased, i.e., no attempt has been made to consider
the effects of stage geometry. This comparison is undertaken simply
0.060"
"N"
(2)
Position (1)
P = 1100 psia
Vs = 0.6866 Cu ft/#
H = 1444.8 BTU/#
(1)
35.15"
Position (2)
P = 1030 psia
Vs = 0.7250 cu ft/#
H = 1437.6 BTU/#
Fig. 10.5.13(a)—A stage with a variable number
of radial seal strips.
Leakage Flow #/sec.
45
40
Martin Equation
Equation 10.4.11
35
30
25
20
15
10
0
1
2
3
4
5
6
Number of constrictions "N" in series
Fig. 10.5.13(b)—A comparison of the calculated leakage quantities for the
Martin and enthalpy drop methods.
584
Seals, Glands, and Sealing Systems
to demonstrate the magnitude of difference between the two methods. In both cases the coefficient of discharge “ψ” has been assumed
to be 1.0. It can be seen from the curves of Figure 10.5.13(b), that
the difference between these two curves is relatively constant at
about 30%.
Example 10.5.1
Consider the impulse stage (diaphragm and disc) construction
shown in Figure 10.5.14. Here the steam conditions are shown at the
various stage points (1).....(4). Applying Martin’s equation to the
leakage under the diaphragm, and above the rotating blade tip, the
following leakage quantities can be found (assuming clearances
under the diaphragm have a design value of 0.025" at a seal diameter of 33.6", and above the blade tip a clearance of 0.060", on a seal
diameter of 41.6"). The flow coefficient should be considered constant at “ψ” = 1.0.
Cl =0.060"
N=2
(3)
Posn. P Vs
H
(1) 1945 0.390 1442 (1)
(2) 1690 1426
(3) 17250.426 1428
(4) 1662 1424
(4)
(2)
Ds = 41.6"
C l= 0.025"
N = 12
Ds = 33.6"
Fig. 10.5.14—The stage details for example 10.5.1. Showing steam
conditions and design dimensions at the seal positions.
585
Turbine Steam Path Troubleshooting and Repair—Volume Two
Solution
Diaphragm packing leakage
The power loss is found from kW
The effect of increased clearance can also be calculated, assuming
the value of “ψ” does not change as the clearances open due to rubs:
Cl
0.025
0.035
0.045
0.055
m
12.496 17.494
22.493
27.491
dh ——————
1442-1426 = 16.0 BTU/#
dkW
211.0
295.4
379.8
464.2
0.065
0.075
32.490 37.488
————————
548.6
633.0
Blade tip leakage
and the effect of clearance changes on leakage flow quantities,
assuming an unchanged value of “ψ,” and the power loss is:
Cl
0.055
0.065
0.075
m
40.508 47.873
55.238
dh ——————
1428-1424 = 4.0
dkW
170.96 202.04
233.12
586
0.085
62.598
BTU/#
264.19
0.095
0.105
69.963 77.327
————————
295.27 326.35
Seals, Glands, and Sealing Systems
This variation of leakage loss is shown graphically in Figure
10.5.15. The effect of changing the number of series constrictions
“N” can be found using the same equation to as follows:
800
Leakage Loss: kilowatts
700
Diaphragm
Leakage
600
500
400
300
200
Blade Tip
Leakage
100
0
20
30
40
50
80
60
70
Radial Clearance: 0.001"
90
100
110
Fig. 10.5.15—The leakage loss in kilowatts, as a function of the radial clearance for the diaphragm (12 seals effective), and the blade tip.
Diaphragm packing leakage, (sensitivity to “N”). In the event
the seals under the diaphragm are damaged, and the number of
effective seals are reduced, there is a change in the leakage quantity. This is shown below:
587
Turbine Steam Path Troubleshooting and Repair—Volume Two
These losses are shown graphically in Figure 10.5.16. From this
numerical example, it can be seen that opened clearances, as in the
impulse unit, have the potential to reduce the steam path efficiency,
by reducing the total kilowatts developed in the section. This is
because any steam that leaks through the seals and bypasses the
blade rows is unavailable to do work there.
Another loss that cannot be quantified is that associated with the
steam reentering the main steam flow, and disrupting the streamline
effect after it has leaked past the seals. These losses could be as high
as those suffered through leakage, particularly on the small radial
height blades.
320
Loss in output - kilowatts
300
280
260
240
220
200
180
5
7
9
11
13
Number of Effective Seals
15
Fig. 10.5.16—The sensitivity of steam leakage quantities to the number of
effective seals.
588
Seals, Glands, and Sealing Systems
Measuring radial seal clearance
When a unit is opened for inspection, a normal maintenance
action is to measure the seal radial clearances. This is normally done
at the horizontal joint. At this time the unit is cold. In addition, the
top half casing has been removed, and therefore the casing could
have distorted to an elliptical form without the tension from the bolts.
These factors can combine to give false readings, and the field operator is left to interpret these values. It is important to consider what
should be measured, and how the values should be interpreted.
At the horizontal half joint. The readings recorded at the half
joint do not represent the actual running clearances. The clearances
at these various positions are affected by three major influences,
which include:
•
The relative radial movement of the steam path components
during operation due to temperature and stress effects (discussed in chapter 2)
•
When stationary, the rotor does not lie on its true center of
rotation, and therefore the seal gap may not be measured to
be the same on the two sides. It is acceptable to assume a
mean
•
The casing may have assumed an elliptical form as the horizontal joint bolts were released
A word of caution, however, after the readings are taken. If they
show a significant localized discrepancy on the two sides, it is suggested that when the rotor is removed, the seal strips be examined to
determine if there is heavy localized damage affecting the true readings. It is also recommended that a visual inspection be made (before
the clearances are measured) to determine if there are any abnormalities in the seal strip condition. The “feeler gauges” should be
inserted into the horizontal joint as far as possible, and at a minimum of 1.5".
589
Turbine Steam Path Troubleshooting and Repair—Volume Two
Casing distortion. When the studs are removed from the horizontal joint, it is possible for the casing to distort and move either
inwards or outwards due to some metallographic change in the casing material structure, and the release of residual stress. In fact, some
movement can be so severe the casing will “grip” the rotor, making
it difficult to remove.
Consider the casing, in its design position (shown diagrammatically in Fig. 10.5.17). In this position, there are clearances at the horizontal joint of “Kl” and “Kr,” and this clearance would also be
measured at the vertical centerline as “Kb” at the bottom, and “Kt”
at the top dead center. The seals also have a radial height “H” at
each tangential location. Also shown are details of the left hand side,
showing seal strip height as “Hl.”
Kl
Kr
Dc
Hl
Kl
Dc
Design condition with
the rotor central, and
no casing distortion.
Rotor
Casing
Seal
Strip
Kb
Fig. 10.5.17—The theoretical clearances ‘K’ and seal strip height ‘Hl’ around the
steam path.
When the casing has distorted, the clearances will change, and
possibly the seal height “Hl” will not be equal at all tangential positions, particularly if there have been localized rubs during operation.
590
Seals, Glands, and Sealing Systems
When the casing is disassembled, the horizontal joint diameter
can change, either increasing or decreasing. When the horizontal
joint diameter increases, there will be a corresponding decrease in
the vertical centerline height. Similarly, if the horizontal joint
decreases there will be a corresponding increase in the vertical centerline height. These moves must be taken into account in determining the radial clearance, and thus in calculating the leakage quantity. However, the amount of horizontal joint movement, “in” or “out,”
will not determine the vertical movement. This will be determined
only by the degree of distortion.
The horizontal joint details must be measured to establish the
clearances and the seal heights. The seal strip heights “Hl” and “Hr”
are measured on both sides, with the measured clearance “Kl” and
“Kr” on the left and right hand side of the casing.
The fact that there has been casing movement can be determined from accurate measurements of the half joint diameter “Dc”
at the horizontal joint and from a vertical centerline height measurement “Vd,” shown in Figure 10.5.18. These measurements must
be made to the casing outer sidewall, not the seal edge.
The effective clearances can be determined in the following
manner:
Mean horizontal joint clearance = (Kl + Kr)/2
From measurements and comparison of the diameter “Dc” and
drop “Vd,” the casing distortion can be determined from the fact that
for a totally cylindrical casing:
2. Vd = Dc
Assuming any movement in the casing is the same magnitude on
both sides, gives side joint movement per side of “dK”:
dK = (Dc/2) - Vd. Casing width increasing at the horizontal joint
dK = Vd - (Dc/2). Casing width decreasing at the horizontal joint
591
Turbine Steam Path Troubleshooting and Repair—Volume Two
When these casing half joint measurements are complete and
the casing movement “dK” has been determined, it is then necessary
to determine the cold clearance at the top and bottom “dead center
positions.” If the measured heights of the seal strips at the left and
right hand sides are “Hla” and “Hra,” (the “a” indicating actual
measurements), then the cold clearances can be determined:
Cold clearances “Kla” and “Kra” with casing bolted closed are:
Kla = Kl +/- dK, and Kra = Kr +/- dK
It is then necessary to establish the seal height “Hl,” “Hr,” “Ht,”
and “Hb” at the four quadrant positions. The cold clearances at the
top and bottom positions “Kta” and “Kba,” can then be determined
in terms of the seal heights at the left “Hl” and right “Hr.”
Recognizing that these readings are taken without the casing
bolted, and cold, it must be accepted that errors can, and will, exist.
However, the most significant consideration in taking clearances and
determining leakage losses is to be sure the procedures for measurements are always the same. This is necessary so that differences from
one time period to the next are consistent. The leakage loss calculations are used to determine clearance differences, and these differences can only be compared by the use of a repeatable procedure.
Clearances at the top and bottom dead center “Kb” and “Kt” are
also taken with leads. However, unless the casing is bolted the clearances will not define the ellipticity of the casing. Also, if the casing
half joint has increased as the joint bolts were removed, it will be
impossible to bolt after the leads are used, as the rotor in the unbolted condition will already compress them.
Measuring seal axial clearances
The axial movement during operation is controlled by the differential expansion (see chapter 2), and the relative movement of the
592
Seals, Glands, and Sealing Systems
Dc
L
Vd
Fig. 10.5.18—The drop check (vertical centerline height) from
the horizontal joint, to establish the presence of ovality within
the casing.
rotor from the thrust block. There are situations where the running
axial clearance between the stationary and rotating components of
the unit change and decrease. This movement can occur to the
extent that the axial seals, possibly produced integrally with the
coverband will rub, removing material from the knife-edge, opening
the axial clearance, and increasing leakage loss.
These clearances must be measured as the steam path becomes
available for inspection, and the leakage loss determined. The equation of Martin will allow such determination.
Calculated values of incremental leakage loss
When a unit is opened for inspection, and clearances are measured, it is clear that even if no wear has occurred, the measured
value of clearance does not represent the “hot running” condition.
The actual running clearance is modified by those phenomena that
modify the spatial relationships within the steam path (see chapter 2). However, these changes are relatively small in the radial
593
Turbine Steam Path Troubleshooting and Repair—Volume Two
direction, and tend to be of a value that they can be ignored in terms
of their effect at the “hot end” of the unit, which is where leakage is
most significant. Therefore, if the losses are calculated on these
measured values, they will provide a loss that is in excess of the
actual value. This is because the rotor will tend to move towards the
casing during operation, reducing this measured value.
In terms of determining “increased leakage losses due to rubs,”
the measured clearance values at an outage can be used with complete accuracy, when compared to previous and design values, to
establish the increased losses. As an example, consider a location
within the unit at which the design radial clearance is specified as
0.025", and at installation the measured clearance was 0.027"
(which represents design conformance, and is within tolerance). If
this same clearance is measured on removal of the unit from service,
and the measured value is 0.056", then there will have been an
opening of the seals of (0.056-0.027 = 0.029"), if the loss rate is 3.9
kW/mil. Under these circumstances, the losses are:
Anticipated design loss:
As installed:
As removed from service:
25 x 3.91 = 97.75 kW
27 x 3.91 = 105.57 kW
56 x 3.91 = 218.96 kW
Therefore, the recoverable losses are: 218.96 - 105.57 113.39kW
or alternately:
218.96 - 97.75 = 121.39kW
This recoverable loss is dependent upon restoring the clearance
to its design value.
Therefore, the measured opening represents a true loss of (56 27) = 29 x 3.91 = 113.39 kW, or 121.21 kW, which loss can be
recovered by replacing the seals, and re-establishing the clearance
at 0.027" or 0.025". If the clearance were re-established at some
other value, the output loss would change accordingly.
594
Seals, Glands, and Sealing Systems
THE ECONOMICS OF
SEAL MAINTENANCE
The one category of loss over which station maintenance staff
can exert a significant influence, in terms of being able to plan for,
gauge and take corrective action, is the control of the internal leakage, allowing the steam to bypass the steam path blade elements.
During operation the most likely factor to influence the sealing
efficiency is for a rub to occur. Such a rub wears the seal, and has
two detrimental effects. First, it will open the clearance, causing an
increase in leakage area. Secondly, the rub will modify the form or
shape of the seal strip, possibly increasing the “discharge coefficient.” Unfortunately, both these effects tend to increase leakage
flow and therefore cause a deterioration of expansion efficiency.
It is important for the operator to know what level of increase in
leakage area can be tolerated, and the financial penalty associated
with such leakage. It is essential to consider these questions, and
while it is difficult to provide an all encompassing answer, operators
should be aware of the magnitude of this leakage for their units, and
what this represents in terms of financial penalties.
Some manufacturers provide operators with guidance concerning the loss rate in kW/mil, and this can be used to establish the need
for replacement of the seals, assuming they are a type that can be
replaced. These recommendations are made usually as a function of
seal wear. Such “rules of thumb,” while being an acceptable guide
to fuel costs, do not allow the operator to readily assess the effect of
changing fuel costs, load factors, and the value of incremental kilowatts. The following analysis, although requiring a greater level of
information preparation on the part of the owner, will normally
allow him to factor in more variables that influence overall operating costs.
595
Turbine Steam Path Troubleshooting and Repair—Volume Two
If a unit is opened only every four to eight years for planned
maintenance, there are considerable advantages to having tools
available that allow a quick and accurate assessment of the losses,
and financial penalties that can be anticipated during the next operating period. The following method allows the owner to predict
additional anticipated fuel costs associated with leakage. However,
there are several factors that must be considered when using this
analysis:
•
To be of real value, it is necessary for the operator to know
the mil loss rate for the various seal positions in the unit. This
can best be achieved by the construction of curves, or a calculated loss rate for each position in the unit
•
The operator should be able to predict, with reasonable accuracy, the fuel cost changes during the next operating period,
and before the next planned outage. While difficult, this will
allow a more accurate prediction of total revenue loss
If the actual fuel costs change in an unpredictable manner during the operating period, such an analysis can be used to help establish the possible economies of reducing the operating period to the
next outage, if the system can tolerate a change.
•
The operator must know the station heat rate (SHR), i.e., the
heat rate for the boiler-turbine-generator cycle
•
The operator must be able to predict the load factor (LF) and
possibly its variation, with some accuracy for the next operating period
With these factors established, within acceptable limits, it is possible to determine, with reasonable accuracy, what additional fuel
costs will be incurred during the next operating period due to excess
leakage. Alternatively, it is possible to estimate the fuel cost savings
available from making improvements to the seal system, and to justify the cost of changing seal strips or segments.
596
Seals, Glands, and Sealing Systems
Example 10.6.1
Consider a unit with an annual load factor of 80% (0.80), having
a station heat rate (SHR) of 10,000 Btu/kW-h, and burning fuel that
costs 200c/million Btus. It is necessary to predict the annual additional fuel cost per kilowatt, as the result of seal wear and increased
leakage.
With this unit, it costs 200c to generate 1,000,000/10,000
= 100.00 kW for 1 hour.
Therefore, any action that can be taken to improve output by one
kilowatt represents a savings. The annual fuel saving per kilowatt is
equal to:
Fuel Cost
= 200c/10,000kW
= 0.020c/kW
$ (Savings) = 1 x 8,760 x 0.8 x 0.020 = 140.16 $/kW/annum.
It should be noted that this savings is independent of the unit rating,
i.e., a kilowatt lost from a 30,000 kW unit is just as expensive as a kilowatt lost from a 750,000 kW machine, if the fuel costs are the same.
Figure 10.6.1 shows a nomogram, which permits additional
costs on a per kilowatt basis to be established for a variety of fuel
costs and load factors (LF) for a unit with a station heat rate of 10,000
Btu/kW-h. For other station heat rates, the losses/kilowatt can be
determined by the ratio of the actual “SHR” to 10,000 Btu/kW-h.
597
Turbine Steam Path Troubleshooting and Repair—Volume Two
Figure on Curves are fuel
costs in c/million BTUs
SHR = 10,000 BTU/kw-Hr.
220
300
250
200
Incremental Fuel Costs $/Annum/Kilowatt loss
275
225
180
200
160
175
140
150
120
125
100
100
80
75
60
40
20
30
40
60
70
80
90
50
Annual Load Factor %.
100
Fig. 10.6.1—The annual fuel cost savings in $/kW as a function of fuel
cost and unit load factor.
Example 10.6.2
A unit consumes fuel costing 200c/10E6 Btus, has a load factor
of 80%, and a station heat rate of 9,650 BTU/kW-hr. What are the
losses per kilowatt of lost output due to leakage?
598
Seals, Glands, and Sealing Systems
From Figure 10.6.1, the loss for a unit with a SHR of 10,000
Btu/kW-h is:
$140.16/kW/Annum
The loss per kilowatt for an SHR of 9,650 Btu/kW-h/year is:
= $140.16 x
9,650
= 135.25$/kW/Annum
10,000
In terms of assessing the losses over an extended period of operation, and increasing clearance due to rubs, the following analysis
can be made.
Predicting kW lost due to excess clearance
The normal manner of predicting leakage quantities and converting
these to kilowatts is to use Martin’s equation for leakage past labyrinth
seals. However, a simplified, although less accurate method, is to determine a mean leakage for various stage and unit configurations, and apply
these mean losses to measured clearances. It is difficult to specify an overall loss rate for any location. Typical ranges are shown below; the actual
values will depend upon the unit arrangement:
2,400 psi/1,000/1,000°F
High-pressure section/flow
Reheat section/flow
Low-pressure section/flow
N1 shaft packing
N2 shaft packing
N3 shaft packing
4.25
2.25
0.75
6.75
16.00
1.50
to
to
to
to
to
to
10.00 kW/mil
8.25 kW/mil
1.25 kW/mil
19.00 kW/mil
55.00 kW/mil
5.00 kW/mil
599
Turbine Steam Path Troubleshooting and Repair—Volume Two
3,500 psi/1,000°F/1,000 F
High-pressure section/flow
Reheat section/flow
Low-pressure section/flow
N1 shaft packing
N2 shaft packing
N3 shaft packing
5.00
3.25
0.75
7.00
20.00
1.75
to
to
to
to
to
to
12.25 kW/mil
9.50 kW/mil
1.25 kW/mil
21.00 kW/mil
55.00 kW/mil
6.25 kW/mil
Example 10.6.3
As an example of this estimation, consider a unit having a combined high-pressure (HP)/reheat section, and a single double-flow,
low-pressure section. The high-pressure section contains 7 stages (6
diaphragms and a nozzle box), the reheat section has 6 (6
diaphragms), and each low-pressure section has 7 stages (6
diaphragms and an inlet nozzle box). This unit is shown diagrammatically in Figure 10.6.2.
Loss Rate/0.001"
8.00
12.50
HP
Cl (N1) = 28
1.00
6.25
Double Flow LP
Rht
Cl (N2) = 28
Cl (HP) = 37
LP 'A'
LP 'B'
Cl (LPb) = 28
Cl (LPa) = 28
Generator
Cl (N3) = 28
CL (Rht) = 30
Clearance in 0.001"
1.00
3.25
40.00
Fig. 10.6.2—The radial clearance and loss rate in kW/0.001" for the unit in
example 10.6.3
This unit has steam conditions of 2,400 psia/1,000ºF/1,000ºF an
“SHR” of 10,150 Btu/kW-h, and a mean load factor of 75%. The fuel
costs 175c/10E6 Btus.
600
Seals, Glands, and Sealing Systems
As removed from service, the high-pressure stages have rubbed
open an average of 37 mils, the reheat section has an increased
clearance of 30 mils, and the low-pressure stages have an increased
clearance of 25 mils. The N1, N2, and N3 packings have opened an
average of 28 mils.
In this unit the assumed loss rates in kilowatt/mil are: HP 8.00,
Rht=6.25, LP=1.00, N1=12.50, N2=40.00 and N3=3.25 kilowatts
/0.001".
The total leakage losses are:
HP section:
Reheat section:
LP sections
N1 packing:
N2 packing
N3 packing:
6 x 37 x
8.00
6 x 30 x
6.25
2 x 6 x 28 X
1.00
1 x 28 x
12.50
1 x 28 x
40.00
1 x 28 x
3.25
Total losses are:
=
=
=
=
=
=
1,776.00
1,125.00
336.00
350.00
1,120.00
91.00
4,798.00
kW
kW
kW
kW
kW
kW
kW
With a fuel cost 175c/10E6 Btus, and a load factor of 75%, it can
be determined from Figure 10.6.1 that the annual cost of leakage
was $115.00/kW.
Therefore, for the last year of operation the annual fuel cost
increase was:
4,798 x $115 = $551,770
The cost in the previous years would naturally have been less,
assuming the damage in those years (immediately after startup) was
less severe.
601
Turbine Steam Path Troubleshooting and Repair—Volume Two
Example 10.6.4
As an example, consider a unit with an annual load factor of
70% (0.70), having a station heat rate of 10,000 Btu/kW-h, and using
fuel that costs 200c/million Btus. It is necessary to predict the annual additional fuel cost as the result of seal wear. It is assumed the
SHR and LF will remain constant over the next operating period.
Solution
With this unit it costs 200c to generate 1,000,000/10,000 = 100
kW for 1 hour.
Therefore, any action that can be taken to improve output by one
kilowatt represents a savings. The annual fuel saving per kilowatt is
therefore equal to:
Fuel Cost
= 200c/10,000kW
$ (Savings) = 1 x 8,760 x 0.7 x 0.02
= 0.020c/kW
= 122.64
The curve in Figure 10.6.1 shows the annual fuel cost savings for
fuel costs from $0.75 to $3.00 per million Btus (75 to 300 cents), as
a function of unit load factor, and for a SHR of 10,000 Btu/kW-h.
As an example of the application of this curve, consider a
500,000 kW unit having an SHR of 10,000 Btu/kW-h, and a predicted load factor of 70%. Also assume maintenance work has reduced
the leakage loss by 2,350 kW. In this case, this represents an annual
fuel saving of:
2,350 x 8,760 x 0.7 x 0.02 = $288,204
(with a fuel cost of 200 cents/million Btus)
Note: There are 8,760 hours in a normal 365-day year)
This amount represents the savings in the first year after return to
service. For subsequent years it is possible the fuel costs will
increase. It is also certain the initial improvement in clearance will
602
Seals, Glands, and Sealing Systems
not be sustained throughout the operating period to the next maintenance outage. Based on these assumptions, and a five-year maintenance cycle, the following cost savings can be anticipated, when
the load factor remains at 70% (0.7).
Years after
return to
service
Remaining kW
improvement
(assumed)
2
3
4
5
2,010
1,860
1,760
1,670
Predicted
Fuel Cost
Annual Fuel
Cost Savings
210
221
232
243
Total savings
$1,258,832
$1,252,062
$1,250,382
$1,248,843
$1,010,119
These savings of $1,010,119 together with those in the first year
($288,204) represent a total potential saving of $1,298,323. If these
later years are returned to original year dollars at the rate of 8%, this
represents a total present day worth of $1,125,000.
It is clear from this example that considerable savings in operating costs can be achieved by attention to maintenance of the sealing
system. Or, when output is limited by equipment of the cycle other
than the turbine, limiting, or controlling the leakage flow can
achieve some increase. One factor that needs to be considered,
however, is that the clearances can be opened to the five-year level
by one bad operating experience. Therefore, it is necessary to limit
transient, and other phenomena known to cause seal damage as
much as possible.
The extent to which such improvement can be maintained, is
dependent upon the manner in which the unit is operated, and the
transients to which it is subjected. However, if seals have been maintained at, or near, the design or installed values, it is usually possible
and economically justifiable to examine the unit to determine what
level of improvement can be made.
603
Turbine Steam Path Troubleshooting and Repair—Volume Two
It is recommended the unit seal system loss rates be analyzed
before the unit is removed from service. In this manner, measured
clearances, and the predicted loss rates can be determined. Then
when clearances are measured at the outage, decisions can be made
immediately in terms of the extent of wear and damage that can be
tolerated at the various seal points.
It is normal for both “utility” and “industrial” operators to carry, as
inventory spares, many of the seal strips and gland segments that could
be used in making upgrading repairs. This is a prudent practice and
allows remedial action to be taken at short notice.
In assessing changes and measuring the effect of any improvement on
the unit after maintenance, it is recommended, and considered necessary,
to calibrate the unit performance by means of some performance test. It
is strongly suggested that in the undertaking of such tests, that instruments
and procedures, which accord with a recognized Power Test Code be
observed. The most important aspect of such tests is to ensure repeatability of the results, and then to ensure they are run at a frequency that monitors condition, and helps assess the need for further maintenance.
FORMS OF THE SEAL
KNIFE-EDGE DISCHARGE
COEFFICIENTS
The form of the seal strip (or tooth edge) at discharge from the
upstream to downstream positions has a considerable influence on the
total flow that occurs at any seal constriction. In establishing the flow
using the equation of Martin, a discharge coefficient of “ψ = 1.0" was
assumed. In fact, the coefficient of discharge will be other than this value
and a reasonable mean of “ψ,” if no better data are available, is 0.82. This
604
Seals, Glands, and Sealing Systems
value represents an acceptable mean and can be used with the knowledge that it represents a value suited for many seal configurations. However, leakage flow is directly proportional to the discharge coefficient.
Therefore, there could be situations where the use of a more accurate
value is justified.
The difficulty in establishing absolute values of “ψ” is that the shape
of the seal strip influences the value, and the only absolute method is to
undertake expensive tests to determine the most appropriate coefficient
in each design. It is also necessary to consider the seal location, the form
of the surface on which the seal constriction is produced, and general
stage geometry at that location.
Knife-edge form
The discharge point on the seal has a marked effect on the discharge
coefficient. Considerable effort and manufacturing expense can be justified in achieving the design requirements. Figure 10.7.1 shows the simplest form of sealing strip, which is a simple constant section strip of
width “d.” Figure 10.7.2 shows various forms of the tapered knife-edge.
In (a) is the form where a taper is produced on one side, and in (b) there
is a two-edge taper. In (c) is the stepped tapered form, in which the final
thickness is “x,” and is maintained over a radial height of “Li.” In each
case, from a thickness “d” to a final knife-edge of thickness “x,” the taper
occurs over a radial length of “L.” There are two reasons for bringing the
discharge point to a knife-edge—first, to reduce the leakage quantity by
reducing the flow coefficient, and secondly, to cause less heat to be generated in the event of a rub on the mating surface.
In the event of a rub, the knife-edge will be deformed to a shape like
that shown in Figure 10.7.3(a). If the inner diameter of the seal has been
increased from “Ds” to “Dt,” there are considerable advantages to “dressing” the remaining strip to a form like that shown in Figures (b), (c), and
(d), if the preferred option of changing the strip and restoring the original
seal diameter “Ds” is not possible.
605
Turbine Steam Path Troubleshooting and Repair—Volume Two
The forms of strip described in Figure 10.7.1 and 10.7.2 are
inserted, with the seal strip being in a true radial direction. There are
also seals where the inclination of the seal strip is at an angle “λ” as
shown in Figure 10.7.4. These strips are normally produced in segments, and the final production of the seal diameter “Ds,” can be difficult to machine if the segment is spring loaded, but can be trimmed
to the correct diameter to achieve the design clearance.
L
d
Fig. 10.7.1—The simple
inserted seal strip
Fig. 10.7.2—Various options for forming the tapered
knife edge.
606
Seals, Glands, and Sealing Systems
Fig. 10.7.3—Methods for finish trimming a damaged seal strip after a ‘rub’.
With the exception of the angled strip (shown in Fig. 10.7.4), the
strips are inserted into the main body of the carrying component,
either the casing, the rotor, or gland ring. However, other forms have
the strips formed integral with the ring of the seal gland. Again, these
strips are tapered to a knife-edge of thickness “x” and have an
included angle “λ.” These seal strips cannot be replaced, and should
a significant rub occur, the seal must be replaced, or sharpened, but
to a larger diameter, which increases the leakage flow.
λ
E
X
Fig. 10.7.4—A formed seal strip. This
shape of strip is manufactured using
form tools and provides for greater
axial strength.
607
Turbine Steam Path Troubleshooting and Repair—Volume Two
Seal location geometry
The geometry of the component at the seal location has considerable influence on the leakage quantity. The seal operates on the
basis of converting the steam thermal potential energy to kinetic
energy, and then destroying that kinetic energy to the greatest extent
possible. Consider the seal shown in Figure 10.7.5(a). Here a simple,
single constriction is provided, and the seal provides a barrier to
flow, which reduces the leakage. However, the kinetic energy is not
destroyed, and the steam flows freely into the downstream space. In
Figure 10.7.5(b), a vertical face is presented to the leaking steam,
which effectively reduces the velocity, and therefore lowers the leakage quantity. This is because a portion of the kinetic energy is reconverted to pressure at discharge, reducing the pressure ratio across the
seal and therefore reducing the leakage steam quantity.
At some locations a vertical face is produced by the inclusion of
a diverting device, as shown in Figure 10.7.6 for two locations, in (a)
on the coverband, and in (b) as a rotor castellation.
In Figure 10.7.7(a) two strips are used, and a chamber is formed
between them. This chamber acts to completely destroy the velocity, but some small quantity will be carried through the second leakage area. It can be seen that the horizontal pitching “p” between the
strips can become too close to allow the effective destruction of the
kinetic energy. The design engineer must consider this, and there is
no advantage to reducing the pitch to increase the number of seal
strips that can be accommodated in any axial length.
In Figure 10.7.7(b) the strips are mounted alternately in the stationary and rotating components of the seal region. The steam is provided with a tortuous path, and a considerable portion of the energy
is destroyed by the continual change of stream direction. Here the
axial gap between stationary and rotating seals strips is governed by
the differential expansion that occurs, and this pitch “p” will change
with changing operating conditions.
608
Seals, Glands, and Sealing Systems
Cl
(b)
(a)
Fig. 10.7.5—Two arrangements of a single seal. In
(a) the seal is formed on a plane cylindrical surface, and in (b) having a vertical face following discharge from the construction.
Cl
Clu
Clu
Cl
(a)
(b)
Fig. 10.7.6—A central diverting ‘rib’ in (b) helps destroy the
kinetic energy of the steam.
Clu
p
(a)
Cr
p
Cll
(b)
Fig. 10.7.7—Two seal strips forming a chamber between them. In (a)
the seals are mounted in the same component, producing a clearance ‘Cr’. In (b) the seals are mounted in alternate components forming an upper clearance ‘Clu’ and a lower clearance ‘Cll’.
609
Turbine Steam Path Troubleshooting and Repair—Volume Two
Fig. 10.7.8—Discharge coefficients for a single seal in terms of the seal
geometry.
Relative
Leakage Flow
Sealing
Design
Cl
1.00
Cl
0.93
Cl
Cl
Cl
0.92
0.88
0.86
0.73
Cl
0.62
Single
Strip
Small
Groove
Medium
Groove
Large
Groove
Coned
Groove
HoneyComb
Double
Strip
Labyrinth
0.40
0.75
Spaced
Labyrinth
Fig. 10.7.9—Various seal configurations and
their flow coefficients.
610
Seals, Glands, and Sealing Systems
Figure 10.7.8 shows a base discharge coefficient “ψ” for a straight
through seal. These values were established for a single seal, and can
be used for different values of the ratio of pressure drop across the
seal. This curve makes some attempt to factor in the geometry of the
seal. Figure 10.7.9 shows a variety of seal configurations, and relative
flow corrections as affected by the seal geometry relative to the
straight through type. These are “shape factors” and will influence the
flow coefficient. Figure 10.7.10 shows a series of curves representing
the flow rates, as a percentage of the straight through type, for different configurations of the strips, and general stage geometry. Similar
information has been presented for tip seal configurations, and the
form effect on flow coefficient. These various forms and corresponding flow coefficients are shown in Figure 10.7.11.
Fig. 10.7.10—Flow factors for various arrangements of the seal strip.
611
Turbine Steam Path Troubleshooting and Repair—Volume Two
Flow
coefficient
0.58
0.46
0.46
0.38
0.30
Fig. 10.7.11—Flow coefficients for various
seal strip configurations at the blade tip.
Multi strip seal configurations
In addition to the importance of the individual strip form, the
arrangement of the seals in a multi strip configuration must also be
considered, in terms of their spatial position relative to each other.
Some possible arrangements will now be considered. In each
design, the arrangement is influenced by certain dimensional and
operational factors, and the designer makes the ultimate selection to
provide the most effective seal possible at the axial location being
considered. The most common configurations are:
612
Seals, Glands, and Sealing Systems
The straight through design. The straight through design is the
simplest multi strip arrangement, and is shown in Figure 10.7.12(a)
and (b). In (a) the strips are mounted to the stationary component of
the unit, and in (b) to the rotating component. In each design a series
of seal strips, each having a radial clearance “Cl” are arranged along
the leakage path, pitched at “P” apart. The distance “P” is selected
to provide an adequate chamber between the strips, and enough to
destroy the steam velocity at entrance.
The seal diameter is marginally different from one arrangement
to the other, being a function of the seal strip height “h.” This difference is minimal when determining leakage quantities.
P
Stationary
Cl
h
(a)
Rotating
Ds
P
Stationary
(b)
h
Cl
Ds
Rotating
Fig. 10.7.12—The straight through design, with the seal strips
mounted in the stationary portion of the unit in (a), and in the
rotating portion in (b).
613
Turbine Steam Path Troubleshooting and Repair—Volume Two
The seal arrangement is convenient when the differential expansion between the stationary and rotating components is such that
clearances at different diameters could not be accommodated into
the arrangement.
The alternative stationary/rotating configuration. In this design,
the caulked attached strips are located alternately in the stationary
and rotating components of the unit, as shown in Figure 10.7.13.
The strips are pitched at a distance “P” apart, and are located from
stationary to rotating component, a distance “Q,” which is selected
so that during operation there will be no contact between them.
There is also a reference setting from some point, such as the thrust
block or coupling flange, such as “G.” This will assure axial positions are predictable under all operating conditions.
It is normal for the upper clearance “Clu” and the lower “Cll” to
be of the same value, but the seal pitch does vary. When calculating
leakage quantities, the mean of “Dsu” and “Dl” should be used.
This is an effective seal, and as seen in Figure 10.7.9, does produce an effective barrier to leakage. This figure illustrates that the last
two arrangements of the pitching of the strips has a considerable
effect on sealing efficiency.
Cu
Stationary
Cl
Dsu
Rotating
Dsl
Q
P
G
Fig. 10.7.13—The alternate stationary and rotating locations of
the seal strips produces a tortuous leakage path.
614
Seals, Glands, and Sealing Systems
The simple hi-lo staggered configuration. The “hi-lo” configuration can be used where differential expansion is relatively small, and
the designer wishes to avoid the need to locate seal strips in the
rotating components of the unit. This arrangement is shown in Figure 10.7.14, where the clearances on the rotor body “Cll,” and the
castellation “Clu” are the same, and there is a minimal difference in
the seal diameters “Dsu” and “Dsl.” The requirement of cold setting
“Q” and “G” are the same at the previous arrangement.
In designing this arrangement, it is necessary to ensure the
castellation width “W” is sufficient, such that under all operating
conditions the seal strip is located above the seal platform.
The multi high strip staggered configuration. In those locations
where differential expansion will not permit the use of the form of
seals shown in Figure 10.7.14, the form shown in Figure 10.7.15 can
be used. In this design there are two high strips. Dependent upon the
extent of load, and temperature distribution throughout the steam
path, only one of the two high strips is effective.
The high strips are pitched in a way that at all rotor axial positions one strip is active, and the other has a radial clearance equal
to “Cu + s,” which is too high to be considered effective in limiting
P
Stationary
Cl
Rotating
Dsl
W
G
Q
Fig. 10.7.14—The simple ‘Hi-Lo’ staggered labyrinth arrangement used on a castellated rotor location.
615
Turbine Steam Path Troubleshooting and Repair—Volume Two
steam leakage. Therefore, when determining the value of the number of strips “N” in Martin’s equation, only one of each pair should
be counted as limiting leakage.
Stationary
Clu
Cll
s
Rotating
Dsl
W
Q
Dsu
G
Fig. 10.7.15—The ‘hi-lo’ arrangement for axial locations
with a large differential expansion. Here two ‘hi’ seals
at each location help ensure one seal is effective at all
axial positions of the rotor.
A shaft end packing arrangment is shown in figure 10.7.16,
which shows a shaft end packing arrangement. This shows the strips
that are located in a gland ring, and the horizontal joint securing
screws that are required to prevent the gland rings from tangential
migration during operation, which would prevent the disassembly of
the top half of the gland housing.
The multi high strip configuration. Shown in Figure 10.7.17 is
the seal system employing what is known as “herringbone seal
strips,” which are used by some manufacturers. In this type of seal,
the seal strip can be formed as a continuous helix, with seal strips
machined in such a manner (on both the stationary and rotating
components) that a small positive radial clearance “Cl” is formed
between them [see Fig. 10.7.17(a)]. The actual leakage area is in fact
controlled by the total clearance shown as “Ct” in Figure (a). However, this a very effective seal due to the geometry, and the fact that
there is (during operation) relative movement between the stationary
and rotating components.
616
Seals, Glands, and Sealing Systems
Fig. 10.7.16—Gland rings at a shaft end sealing position.
Stationary
Rotating
Cl = + ve
(a)
Ct
Stationary
Rotating
Cl = 0
(b)
(c)
Cl = - ve
Stationary
Rotating
Fig. 10.7.17—The ‘herringbone seals’.
This design can be used with (a) a positive clearance, with (b) zero clearance,
and (c) a negative clearance, where
there is no large differential expansion.
617
Turbine Steam Path Troubleshooting and Repair—Volume Two
The major disadvantage to this seal is that in the event of rotor
vibration, the seals will rub and cause damage that is difficult or
impossible to repair.
The effective clearance at any point “Ct” can be seen from this
Figure (a) to be considerably larger than “Cl.” However, it is difficult
to determine the actual value of “Ct” at any point because of the relative motion between the parts. If the seal strips are cut to a helical
path there is no one value to “Ct,” and the strips’ clearances vary due
to their relative and changing pitch position around the circumference.
This type of seal has been found to be very effective, and having
a tendency to reverse the flow direction, destroys the kinetic energy
of the steam. Depending upon the amount of differential expansion
in the vicinity of the seals, they may be arranged with a zero (or negative) radial clearance “Cl,” as shown in Figure 10.7.17 (b) and (c).
In such a design the seal strips must be rings; the helical form cannot be used in (c). These herringbone seal strips are also used on
low-pressure segments where there is large differential expansion.
FORM OF THE GLAND RINGS
Many of the seals formed in the unit are produced as rings,
which can be inserted into the stationary components of the unit,
and then located around the rotating components. These rings can
be accurately located in both the axial and radial directions. This
represents a convenient arrangement, as it allows rings to be
replaced with relative ease when the unit is open for maintenance
inspection. These rings can be classified into two basic types. First
are those in which the seal strips are produced integral with the ring,
in which design the strips are generally of the form shown in Figure
618
Seals, Glands, and Sealing Systems
10.7.4. The second type are those in which the seal strips are inserted, and then staked into the ring. The latter type can, in certain circumstances, have the strips removed and replaced.
The rings must be split at their horizontal joint to facilitate assembly and disassembly, and to allow the segments to make a radial
adjustment during operation. It is normal for the rings to be produced
so a complete ring is formed by 4, 6, or 8 equal arc segments. This
allows the segments to be capable of individual radial alignment.
Depending upon the local environmental temperature, these rings
can be produced from either chrome steel, or a copper base alloy.
Carbon rings
The type of seal used by earlier design units, of smaller output,
was a carbon ring. Such a ring arrangement is shown in Figure
10.8.1. In this design, the seal is produced by a three-segment carbon ring. This ring is arranged and broken at the horizontal joint on
one side of the rotor. Also, at this same location it is keyed to the
gland housing to prevent operational rotation. The ring is held in
intimate contact with the rotor with a garter spring. These carbon
segments normally have closed-butt joint segments.
By providing a large axial taper on the outer diameter of the
carbon ring, as shown in Figure 10.8.1, the spring is also able to
exert enough axial force on the ring to maintain contact between
the ring and housing, and so minimize end leakage over the assembly. The bore of the carbon ring is normally sized to provide a “cold
clearance,” which is taken up as the carbon ring reaches operating
temperature.
This type of carbon gland (on earlier design) had a small butt
clearance “g” between the ring elements, as shown in Figure 10.8.2.
However, these were suitable only for low-pressure wet steam applications, and were used at low rubbing speeds.
619
Turbine Steam Path Troubleshooting and Repair—Volume Two
Load applied by garter
spring both radially
inward and axially
to produce a steam seal.
Atmospheric
leak off
Carbon
Rings
Turbine
Shaft
360° Garter
Spring
Drain
Fig. 10.8.1—The carbon gland ring system with ‘garter springs’.
g
8 - 45°
segments
g
Stationary
Securing
g/2 Screw
Gland
Ring
Fig. 10.8.2—The gland ring, showing the
tangential gaps between the segments.
Butt clearances and tangential location
As gland rings are mounted into the stationary components of the
unit, they are adjusted to have sufficient tangential space (butt clearance) between the individual segments. This allows them the ability
to move radially to accommodate movement of the rotor, and not
bind, or become “arch bound” with changing steam temperatures
when they heat at different rates compared to the major components
620
Seals, Glands, and Sealing Systems
of the unit. Therefore, it is necessary for them to have carefully selected tangential clearances between the segments, to allow for conditions of temperature transients, both at “start-up,” “shutdown,” and at
any prolonged temperature change during operation. It is also necessary to ensure this butt gap should not be so large that excessive
bypass leakage will occur.
The butt clearance is normally specified by the design engineer,
and must be achieved as the rings are assembled to the stationary
components of the unit. The butt clearance is normally a “field
adjustment.”
The individual gland rings are also restrained from tangential
migration during operation. Securing devices at the horizontal joint
provides tangential locking.
Gland ring spring loading
During operation the gland ring is held in a radially inward position by the steam force developed behind it. This steam pressure also
forces the gland ring axially downstream enough to give it positive
location, and to form a steam seal. At “start-up,” and before the
steam pressure develops in the steam path, it is necessary to ensure
the gland ring moves to a radially inward position so the steam seal
can be formed between the stationary carrier and the ring segments.
To achieve this initial seal, it is normal to employ a spring behind
the gland segments, to initiate the seal, and maintain alignment until
the steam pressure can become effective. There are various forms of
springs, from leaf to close coiled. The selection for any application
depends upon the preference and experience of the designer, and
the ease of achieving the initial assembly.
621
Turbine Steam Path Troubleshooting and Repair—Volume Two
The gland ring and carrier geometry
Figure 10.8.3 shows the basic geometry of a gland ring, and the
receiving slot into which it is mounted. In this case there is a closecoiled helical spring, but other spring systems are also used with
equal effect.
W
w1
w2
Co
G
l
H
Ci
Cl
Dr
Dt
S
H
Ci
Fig. 10.8.3—The basic geometry and principle
dimensions of the “T’ head for the gland ring.
In this design the gland ring “T” head has a total width “W,” and
the “tee” slot has a receptacle head width of “w1+W+w2,” which is
sufficient to permit assembly without interference. This provides for
clearances of “w2” on the high, and “w1” on the low-pressure side.
In fact, the pressure in both slot clearances is identical, and equal to
the steam admission pressure to the “T” slot.
622
Seals, Glands, and Sealing Systems
In the radial direction, the seal strips of the gland ring achieve a
radial clearance “Cl” between a rotor diameter “Dr,” and the seal
strip inner diameter “Dt.” In the event the rotor causes a rub, and the
gland segments move radially outwards, there must be sufficient
clearance between the ring surface and the inner surface of the carrier “Di,” so that contact will not occur. This requires that radical
clearance “Ci” is sufficient. There is a further radial clearance
between the carrier ring inner surface and the “T” head of the gland
ring equal to “Co.” This is of no significance except in establishing
the required length or geometry of the springs.
To ensure steam is admitted to the space behind the gland segments, a slot “G” is provided in the segments, on the high-pressure
side to allow easy admission.
The steam seal
To achieve a suitable seal, it is necessary to produce seal faces
(both axial and tangentially) between the gland segments and housing faces. Figure 10.8.3 shows in detail the downstream portion,
where the gland ring mates with the housing surfaces. To achieve an
acceptable seal, it is necessary that axial distance “S” and radial distance “H” has a surface that is at least 125√µ-inches, and preferably
64õ-inches. However, the difficulty of producing these surfaces on
the gland housing must be recognized. On the gland rings this
becomes relatively easier, and for normal machining methods it
presents no difficulties.
When replacing gland rings in a unit that has been in service, it
is common for there to be oxide scale and other impurities present
in the regions where seals are formed. It is a good practice to clean
the surfaces intended to form seals. Since it is not possible to blast
these surfaces with any expectation of success, a removed gland ring
with grinding paste on the seal surfaces can be used to remove a significant portion of any debris and scale.
623
Turbine Steam Path Troubleshooting and Repair—Volume Two
The seal and radial/axial steam forces
The steam forces developed on the gland segments, and holding
them in a radial inward location, need to be considered, because if
the seal fails, excessive leakage can influence these total forces,
causing the seal to become partially ineffective. Figure 10.8.4 shows
a single gland with eight seal strips, each reducing the pressure from
inlet conditions “Pa” to discharge “Pj.” In this gland there are forces
developed in the radial inwards and outward directions, as well as
forces acting both up and downstream. Curves of these forces are
also shown.
Figure 10.8.4 (a) shows the principal dimensions around a gland
segment. In (b), (c), (d), and (e) the principal forces developed by the
steam on this ring are active during normal operation.
The steam radial inward force “Fi” is due to the steam pressure
“Pa” acting on the outer surface of the inlet shoulder, the “T” head,
and the exhaust pressure “Pj” acting on the discharge shoulder.
These steam forces are shown in (b), and are equal to:
Radial Inward Force “Fi” = Pa.[d.Db.(Kt - Ka)] + Pa.[d.Di.(ki - Kn)] +
Pj.[D.Di . (Km - Ko)] +/- Segment Weight
In addition, the radial inward force is supplemented by the force
developed by the spring, which is relatively small compared to the
steam force when the steam pressure is at design conditions. This
inward force is also affected by the weight of the segment, which can
be added for the top section subtracted for the bottom.
In operation, these radial inward forces are opposed by the radial outward forces “Fo” produced by the steam in the individual
pockets formed between the “N” seal strips. These total steam forces
are shown in (c), and are equal to:
624
Seals, Glands, and Sealing Systems
Pa
Downwards
Pressure
(b)
Pj
Axial Length
Ki
Kt
Ka
Pa
Ko
High
pressure
side
(a)
Db
Dh
Low
pressure
side
Km
Kn
Di
Pj Ph Pg Pf Pe Pd Pc Pb
Pressure
Pressure
Pj
(e)
Pj
Pa
(d)
Dt
Pa
Stage Pressure
Drop = Pa - Pj
Upwards
Pressure
(c)
Pj
Axial Length
Fig. 10.8.4—The steam pressure forces on the gland ring. In (b) is the radial
inward steam force, and in (c) is the radial inward direction. These forces are
directly opposite. In (d) is the downstream steam force, and opposing these are
the axial upstream forces, shown in (e).
Note: In determining the “inward “ and “outward” radial forces,
it must be recognized that the projected inner surface of the gland
segment is smaller than the outer because this surface is formed at a
smaller radius. This may need to be taken into account on some
lower pressure stages where the weight of the segment represents a
larger proportion of the total forces.
625
Turbine Steam Path Troubleshooting and Repair—Volume Two
In the axial direction, the steam forces in the downstream (direction of steam flow) direction. “Fd” is determined by the product of
the steam pressure at inlet “Pa,” and the total area of the gland ring
segments exposed to the steam. This total force is shown in (d), and
is equal to:
In the upstream direction, the steam force “Fu” is developed on
the upper face of the “T” head by the pressure “Pa,” and on the
lower shoulder vertical face by the steam pressure “Pj.” These forces
are shown in (e), and are equal to:
In the event there is steam leakage past the seal surfaces “S” and
“H,” shown in Figure 10.8.3, then these values will modify, and
while the pressure differential will be upset, it is unlikely the glands
will fail to function effectively. However, there will be an additional
leakage loss around the gland rings.
In addition to consideration of the radial and axial forces produced on the gland segment by the steam pressure, it is necessary in
certain applications to consider the turning moments these forces
produce. Figure 10.8.5 shows these total forces, and it can be seen
that there is a resultant moment, which would cause the gland ring
to rotate about the point of contact “T.” Because of the curvature of
the segments, the rotation of the segment about “T” would not be too
great before the radial clearance “s” was consumed, but this would
provide an additional leakage path past the seal surfaces, and if
water were present could lead to “washing erosion.” This would also
open the clearances between the seal strips and the rotor.
626
Seals, Glands, and Sealing Systems
Inward
force
s
T
Axial
force
Pj Ph Pg Pf Pe Pd Pc Pb Pa
Weight
Outward
force
Axial
force
Spring
force
Fig. 10.8.5—The forces on the gland segment tending to
rotate it about the turning point ‘T’.
FORMS OF THE SEAL STRIP
AND ITS TRIMMING
The seal strip is formed in cross-section to achieve two concurrent objectives:
•
It must have sufficient thickness or depth “d” of Figure
10.9.1, to be able to withstand the stresses induced in it due
to the pressure drop “dp,” which occurs across it
•
The seal strip will preferably have a knife edge or very thin
edge formed on it, shown as “x” in Figure 10.9.1. This edge
must be able to be machined onto the strip after its assembly
without deforming the strip, or changing its axial or radial
position. This knife-edge is to ensure that should rubbing
627
Turbine Steam Path Troubleshooting and Repair—Volume Two
occur during operation, the strip will wear without causing
significant damage either through mechanical deformation,
or gouging of the rubbed components, or the generation of
significant amounts of heat
Seal
strip
Stationary
Carrier
Caulking
Pressure
drop 'dp'
Thickness 'd'
Steam flow
direction
x
Fig. 10.9.1—The basic dimensions of the
seal strip.
When seals are inserted using caulking material to fasten them to the
carrier, there are several considerations as to how they are arranged:
628
•
The insertion so that the seal strip projection is firm against
the downstream side of the carrier material as shown in Figure 10.9.1. This provides greater strength to the strip, which
could be of excessive length if the caulking material were to
form the base
•
The inserted knife-edges (shown in various figures) are
arranged to be at 90 degrees to the seal surface
•
The taper face, if only one side is tapered, should be on the
lower pressure side of the strip. This will produce a minimally smaller flow coefficient
Seals, Glands, and Sealing Systems
From manufacturing, assembly, and maintenance considerations, there are two types of seal strip that are used. There are those
that are finished in the inserted condition, i.e., no further machining
operations are required after assembly. Also, there are those requiring trimming after installation. When a strip that requires final trimming has been installed, particularly those at a large diameter, it is
necessary to undertake the trimming operation using special tooling.
This may include the use of a boring bar, or in the case of a rotor,
placing the element in a lathe. Because seals can be either axial or
radial, there are different methods used to form the final finished
dimensions. The most appropriate method in any situation is a maintenance decision.
h
Casing
x
Cr
s
L
Dc
Ds
Ca
Fig. 10.9.2—A combined axial and radial seal above a rotating blade
coverband.
When the stage design contains a combined axial and radial seal
of the type shown in Figure 10.9.2, the seal platform with an outer
surface “s” has a radial depth that is required to provide stiffness to
the coverband against bending forces. Therefore, the radial trimming
of the rotor is undertaken before the seal strips are trimmed. However, these seal strips have a pressure drop across them, which if the
629
Turbine Steam Path Troubleshooting and Repair—Volume Two
design requirements of “Dc” and “Ds” are not maintained within
design specification, can make the length of the seal strip “h-Cr” to
be excessive, causing an increase in its bending stress. Such an
increase is normally not likely to exceed the strip capabilities, but
this should be considered in any trim machining.
INSERTION AND SECURING OF
SEAL STRIPS
There are various methods of attaching seal strips to stationary and
rotating components of the unit. There is obviously a need to make a
more secure attachment when the major carrying or locating component is rotating. This is because such strips will then be subjected to
the effects of centrifugal load due to their own mass during operation.
However, because of their design function there is a pressure drop
across the seals, and therefore a bending stress induced in them.
Possibly the simplest forms of attachment are those shown in Figures 10.10.1 and 10.10.2. Figure 10.10.1 shows a solid form of strip, in
which there are small grooves, at a depth “L” produced on one or both
sides of the strip. This strip, when inserted into a prepared groove in the
major component, can be staked as shown. In the event these strips
require changing, it is sometimes difficult to re-stake, because the major
component has already been deformed by the initial staking action, and
there could be insufficient material to reattach the strip securely.
The caulked strip of Figure 10.10.2 is secured by inserting a strip,
as shown, and then using a malleable material to form a tight enclosure, holding the strip securely in place. This type of attachment can
be used successfully on both stationary and rotating components of
the unit, and provides the capability of having the strips easily
removed and changed. The prepared grooves are normally arranged
630
Seals, Glands, and Sealing Systems
L
Staked
x
Fig. 10.10.1—The staked seal strip.
Stationary
carrier
Soft
caulking
material
d
L
α
x
Fig. 10.10.2—Details of the caulked seal
strip.
631
Turbine Steam Path Troubleshooting and Repair—Volume Two
to have sloped or tapered walls at an angle “α” as shown. These
strips can be trimmed to diameter “Ds,” plus clearance and knifeedged to “x” after installation. However, it often becomes necessary
to undertake some straightening of the strips after caulking. This
straightening is normally undertaken by hand; if rolls are available
they are preferred. The ultimate success of the caulked strip depends
upon the selection of a caulking material that can operate satisfactorily at the local environmental temperature of the stage. There are
suitable materials available, and these make this means of securing
the strip relatively easy, and secure.
Locating
grooves
P
Dt
Fig. 10.10.3—The inserted and replaceable seal.
Figure 10.10.3 shows a more sophisticated means of attaching a
simple single, or multiple seal to a stationary component that is
accessible from the horizontal joint. In this case, the seals are produced from solid material, and have the strips machined into them
with form cutters. The location within the major component is
achieved by means of a simple dovetail, as shown. The actual geometry of the strip is dependent upon manufacturer preferences, and
can have considerable variation, and yet remain a successful type
component. It is normally necessary to secure such strips against tan-
632
Seals, Glands, and Sealing Systems
gential migration within its locating slot. These seals are normally fitted as segments that are of a length that allows their easy assembly.
A typical seal system for a Curtis stage is shown in Figure 10.10.4,
where a total of 11 inserted strips, of the type shown in Figure
10.10.2, are used to provide effective sealing between the nozzle
plate, the stationary carrier, and two rotating blade rows. These seals
are used for both axial and radial flow control, and are intended both
to minimize leakage, and to guide the steam from one row to the
next. Each of the 11 seals is located in a stationary component of the
stage, and therefore is not subject to centrifugal loading.
Ca2
Ca1
Cr1
Cr3
Cr2
Cr5
Cr4
Ca4
Ca3
Fig. 10.10.4—Details of the seal system around a ‘Curtis’ stage. The total system
contains 11 strips, providing individual seals between the stationary and
rotating blade rows.
633
Turbine Steam Path Troubleshooting and Repair—Volume Two
When multiple series seals are required, the type shown in Figure 10.10.5 can be used. Here a number of individual gland rings
are employed to provide a tortuous leakage path. A similar arrangement is shown in Figure 10.7.16 with alternate “hi-lo” teeth which
can also be used. The inverted “T” type root that located the segment in the stationary component is spring loaded, and prevented
from tangential migration by use of a securing screw at the horizontal joint.
Fig. 10.10.5—Inserted gland rings seen from the horizontal joint. Tangential migration is
prevented by means of a button screw at the horizontal joint.
Gland and seal strip assembly
There are several aspects of the assembly of seal strips and gland
segments that need to be considered to help ensure the seals are as
effective as possible, and to minimize leakage quantity.
634
Seals, Glands, and Sealing Systems
Staked strips. This strip is inserted either in short segments, or as
complete 180-degree strips. The length of strip depends upon the
overall geometry. These are then held in position by staking. The
staked strip, shown in Figure 10.10.6, shows the principal dimensions that should be controlled. The location within the major component is set by the depth of the groove “G” and its width “w.” These
dimensions must be controlled to ensure the strip has the two staking slots located at the correct radial height, and the strip is not loose
in the groove. The control dimension of the groove “V,” which is set
from a referenced surface, sets the axial position of the seal.
w
90°
L
X
Dt
Fig. 10.10.6—Dimensional
details of the staked seal strip.
After installation the strip is checked to ensure it is at 90 degrees
to the plane of the surface. The strip is possibly trimmed to a length
“L” to achieve the design seal diameter “Dt.” The knife-edge thickness is “x.” (There may need to be some small degree of flexibility in
this value, and the shape of the strip required to achieve an acceptable compromise.) Many manufacturers do not trim such shaped
strips, as they are able to control the manufacturing process of both
635
Turbine Steam Path Troubleshooting and Repair—Volume Two
the major component, and the strip to the extent the design diameter “Dt” is achieved within design tolerances.
This is a replaceable strip. However, after a number of replacements it would be advisable to re-machine the casing groove to a new
width “w1,” and use a new form (width) of strip. This would provide
new caulking material and new groove surfaces to locate the strip.
Caulked strips. The row of caulked strips shown in Figure
10.10.7 are arranged to be at a pitch “P” apart, and after assembly
the strips are adjusted to achieve the 90 degree projection from the
major component. After assembly, these strips are trimmed to a
length “L” to achieve the design seal diameter “Dt.” The strips are
machined to an angle “α” at their tip, producing a knife-edge thickness “x.” The requirements for caulking are similar to those for the
staked strips. The advantage of these strips is that it is normally possible to easily replace rubbed or damaged elements because the
assembly process does not affect the major component. Like the
staked seals, the axial, or radial position is set by the control position
of the groove “V.”
α
90°
L
Dt
P
V
x
Fig. 10.10.7—Dimensional details of the caulked seal strip.
However, as a unit ages there is always a tendency for the
integrity of the groove to deteriorate, particularly if the seals have
been changed several times. Also, the general condition deteriorates
636
Seals, Glands, and Sealing Systems
if there is water at the seal location. Under such conditions it is
sometimes necessary to reform the groves with slightly larger dimensions, ensuring the same axial distance “V” is maintained.
Inserted segments. The segments that are inserted into specially
prepared dovetail locations have (in addition to the requirements of
being correct dimensionally) the need to be assembled correctly so
they can achieve both an acceptable seal, and so they can be
removed for replacement as they become worn, and/or in certain
designs to have inserted strips replaced.
Irrespective of the type of tooth or seal strip (inserted or integral),
there are two basic assemblies of these segmented glands. There are
those designed to have a “spring” backing to them so they are able
to move radially towards and away from their mating part as it
moves during transient conditions. This “spring back off” is intended to minimize the wear that occurs on the segment as the seal rubs
on the mating part. There are also solid segments, which have no
spring action, and wear in the same manner as the inserted strip if a
rub should occur.
Figure 10.10.7 shows the cross section of a segmented gland with
four straight through strips. This design has no provision for backing
away under spring loading. Therefore, if a rub occurs the strips will
be worn. The principal dimensions are shown. These are similar to
the inserted strips, having a length “L,” being pitched “P” apart, and
having a knife-edge thickness “x.” The seal diameter is “Dt.”
Figure 10.10.8(a) shows a gland ring having four strips, but this
gland section is designed for spring backing. The strips have a seal
diameter “Dt.” Figure 10.10.8(b) shows a similar segment section,
but in this design there are four “hi-lo” strips. Therefore, this section
has two seal diameters “Dto” and “Dti,” with seal strip lengths “L1”
and “L2.” With this (b) design, because of the possible geometry
changes with the different seal diameters, and a constant pitching, it
is possible there will be differences in the strip thickness “x” or the
637
Turbine Steam Path Troubleshooting and Repair—Volume Two
angle of the strips. This is a design detail, and may vary from one
manufacturer to another.
The one factor that is different on these two segments is the distance “E” from the center of the shoulder to the steam face. This
dimension “E” is made purposely different on the “hi-lo” design to
prevent segments from being installed incorrectly. If incorrect assembly should occur, it is possible the high-low strips would be in an
incorrect axial position, or in the case of straight through strips, there
could be pitching difficulties.
E
E
E1
E
L
L1
Dt
Dto
L2
X2
X
P
(a)
P
X1
Dti
(b)
Fig. 10.10.8—Details of the inserted gland rings with integral seal teeth. In (a)
is shown the straight through form, used when the axial location has large
differential expansion.
In addition to these general requirements, for the spring loaded
segments there are other requirements necessary to ensure they
operate as intended by design:
•
638
The production of a suitable steam seal face on the downstream side. Such a seal is required to minimize leakage
around the segment. This steam seal is produced principally
on the axial face, as shown in Figure 10.8.3 as surface “H.”
There is also a sealing effect on the axial surface “S”
Seals, Glands, and Sealing Systems
Some designs ensure the steam seal face is always in relatively
hard contact, by providing close-coiled, helical spring, as shown in
Figure 10.10.9. These springs provide an axial force sufficient to produce contact at all times.
Close coiled
helical spring
Fig. 10.10.9—A close coiled helical spring applying an axial load to close the steam seal face.
This steam seal face is particularly important in stages with highpressure drops across the segment, and free moisture, as occurs in
the high-pressure section of nuclear units. Under such circumstances, washing erosion could occur if leakage became excessive.
It is important that when gland segments are changed, before installation of the new segments, all damage and dents are removed from
the seal face within the diaphragm or packing head grooves.
•
Positive radially inward pressure is required to maintain axial
alignment. This is done initially by a spring located between
the housing and gland segment, and then by steam pressure
during operation
639
Turbine Steam Path Troubleshooting and Repair—Volume Two
For normal metallic seals, some form of leaf or close-coiled
design is used. For carbon rings, a garter spring extending 360
degrees around the gland ring is used.
•
Because the segment will need to move radially during operation, there is a need for a small tangential “butt” gap
between the segments. This gap represents a leakage path to
the steam. However, without the gap, the segments could
“bind” at some loads and steam conditions, holding the segments off their shoulders and preventing effective sealing.
Also, the segments and carriers are manufactured from materials that can have different coefficients of expansion and
will accept and reject heat at different rates. Therefore, the
“butt” gap is required to prevent “binding”
These gaps must be checked at final assembly. This is most
important when a bronze material is used for the seals, because this
material has a coefficient of expansion considerably different from
steel, and failure to maintain the gap at design values could introduce problems during operation.
•
At “shutdown” there can be moisture collecting at various
parts within the steam path. Often a provision is made at the
bottom dead center of the gland housing for collected water
to drain through the seal to some convenient collecting point
within the turbine steam path, as shown in Figure 10.10.10.
This seal has been in service for an extended period and also
shows evidence of rubs
The details of the components and seal system will vary from
design to design. These variations represent a preference and the
experience of the designer. When any maintenance work is undertaken, it is necessary to return the measured characteristics as close
as possible to the original design, provide positive location, and a
steam seal surface to prevent excess leakage.
640
Seals, Glands, and Sealing Systems
Fig. 10.10.10—A caulked strip at a rotating blade row, showing the drainage slot and a
‘rub’ at the bottom dead center.
SEAL STRIP AND GLAND
RING MATERIALS
Two quite distinct classes of material are used for the manufacture of gland segments and seal strips. The material selected for use
in any particular application is chosen to be of sufficient strength and
mechanical properties, such that it is able to perform adequately
within the temperature of the local steam environment, and where
necessary, be able to limit the damage that can be caused by moisture deposited from the steam.
For the lower temperature, low-pressure applications, a nickel
leaded copper alloy is normally used. While this material is normally
produced as a sand casting, centrifugally cast elements are preferred.
641
Turbine Steam Path Troubleshooting and Repair—Volume Two
These centrifugally cast rings tend to minimize the porosity, which is
common in this material. Should porosity occur in the region of the
thin strip, it can cause a component to be scrapped after a considerable amount of work and time has been invested into its production.
In the high-temperature zones, steel is preferred, as it is better able
to withstand the elevated temperatures and pressures to which these
components are subjected. The transition from steel to the leaded
bronze material usually occurs at temperatures below about 750ºF.
Sand-cast materials
A typical chemical composition for three copper base alloys is
shown in Table 10.11.1. The first two alloys are suitable for sand
casting, while the third is the normal composition for the centrifugally cast material. While gland ring segments are not exposed to
any considerable stress levels, the basic mechanical properties are
shown in Table 10.11.2 for reference.
ASTM
B584-73
Copper
Tin
Lead
Zinc
Antimony
Nickel
Sulphur
Phosphorus
Aluminum
Manganese
Silicon
%
%
%
%
%
%
%
%
%
%
%
949
Cast
976
Cast
Centrifugal
Cast
79.0 - 81.0
4.0 - 6.0
4.0 - 6.0
0.30
0.25
4.0 - 6.0
0.80
0.05
0.005
0.10
0.005
63.0 - 67.0
3.5 - 4.5
3.5 - 5.0
1.50
0.25
19.0 - 21.5
0.80
0.08
0.005
1.00
0.15
61.0 - 68.0
1.3 - 3.0
4.5 - 7.0
0.85
10.0 - 16.0
0.35 Max
0.35 “
0.35 “
0.35 “
Table 10.11.1—Copper Base Alloys for Gland Segments and Seal Strips.
642
Seals, Glands, and Sealing Systems
Tensile Stress psi
Yield Stress
psi
Elongation
%
38,000
15,000
15
40,000
17,000
10
34,500
13,500
30
Table 10.11.2—Mechanical Properties of Copper Base Alloys.
Alloy steel materials
The alloy steel chosen by the manufacturer for the higher temperature regions is again chosen to suit particular applications, such
as nuclear and non-nuclear applications, and influenced by moisture content in the steam, etc. Typical chemical compositions are
shown in Table 10.11.3. This listing is not complete.
Material
Carbon
Manganese
Phosphorus
Sulphur
Silicon
Chromium
Molybdenum
%
%
%
%
%
%
%
1.25CR
2.25Cr
AISI 416 AISI 410
0.12-0.15
0.12-0.15
0.03
0.03
0.03
0.03
1.25
1.00
2.25
1.00
0.15
0.15
1.25Max. 1.0Max
0.06
0.04
0.15
0.03
1.0Max 1.0Max
12.0-4.0 11.5-13.5
0.60Max
Table 10.11.3—Typical Alloy Steels for Gland Segment and Seal Strips.
GLAND SYSTEM
OPERATING PROBLEMS
Problems occur and damage is found at various locations within
the steam path where seals are used. These problems cause damage
to either the seal strip itself, which is the more common, or it can be
on the component against which the seal is formed.
643
Turbine Steam Path Troubleshooting and Repair—Volume Two
The most common fault that occurs on the seal strip is for “rubs”
to occur during operation. These rubs flatten the knife-edge, thus causing both an opening of the running clearance, and also increasing the
discharge coefficient. A rubbed knife-edge is shown in Figure 10.12.1,
where a rub has occurred on the caulked strips in the diaphragm carrier, located above the blade tip. This piece of strip is located at the
“bottom dead center,” and the moisture drainage slot can also be seen.
This seal is located above a blade coverband in an impulse unit. This
rub is not heavy, and a decision to change will be based on an economic evaluation of the additional losses sustained as a consequence
of the increased clearance.
Fig. 10.12.1—A caulked seal showing a ‘rub’ at the knife edge.
Figure 10.12.2(a) shows a gland segment from a diaphragm where
heavy rubs have occurred. In Figure 10.12.2(b) are the measured values of clearance around the total seal. These clearances have been
determined from knowledge of the design clearance, the original seal
strip height “h,” and the measured value of “h” from the removed
gland segments. These measurements, which were taken at three loca644
Seals, Glands, and Sealing Systems
Fig. 10.12.2(a)—A lower gland ring segment, showing rubs at the knife edge.
A diametral record of this complete ring is shown in Figure 10.12.2(b).
tions on each of the six segments, show that the major wear is in the
vertical position. This unit has obviously been either subjected to high
levels of rotor vibration, or was not aligned to its best advantage when
returned to service after the previous outage.
When such rubs occur, there are three possible actions available
to the operator:
•
To operate with the damaged seals and accept the additional
losses that are induced by the increased clearance produced
•
To replace the worn strips with new components, then adjust
them by some means to re-establish the original clearance.
This may involve the expense of removing the seal carrier to
a facility with the capability of machining the seal edge, or
using a boring machine inside the unit to produce the seal
diameter
645
Turbine Steam Path Troubleshooting and Repair—Volume Two
•
To hand dress the worn parts, to re-establish the knife-edge.
It will normally be necessary to return to service with an
increased clearance, but the coefficient of discharge will
have been re-established at or near its original value. This
can normally only be justified when the damage occurs over
a relatively small length of the knife-edge
The choice depends upon various factors, most importantly the
economics of the situation, but also by the ability of the design to
accommodate the installation of new parts, or the dressing of the
existing. Many seal arrangements do not easily lend themselves to
replacement without incurring excessive cost, and one that is greater
than any savings anticipated as a result of fuel cost savings.
Fig. 10.12.2(b)—A diametral record of the radial clearances around a complete
gland ring set. These clearances have been measured at three tangential
positions on each of the six segments. The design clearance was set at 0.015".
646
Seals, Glands, and Sealing Systems
Fig. 10.12.3—A multi knife edge seal showing the effects of
a ‘rub’. These seals are produced integral with the gland
ring carrier.
The forms of some strips make refurbishment difficult, and the
only recourse is replacement. Figure 10.12.3 shows the complex
seal form from an impulse unit, where each pitch had an original
three knife-edges. After rubs, these edges have been destroyed, and
badly deformed. However, these gland rings can be replaced with
relative ease.
After extended operation at high-steam conditions, some seal
strip material tends to become brittle. This is particularly the case if
there have been rubs that would have heated the material, and then
its having been immediately quenched by the flowing steam. Such
quenching will increase the tendency towards embrittlement. Under
these conditions, relatively light impacts from mechanical debris (or
647
Turbine Steam Path Troubleshooting and Repair—Volume Two
even the steam forces) can cause the thin seal metal section to rupture and deteriorate. Figure 10.12.4 shows a seal where a significant
amount of material has been lost, increasing the leakage area. This
shows that there have been rubs on these spring-loaded elements.
Fig. 10.12.4—A seal showing the loss of seal strip material. This seal has been exposed
to high temperatures for extensive periods.
Another form of damage sustained by seals is “knock” damage,
which is sustained as a consequence of small impacts when the unit
is open for inspection or repair. For this reason, operators must take
the greatest care to protect these seals when they are removed from,
or left exposed in the unit during an outage. It is sometimes possible
to straighten bent or deformed seal strips. However, it must be recognized that microcracks may then exist, which could cause rupture
after return to service. The missiles generated by subsequent failure
of the seals are unlikely to cause impact damage, but there will be a
resulting loss in efficiency and output
648
Seals, Glands, and Sealing Systems
Fig. 10.12.5—Rubs on the inner coverband of stationary blades. These rubs are not normally damaging to the coverband, but will deform the knife edge of seal strips. The
casing carrying these stationary blade elements should be checked for concentricity.
The surface against which a seal strip forms its seal can also be
damaged, by various mechanisms. Figure 10.12.5 shows a portion of
a coverband in which small grooves have been cut by rubs that have
occurred during operation. This form of damage is common and can
also occur on the rotor body. Often, such rubs will cause localized
heating and consequently hardening, making the coverband and/or
rotor materials more susceptible to corrosive attack from aggressive
ions that have found access to the steam path.
When seal strips are caulked into a stationary component, they
must be checked to ensure they are secure against the steam forces
that will act on and affect them during operation. Figure 10.12.6
shows a portion of a seal system located in a casing above a blade
coverband. Here the seal strips have worked loose, detached, and
passed down the steam path. Figure 10.12.7 shows a caulked seal
strip from above a rotating blade row. This strip has been damaged
by mechanical impact, but can be replaced relatively easily at a
maintenance outage.
649
Turbine Steam Path Troubleshooting and Repair—Volume Two
Fig. 10.12.6—Seal strips, which have detached from their holder grooves, have entered
the steam path, causing some gouging and damage to the blade elements.
650
Seals, Glands, and Sealing Systems
REFERENCES
1. Martin, H.M. Steam Turbines, published in The Engineer,
London 1913, 1,610
2. Egli, A. The Leakage of Steam Through Labyrinth Seals,
Transactions ASME, Paper FSP-57-5
3. Kearton, W.J. Leakage of Air through Labyrinth Glands of
Staggered Type, Institute of Mechanical Engineers, September, 1950
4. Meyer, C.A., and J.A. Lowrie. The Leakage Thru Straight and
Slant Labyrinth and Honeycomb Seals, ASME Paper 74WA/PTC-2
5. Neuman, K., G. Stannowski, and H. Termuehlen. Thirty Year
Experience with Integrally Shrouded Blades, The Joint Power
Generation Conference, Dallas, Texas, October, 1989
6. Cofer, J.I., S. Koenders, and W.J. Sumner. Advances in Steam
Path Technology, GE Power Generation Paper GER-3713C
651
Chapter
11
Quality Assurance for
Replacement and
Refurbished Steam
Turbine Components
INTRODUCTION
There are various causes that can initiate mechanical failure of
the steam path components (see chapter 1). A major, but often-overlooked contributor, is the quality of the manufacture and/or assembly, and their final compliance with design requirements.
There are many critical components, each with a number of
characteristics that have the potential to affect both efficiency and
reliability within the unit. The quality requirements of these components are established by the designer in terms of the tolerances
applied to the individual components, the processes that will be
used to produce them, and the manner and expertise with which
653
Turbine Steam Path Maintenance and Repair—Volume Two
they are applied and assembled. Many of these technical requirements are not obvious from casual observation, and to ensure the
design specifications are achieved in the delivered product, it is necessary to investigate these during the process of awarding a contract,
and then to monitor their application during manufacture, and also
to have access to any instances of nonconformance.
This availability of access to (and even involvement in) the disposition of nonconforming situations provides considerably greater
confidence in the quality of the final product. Such participation also
allows the purchaser to be able to anticipate, and identify manufacture as the cause of any specific mechanical problems subsequent to
going into service.
The purchaser/user has an implied responsibility to monitor the
manufacture and assembly of these components. This monitoring
does not require the physical measurement of the components themselves, but can normally be achieved by the monitoring of the supplier’s quality program, and also by directing inspection or surveillance attention to those critical characteristics that must be achieved
if the unit is to perform as anticipated.
To undertake a “quality review and monitor” it is necessary to
compare the manufactured components against the standards
defined by the design engineering function. These engineering standards establish component requirements, and the quality of the
products is dependent upon the ability of the manufacturing department to achieve an acceptable level of compliance.
This synopsis is not complete, as the requirements from one
manufacturer to another may be defined by different methods. This
is not unreasonable, and these differences reflect the differences in
both the manufacturers’ method of manufacture and their experience. It is only by seeing that the manufacturer has (and follows)
their own documented standards, that an acceptable component
can, from a manufacturing/refurbishment perspective, be achieved.
654
Quality Assurance for Replacement & Refurbished Steam Turbine Components
The design specified requirements for components of the steam
path are stringent. The engineering specifications include definition
of material properties, physical dimensions, process, and performance requirements. Many of these engineering defined requirements
will have an adverse effect on the total performance of the turbine if
not followed in detail.
Component failures resulting in forced or extended outages are
common within the steam path, and often the cause of such failures
can be attributed to failure on the part of manufacturers to adequately control either the metal forming and manufacturing process,
or the assembly of the individual components into the final product.
An important aspect for any purchaser of new or replacement
parts, is the ability to establish that components are produced in a
way that they will assemble to the turbine, and provide performance
levels that comply with system requirements. This surveillance, or
monitoring of product quality should be an integral part of any procurement program, and could have a direct effect on the quality and
suitability of the parts for their intended application. The manufacturer of new and replacement parts will normally undertake inspection and produce records to verify compliance with design requirements.
It is important to recognize the requirements of a quality program,
and in qualifying a manufacturer determine that their quality program
will help ensure components meet the design specification. In addition to design, manufacture, and assembly, the program can also be
extended to include the requirements that the components are packed
and shipped so they will be in an acceptable condition when mounted to the turbine. In the case of inventory components, the time of
installation may not be for several years, and therefore will require to
be protected to prevent any deterioration during the storage period.
There is, of course, a cost associated with the review and monitoring of a quality program—a real cost. However, if this prevents an
655
Turbine Steam Path Maintenance and Repair—Volume Two
incorrect component being delivered, or the extension of an outage so
that some defect can be corrected, then the costs are easily justified.
Figure 11.1.1 shows the possible involvement of a purchasing
engineer, in reviewing and establishing a quality program at the outset of a project. The most appropriate (but costly) approach would
be to review the potential supplier Quality Assurance (QA) program
prior to placement of the contract, but this can become expensive in
terms of a small contract, and would be difficult to justify. Therefore,
a normal manner of undertaking this review is after placement of the
contract, and at an engineering review, which is held after the supplier engineering function has prepared the engineering specification for the component. At that time, the drawings, material specifications, special process specifications, and any other details (togeth-
Engineering
Definition
Prepare
Engineering
Specification
Review
QA Program
* Manual Review
* Program
Implementation
Engineering
Review
*
*
*
*
*
Original Drawings
Reverse Engineering
Material Specifications
Process Specifications
Special Processes
* Inspection
Records
Inspection and
Test Plan
Product
Surveillance
Quality
Records
Fig. 11.1.1—Purchaser’s involvement in establishing project scope
and QA involvement.
656
Quality Assurance for Replacement & Refurbished Steam Turbine Components
er with the QA program) can be examined for suitability, the contract
being set on the basis of acceptability.
At the review meeting, the inspection and test plan (I&TP) are
reviewed and accepted, or modified, and “hold points” agreed
upon. Based on this I&TP, the purchaser will define and prepare his
or her surveillance activities.
RESPONSIBILITY FOR QUALITY
There are various factors that can contribute to poorer-thananticipated reliability of the turbine steam path, and in certain circumstances these factors will combine to degrade the total availability of the unit. To begin to define the technical requirements of
any component to operate adequately, it must be recognized that the
prime responsibility for the adequacy of the component is vested in
the design function. Engineering must ultimately be considered
responsible for product quality. Steam path components, many of
which are technically complex, are designed and manufactured to
stringent standards. Such standards are intended and expected to
make the turbines suitable for continued, reliable operation.
The stresses induced in, and the environmental operating
demands placed on steam path components are such that their margin of safety can be seriously eroded by relatively small manufacturing deviations from the design specified requirements. Many failures
of components within the steam path can be attributed directly to
incorrect or inadequate control of manufacture, while others are
accelerated or contributed to by unacceptable manufacturing quality.
Turbine steam path components operate in an environment that
tends to degrade their material quality. Also, many components can
be subject to excessive stress levels, both direct and alternating, as a
657
Turbine Steam Path Maintenance and Repair—Volume Two
consequence of relatively minor manufacturing deviations. These
factors combine to reduce the operating life of the components in a
number of ways.
The cost of a turbine component failure is composed of several
elements. There is the cost of purchasing replacement components,
the cost of opening the turbines to install them, and the maintenance
staff costs associated with their installation. However, these costs are
normally minimal when compared to those costs associated with the
purchase of replacement power, whether it is purchased from other
producers, or generated within the utility by starting older, less-efficient units, or starting units consuming more expensive fuel. For
these reasons the purchaser has a direct responsibility to monitor
supplier quality to help ensure the possibility of manufacturing errors
is minimized.
DEFINITION OF QUALITY
The quality of any piece of engineered equipment can be
defined by its degree of compliance in meeting primary and secondary objectives, as defined by design engineering, and by providing the owner an acceptable return on investment. To define the
quality of turbine components, it is necessary to be able to state
quality requirements in quantifiable terms, and then be able to measure these against a standard or target value that provides direct comparison with anticipated values.
In general, the quality of a component should be defined in
terms of its performance against both supplier-predicted levels, and
standards for the industry. Establishing what the standard value
should realistically be for any technically complex, contract-engineered item is a problem of considerable magnitude. In making
658
Quality Assurance for Replacement & Refurbished Steam Turbine Components
comparisons, it is very unfair to attempt a comparison of a component or design that either contains a significant number of newlydeveloped features, or will be operating in an environment or load
pattern in which it is untried, to one containing no prototype components and operating in a proven and accepted manner.
Therefore, it is reasonable to expect the quality or performance
level of any component to be a comparison with a standard that represents a realistic assessment for the design, its degree of design evolution, and mode of operation. As an initial step towards measuring
quality, it becomes necessary to define, and then quantify performance, and relate quality to those factors that define it.
The quality of any turbine component can be established against
standards of performance, and against supplier-predicted levels. It
would be appropriate if this could be done in terms of their influence
on both efficiency and reliability of the unit, and the effect such
components have on these qualities. Optimally, each component
should be considered in isolation, but if comparisons are to be made
among different but comparable components, it is necessary to
reduce the parameter of each piece to a common base to allow
meaningful comparisons. It is difficult to establish the efficiency of a
particular component. It is sufficient to assume that if components
meet their engineering definition (from the original equipment manufacturer, or from another supplier determined by reverse engineering), then it will achieve the efficiency and reliability required to
achieve the heat rate and availability of the turbine-generator unit.
659
Turbine Steam Path Maintenance and Repair—Volume Two
DEFINITIONS OF
PERFORMANCE
In describing performance or quality levels, it is valuable to
establish or define terms that can be used, and have specific meanings. The most appropriate are:
•
Guaranteed level—quoted by a supplier during the bidding
phase. Such levels could form part of a commercial agreement between supplier and purchaser. If tests or measurements are taken, it is normally against these levels that comparisons will be made. Efficiency can be measured. However, availability is difficult to quantify because it is influenced
both by manufacturer and operator determined factors
•
Target (tolerance) level—represents an adjustment made to the
guarantee level, up or down, for experience gained during
previous operation with the supplier’s equipment. Because
these values are normally adjusted by the purchaser (based on
his or her accumulated knowledge of the manufacturer) they
form no part of the bid, and have no legal status. These target
values are, however, of great importance both in bid evaluation, and for operators who must measure their performance
to some established and predictable standard
•
Actual level—represents the field measured and observed
values and can be compared to both guaranteed and target
values. Discrepancies, either positive or negative, should be
accounted for
These first two definitions are often used interchangeably as the
predicted level of performance. Most often, the predicted value
refers to the guaranteed level when applied to unit efficiency, but
more often, the target levels, when applied to unit availability. (The
supplier does not normally guarantee availability.) However, most
660
Quality Assurance for Replacement & Refurbished Steam Turbine Components
suppliers have statistical values available upon which a purchaser
can make predictions of how his or her unit should perform in the
environment of his or her power system.
The factors of performance
In discussing quality or performance, it is necessary to establish
what factors contribute to it and how they can realistically be measured. These factors are shown in Figure 11.2.1. It is necessary to
define performance in a quantifiable manner, establish a means of
comparing it to a base that is common for all components, and
allows comparison in such a manner the anticipated differences
between them can be eliminated.
Steam
Turbine
Performance
Unit
Availability
Unit
Efficiency
Reliability
Initial
Efficiency
Sustained
Efficiency
Safety
Mechanical
Integrity
Maintainability
Repairability
Accessibility
Interchangeability
Correctability
Fig. 11.2.1—The factors of performance.
Performance can be considered as being comprised of two distinct factors—efficiency and availability. Of these two factors, availability comprises two sub factors—reliability, which measures the
661
Turbine Steam Path Maintenance and Repair—Volume Two
unavailability or forced outage rate of the unit as a consequence of
component problems, and maintainability, which measures the ability or time interval required to return a unit to the available status
after a forced outage caused by various components.
The influences of availability. The availability of power-generating equipment or components contained within such equipment is a
measure of its availability to generate power when required. This
definition is independent of the equipment actually being in service;
the requirement is that it is available if required. There are two factors of availability that need to be considered in describing the performance or quality level of any unit or component:
•
Reliability—a measure of the forced outage rate of a unit or
component, and is a measure of the period during which,
due to component failure or damage, the unit is not available
to generate power
•
Maintainability—a measure of the ease of access to the various
component parts of a unit to undertake corrective action, and
the individual components’ ability to be reliably repaired.
Therefore, this is a measure of the suitability of a unit to be
inspected, corrected, and returned to an available status after
a planned or forced outage has required some corrective
action
The influences of efficiency. The efficiency of the thermal cycle,
the prime mover, or any significant component of the prime mover,
can (for a cost) be determined by established means, and with relative ease. This value can also normally be determined with a high
degree of accuracy.
•
662
Initial efficiency—It is normal for the builder of power generating equipment to guarantee efficiency for each unit supplied,
or in the event of supplying replacement components, that they
will perform at the same level as the original components they
Quality Assurance for Replacement & Refurbished Steam Turbine Components
are replacing. This guarantee level is one predicted by the supplier from an analysis of past operating information, or from
development tests undertaken on a new design of components
•
Sustained efficiency—Degradation of efficiency will occur as
many of the components age. This degradation can be due
either to deposits formed on them, or any mechanical damage they sustain as a consequence of operation. Any evaluation of the efficiency of the replaced component must be
undertaken before degradation to a significant level occurs
THE DESIGN SPECIFICATION
The turbine steam path utilizes components selected and arranged
by design to achieve a predicted level of output, and one that can
achieve a level of efficiency consistent with an acceptable level of
service reliability. These two requirements of efficiency and reliability define the “performance level” of the turbine. Unfortunately, in
many instances of component selection, optimizing these two
requirements is difficult because they are often in direct conflict.
Therefore, a major responsibility of design is to evaluate alternative
components and make a balanced selection, i.e., a selection that will
achieve acceptable levels of both efficiency and reliability, and can be
utilized in a turbine that can be manufactured at acceptable costs.
In evaluating and making the selection, the designer considers
the total requirements of the various components, both individually
and as an assembled whole. The final component selection is made
from a detailed analysis of possible arrangements. After a total evaluation, the turbine is defined, and a specification is prepared that
will include information to the manufacturing function of the turbine
663
Turbine Steam Path Maintenance and Repair—Volume Two
supplier. This design definition will include at least a portion of the
following information for each component:
A material specification. The material specification defines the
properties of the material from which the components are to be produced. This material specification should include both chemical and
mechanical requirements, the method of material production, heat
treatment, and the delivered (and possibly the final) microstructure.
A material supplier will supply a material test certificate along
with the provision of material. The turbine manufacturer will examine the material supplier’s certification as the material is received. It
is also possible critical materials will be subjected to microstructure
and mechanical properties verification and chemical composition
analysis before being used to produce components.
The most effective method for a purchaser to ensure materials
meet design requirements is to examine the material specification
and then compare the material test certified properties with the specification for conformance. Material conformance must be complete
in every detail, from physical property and chemical compliance,
and heat treatment, to method of manufacture and preparation.
Physical dimensions. Dimensional requirements are more difficult to define, and these data are provided to manufacturing departments in a variety of forms. The most obvious form is for dimensions
and tolerances to be specified on manufacturing drawings. However, other methods, such as operation sheets and manufacturing standards, are also used. If material is to be left on a component for finishing in the field, the specification should define how much additional stock should be left on the component.
Possibly the most effective method of considering the dimensional requirements of the steam path components, from a purchaser’s perspective, is to review them briefly by component, and pro-
664
Quality Assurance for Replacement & Refurbished Steam Turbine Components
vide details of the controlling dimensions that must be achieved in
the finished product to help ensure a reliable component.
Definition of surface finish. The surface finish can be critical on
certain components, and where necessary is defined. As part of the
engineering requirement, this finish is gauged to ensure the surface
finish is correct and has its “lay” direction with the correct orientation.
Specifications for special processes. Any engineering specification can include the use of special processes, i.e., processes that cannot be gauged at their completion without destructive examination.
Such process specifications will include a definition of the requirements, and a means to calibrate the procedures for their undertaking.
Procedures for non-destructive tests. Many components will
require special tests to ensure they meet engineering-defined
requirements. Such specifications may not contain any special tests,
but may define closer tolerances, or the use of specific measuring
devices. Such a test procedure will also define the tolerances within
which the test results must fall.
Special instruction for the assembly and alignment. There may
be special assembly sequences and test requirements for main and
sub assemblies. The engineering definition could define these
requirements, test procedures, and acceptable results.
If after assembly an alignment test or checks are to be undertaken to ensure the suitability of the assembly, these will also be identified and defined.
This information is supplied from the design function in the form
of drawings, specifications, and written procedures. It is the responsibility of the engineering function to control product quality through
the quality assurance/inspection department. A major responsibility
of this control is for engineering to be involved in the evaluation and
disposition of any nonconforming conditions that are reported to
them for evaluation.
665
Turbine Steam Path Maintenance and Repair—Volume Two
REVERSE ENGINEERING
The process of reverse engineering provides an engineering
specification for turbine components based on an analysis of components that have been in service, and found to be in a condition
such that they must be replaced. This process will allow components
to be manufactured by suppliers other than the original equipment
manufacturer (OEM). This reversal process is applied either to obtain
parts at competitive prices, or to achieve a delivery that will allow a
unit to be returned to service in a much faster time. However, acquiring enough data to completely define and specify the part can introduce problems related to operation, if sufficient caution is not exercised during the reversal process, and the operating requirement of
the components not adequately considered.
A component produced as the result of reversal must meet the
critical, dimensional, and material properties of the original component as closely as possible. The component may also require certain
proving tests to ensure these critical characteristics, essential to satisfactory operation of the unit, are present in the replacement parts.
Reverse engineering can be (and is now) applied to many
mechanical components. It is necessary to consider the application
of the process of reverse engineering of the more complex components of the steam turbine (although the process as described applies
to most mechanical components in all equipment). This section will
consider those factors that must be reviewed for these complex components to make them suited to the local environmental conditions.
The components can be subject to high stress levels, and can possibly be exposed to severe transients in terms of load, heating, and
cooling rates. The environment within the turbine can also contain
certain aggressive ions, which can concentrate and become active
under certain conditions.
666
Quality Assurance for Replacement & Refurbished Steam Turbine Components
The reverse engineering process can begin from one of two
bases, in terms of the components that are used to determine the critical characteristics providing its technical definition. These are:
•
Samples that have been in use are showing damage or deterioration, and are in need of replacement. These samples
may require some engineering interpretation to obtain complete definition
•
Inventory spares, in which case the samples are in a new
condition, and provide easy definition for the reversal
process. Such components are normally required to replenish the stock of spares before committing the last spares
In either case, the reversal process requires modeling from the
sample components.
If the component or sample that is being copied has been in service, its replacement is generally necessary because its condition has
deteriorated. This deterioration will represent either a modification
or a change of the physical dimensions of the part, or a deterioration
of its metallurgical properties. Therefore, the component will not be
in a condition that conforms to the original design requirements, and
the form and degree of deterioration will have to be identified and
accounted for in the replacement parts. Evaluation and interpretation of the available information from its gauging is necessary.
Usually, there is enough material and dimensional detail of the
original component remaining that a judgment can be made. If such
data are not available, then it is often impossible to undertake
reverse engineering. It is usually preferable in such a case of modeling from used samples, to have more than one sample or specimen
available. Because of the possibility of severe wear, the engineer
responsible for the reversal process is required to have some product knowledge of the components.
667
Turbine Steam Path Maintenance and Repair—Volume Two
If the component to be reverse engineered is an inventory spare,
the problem of obtaining information of the original design is made
much easier, because there should have been no or at least a minimal amount of deterioration while in inventory. However, before the
reverse engineering process is begun, it is better to ensure the component does in fact fit the requirements for which it was originally
produced. Also, if more than one sample is available, the dimensional and mechanical properties should be established from more
than the one sample.
A component that is specified by reverse engineering should
have properties that are equal to or, because of advanced or
improved technologies, superior to the original. A reverse engineered component acts as a direct substitute for the original, in terms
of function and quality. It will have a life expectancy equivalent to
the original, and will not represent an inferior product whose performance will compromise the efficiency or availability of the unit.
Economies in the purchase of replacement parts can be achieved
by reverse engineering and competitive bidding. However, such
economic advantages cannot be achieved if the reversal process is
not undertaken with due regards to the design dimensional and
material needs. These considerations should also address both short
and long term operating requirements. In addition, certain components may require specialized processes, or manufacturing facilities
that may not be readily available. For small quantities it would be
difficult to justify the installation of such plant. Each case must be
evaluated on its own merit.
The concept
To successfully reverse engineer any component it is necessary
to obtain certain information concerning it. This helps ensure its critical characteristics can be reproduced in the components that are to
be provided to the owner. To undertake this reversal process, it is
668
Quality Assurance for Replacement & Refurbished Steam Turbine Components
necessary to define the technical requirements of the component
based on its operating history, knowledge of the type of component
and the type of service it will, or has, experienced. The principal
considerations of reverse engineering are:
•
What are the critical characteristics and functions of the
component?
•
What caused the failure in the original component, and is the
situation or set of circumstances likely to re-occur?
•
Was the failed component a previously reverse-engineered
component? If so, are there any differences in the repeat failure and the need to replace the two original and reversed
components?
•
Which dimensions are critical, and should be measured and
applied to the components?
•
What level of tolerances should be applied to both the critical and other dimensions?
•
What surface finish requirements should be specified for the
various parts of the component?
•
What are the maximum and minimum clearances between
the component parts and other parts within the steam path?
•
What materials should the components be produced from, or
is there a need to re-evaluate the materials and possibly
make changes?
•
Are there any special heat treatment or other special processes required to provide additional mechanical properties to
the material?
•
Should coatings be applied to the components to reduce
wear and damage, and what advantages will this offer?
669
Turbine Steam Path Maintenance and Repair—Volume Two
•
What functional tests should be applied to the components
to help ensure they will adequately fulfill their purpose during operation?
Advanced planning
To achieve the maximum gain from the reversal process, in terms
of costs and delivery, some advanced planning is required. Such
planning permits engineering to be undertaken and manufacturing
space to be made available within the supplier’s facilities. In the
majority of instances, neither the supplier nor the purchaser will
have drawings and manufacturing specifications available. In this
case an investigation is often of value to investigate if “sister units”
were ever built by the OEM. Several options are available to determine if such units exist:
•
Inquire with your own utility operations and maintenance
department
•
In the United States, the Fossil Operations and Maintenance
Information System (FOMIS) [operated by the NUS Corporation and the Electric Power Research Institute (EPRI)] often
are able to supply such information
•
The manufacturers who are to produce the reverse engineered component, or those who are bidding on its manufacture, may also have information on the component
•
The OEM may also make such information available, either
once a sister unit has been located, or to facilitate manufacture and allow a customer to return to service quickly. This
is usually the situation when the OEM cannot support emergency manufacture
In the event a sister unit (or a unit with similar or identical components) is located, the operation and maintenance department of
670
Quality Assurance for Replacement & Refurbished Steam Turbine Components
the owner of the sister unit should be contacted, and possibly a visit
made to the owner to discuss the problem and the needs. It is possible the OEM has a design flaw in the original component on other
units and has failed to inform the owners of sister units. Such discussions with other owners can be beneficial to the reverse engineering process.
Dimensional requirements
When a component is examined, sufficient dimensions must be
recorded to define its mechanical extremities and production
requirements. These dimensions and tolerances are established by
both direct and computer measurement. Also, the use of optical comparators with at least a 10X magnification allows direct comparison.
Dimensions must be sufficient to allow the component to be
drawn and dimensionally defined. It is necessary to define on the
manufacturing drawings, normally from product knowledge, which
dimensions are critical, and which control (or establish) the quality
and suitability level of the component. Remember that each component will interact with (and possibly be assembled to) others. This
will require that the reverse engineered component does not cause
either interference or looseness beyond what can be tolerated by
these other existing components.
If a component has been in operation for extended periods, possibly
at high temperatures, it is possible this environment could have modified
the dimensions that are measured. Such operation could also have
caused deterioration, which precludes the application of the reversal
process. This is particularly important when creep stresses could have
been a factor causing undue deformation and distortion. Under these circumstances, it may be necessary to make measurements of other components that are unaffected or influenced by the component being
reversed, and with which it will interact in its future operation.
671
Turbine Steam Path Maintenance and Repair—Volume Two
Many components require a level of dimensional accuracy that
is established by the detailing close tolerances. This requirement
normally cannot be determined from an examination of one component, although it can be inferred by an examination of a number
of components. (This requirement requires some detailed knowledge
of the design requirements of the component, its application and
operating environment.)
Tolerances are applied to any component to ensure it can be
interchanged, its stress and efficiency levels are acceptable, and it
will meet overall criteria of engineering performance. Tolerances
that are too demanding can cause the manufacturing costs to be
excessive, and provide marginal or no improvement in the performance of the component. Therefore, it requires mature judgment as to
the level of tolerances that are applied to each dimension on any
component that is reverse engineered.
Surface finishes, like the range of tolerances, must be chosen and
applied to allow the component to fulfill its design function without
causing manufacturing costs to exceed a level that makes them unattractive. There are some applications where the direction of finish is
also important. This again should be followed, and processes that
remove material in a different manner cannot be substituted. The
total reversed component represents a complex engineering evaluation and re-specification.
The final manufacturing specification should, therefore, provide a
clear definition of dimensional requirements, provide tolerances within which the components will be acceptable, and the surface requirements at each location where the performance level can be achieved.
Material requirements
Irrespective of its level of dimensional compliance, a component
must also be produced from a material that is compatible with both
672
Quality Assurance for Replacement & Refurbished Steam Turbine Components
its operating environment, and the levels of stress to which it will be
subjected during operation. The original equipment designer selects
the materials of original components to achieve certain characteristics in operation, and to make them suitable for extended life operation. It should also be recognized that some components, such as
gland rings and studs, are of finite life and could require periodic
replacement. However, this should not be a common practice in
blades, and other major components that should be manufactured to
be suitable for the life of the unit.
Because many components will often be exposed to an aggressive chemical environment, it is also necessary to ensure the application of the component is considered in both determining and
specifying the material of the replacement and reversed component.
The chemical properties of a material can be determined by
spectrographic and chemical analysis. The mechanical properties of
the component can be determined by suitable destructive and/or
nondestructive examination. In determining the mechanical properties of the material, the location of the removed test specimen must
be chosen with care. This is to ensure they are not from a location
where there could have been significant deterioration due to operation. It is normally not sufficient for a critical component, which is
known to be subject to high operating stresses, to rely upon a single
test specimen. In addition, there are some components, such as
rotors and “shrunk on wheels” where it is necessary to determine
multi-directional mechanical requirements.
Material substitution
There are, however, occasions when it will be necessary and/or
prudent to consider the use of an alternate material for the component
being reverse engineered. Generally it will be possible to upgrade the
material. This is particularly the case when the component being
reversed is of an old design, and it is possible to take advantage of
673
Turbine Steam Path Maintenance and Repair—Volume Two
advances in material technology. It is also possible to use an alternative material when the component is being reverse engineered
because of original material deficiencies, or poor selection. Such situations could arise where it has been found that the original material
of the component has suffered from some form of deterioration that
could be eliminated or slowed by using a different material.
Within the steam turbine industry, there are often materials having chemical compositions and mechanical properties not present in
commercially available materials. Under these circumstances, it
may not be possible to manufacture the component from materials
with identical properties. Fortunately, there are often other materials
available that will serve as well, although it might mean upgrading
the material to achieve suitability. With modern material production
techniques, it is often possible to use materials, although not conforming to the analysis of the original component, which can provide adequate service.
When a change of material is anticipated, it is recommended the
important design parameters, or requirements of the component be
reviewed, and these compared with known operating requirements
and restrictions.
The OEMs often use materials with properties that are less common than commercially available products for one of several reasons. The materials that the OEM uses were developed for a specific application, and because of the volume of material required, the
OEM can afford the purchase of a special melt or mill run. In addition, it is possible the OEM, because of their volume of production,
would have designed special tools or processes that do not affect the
properties or quality of the final product, but do facilitate the manufacturing process.
Material substitution is often a very real consideration and even
a necessity in reverse engineering. Such change may be required
through both the desire to improve the quality of the product, and
674
Quality Assurance for Replacement & Refurbished Steam Turbine Components
also because it may not always be possible to duplicate the component material properties.
The following rules concerning the application and substitution
of materials should be observed:
•
No material change should be made when an examination
of the component to be reversed indicates there is no justification for it, even if the change is possible
•
If a material cannot be replicated, an upgrade should be
made using a material with the most compatible chemical
and mechanical properties
•
When a change of material is required, and a change in
blade mechanical characteristics has occurred, the replacement material must be examined for its short and long-term
mechanical properties, and its ability to operate in the
chemical and physical environment of the component it will
be replacing
If these three criteria are observed, it is unlikely a change of
material will cause any significant deterioration in the performance
level of the reverse engineered component. However, each substitution must be carefully evaluated and made only after it is determined
it is safe to do so. There are instances when superior materials can
be used. However, it cannot be assumed that the use of a material of
higher mechanical properties will improve the component. This substitution could, in certain instances, place the component into a situation where it is subject to other forms of deterioration.
Special processes
Many components or sub-assemblies of the steam turbine
require the application of special processes to complete their manufacture. These processes include welding, brazing, nitriding, shot
675
Turbine Steam Path Maintenance and Repair—Volume Two
peening, coating, and a variety of others that join, coat, harden, and
generally modify the components and their characteristics to make
them suited for application within the turbine.
If a component must be manufactured by reverse engineering, it
is necessary that the reversal process should identify and specify
these requirements to ensure the replacement components are entirely suited to their intended application. Welding is possibly the most
commonly used form of the special processes; it is also the process
most likely to cause the reverse engineered component to be unsuited to its intended duty if the welding is not performed correctly.
Therefore, the reverse engineering process must establish parameters that will allow those components needing welding to be completed in a manner that will not compromise their quality. The selection of the welding process will depend upon the material of the
component that will be involved, and the type of duty anticipated.
Other special processes, which are essential to the satisfactory
performance and continued operation of the component, will also
require careful specification as part of the entire engineering specification of the reverse engineered component.
Component evaluation and testing
To ensure the adequacy of many components, it will be necessary
to undertake some forms of nondestructive testing or examination, and
in some instances to undertake destructive testing and examination to
ensure the components conform to the requirements of the design. The
reversal process specifies the requirements of these examinations and
tests, and defines the limits within which these components are considered acceptable. Non-destructive examination must be so defined;
the component will achieve a level of acceptance that is consistent
with that of the original specimen. This phase of the reversal process is
important, and provides a level of assurance that the component produced by reversal will fulfill its intended duty.
676
Quality Assurance for Replacement & Refurbished Steam Turbine Components
In some instances, component testing is used to help ensure the
less tangible characteristics, such as vibration frequencies and levels
of deflection are within acceptable levels. In those instances where
the original design criteria are not known, a direct comparison with
the original sample is used to ensure compliance. To allow this, a test
rig is manufactured, which will allow the reverse engineered components and the original samples to be tested under identical conditions,
simulating the operational conditions as much as is practicable.
Component installation
At completion of manufacture, the reverse engineered component must be capable of being installed into the turbine steam path,
while allowing the unit to be returned to service to operate with an
efficiency and reliability at least consistent with that of the original
components. The installation of the reversed components is often as
critical to the performance level of the turbine unit as the actual
manufacture of the components. For this reason the owner should
take care to ensure that qualified people install the components correctly, and there are records produced of any critical characteristics
having the potential to affect the quality of the unit and its short and
long-term operation.
THE QUALITY ASSURANCE
PROGRAM
A quality assurance (QA) program is prepared and implemented
by an equipment supplier to achieve certain requirements or characteristics in the components produced within his or her facilities.
The QA program is normally not limited to measuring and gauging
products; this inspection activity represents only a portion of the
total QA program.
677
Turbine Steam Path Maintenance and Repair—Volume Two
A quality assurance program should include various items controlling total quality. These would typically include:
•
for an alternative supplier, procedures for reverse engineering. This will ensure the reverse engineered components
comply (are equivalent to or better than the original product)
and that tolerances are to industry standards. It also includes
production planning to ensure delivery dates are met
•
procedures for controlling the purchase of materials and other
items that will affect the quality of the finished components
•
details of inspection and test methods, their implementation,
and a definition of those responsible for this work and their
authority. This inspection includes incoming, in-process, and
final inspection
•
methods for controlling the calibration and use of measuring
instruments
•
methods for calibrating and controlling special processes,
including the qualification of operators responsible for their
application
•
methods for reporting and evaluating nonconforming conditions as they arise in the manufacturing facility as well as the
isolation of nonconforming items from those that comply,
and are suited for shipment
•
methods of packing and preserving the components ready
for shipment
•
details of record storage, retention, and retrieval
Such a program should have a person within the supplier’s
organization responsible for its implementation. This person must
have enough authority to halt production until corrective action is
taken (if it is determined the manufacturing process used to manu-
678
Quality Assurance for Replacement & Refurbished Steam Turbine Components
facture the components is out of control to the extent they would
affect product quality).
It is normal for a purchaser, before placing an order, to review
the QA manual and undertake an implementation check to ensure
the program (as defined in the manual) is operative, and the components have a way of being traced. For small orders, the implementation check is reserved until after placement of the order, and then
corrective action requested, if the program does not comply with the
purchase specification.
THE QA MANUAL
An important component of a complete QA program is the availability of a suitable manual that defines the quality philosophy and
procedures. The manual is also a description of the working program, and provides procedures that are to be followed to assure consistent and acceptable quality is achieved.
The intention of a QA manual is to propound a philosophy. The
QA manual is a working document, which must reflect what is
done within the component manufacturing plant to assure quality.
Therefore, the manual must reflect honestly and accurately what is
to be done within the plant, and the minimum standards that will
be accepted.
It is normal (for major purchases) that the manual will be
reviewed before a purchase order is awarded, and an “implementation check” made to gain assurance that the program is being implemented. If discrepancies are found between the written program and
the actions within the manufacturing department during the implementation check, these differences are normally resolved before a
contract is let. It is necessary to reach a solution that allows practice
679
Turbine Steam Path Maintenance and Repair—Volume Two
to comply with the written standard, and yet causes the least disruption to the existing system. However, it is inevitable some of these
discrepancies will represent faults, or at least the potential for faults,
in the existing program. These need to be acknowledged and
addressed. Such areas must be corrected in the preparation of the
final procedures to be implemented.
THE ENGINEERING REVIEW
The purpose of an engineering review is to address those areas
where it is possible for errors or nonconforming situations to arise,
and where the purchasing engineer wishes to exercise some level of
control over the project. Such control normally takes the form of
approving specifications that influence product quality, and establishing a means of controlling the disposition (recording and correcting) manufacturing errors, should they occur.
During the engineering review, agreement should be reached on
the inspection plan, provided by the supplier to the purchasing engineer, and also on hold points, which should be included in a total
surveillance plan.
The engineering review can also be a pivotal point in the project
where decisions are made concerning the need, application, and
control of special processes, and agreement is met on matters such
as dimensional control of the components themselves.
Therefore, the following matters should be addressed at an engineering review:
•
680
Quality assurance program—This part of the engineering
review should examine the QA program, and its adequacy to
help ensure that components are produced to the engineering
Quality Assurance for Replacement & Refurbished Steam Turbine Components
specification. This includes meeting the purchaser specification and the engineering standards of the supplying company
•
Drawings—The components to be supplied will be produced
to an engineering drawing. These drawings must reflect the
following: dimensional tolerances, surface finish requirements, the processes to be used in the production, and any
assembly phase (although these processes may be defined in
the form of internal standards, which must then be examined)
There will need to be agreement as to the availability of
drawings to the purchaser engineer for approval and record
•
Manufacturing processes—There will be a need to examine
the proposed manufacturing processes to be used, their
method of control, and how the supplier will monitor them.
For special processes (those that cannot be gauged for compliance at completion) agreement on methods of process calibration should be defined and agreed upon, if necessary
With many processes there is often the involvement of proprietary information, which will need to be considered. The
purchaser and supplier will need to reach an agreement on
how this aspect of the contact will be managed
•
Non-conforming components—During the manufacturing
process, it will often happen that components are produced
outside engineering specification. This will normally be
determined by the supplier’s inspection department. When
such situations occur, the resulting component will be examined, and an engineering disposition developed. (This is considered in more detail in following sections)
•
Inspection and test plan—The inspection and test plan is a
chronological listing of all tests and examinations with
which the components must comply to be considered
acceptable. (This is discussed in following sections)
681
Turbine Steam Path Maintenance and Repair—Volume Two
•
Hold points—The engineer will normally undertake some
level of inspection and product verification during the course
of the supply contract. There can also be a point beyond
which the manufacturing process may not continue before
the engineer has had the opportunity to examine the product, its records, and even witness a particular test. This will
need to be agreed between the parties to the contract
It is normal for the supplier to provide warning of the
approach of a “hold point” some time before this point is
reached. The engineer can then elect to be present to witness, or “waive” this opportunity or hold point
Note: The specifying engineer has the overall responsibility for
the quality or compliance with the purchase specification. While the
engineer may in fact become involved with a portion of the more
critical aspects of production, the responsibility is most often delegated to an inspector or surveyor who will undertake the majority of
such verification activities.
682
•
Access to sub-supplier plants—Depending upon the nature
and extent of the contract, there could be a need for access
to sub-supplier plants to inspect subassemblies or material.
The need for such access and the availability for other information should be addressed at a design review meeting prior
to award of a contract. Such access should also be defined in
the purchase specification prepared by the engineer
•
Quality records—An important consideration for the purchaser of any equipment is the level of recorded information
concerning the parts he or she will receive from the supplier. This must be agreed at the engineering review, and will
probably have been defined in the purchase specification
Quality Assurance for Replacement & Refurbished Steam Turbine Components
THE RESPONSIBILITY AND
ADMINISTRATION OF A
QA PROGRAM
In establishing the responsibility for product quality, it is necessary
to differentiate between “quality assurance” and “quality control.”
Quality assurance (QA) is a program that defines the steps considered
necessary to ensure the final product of a manufacturing process
meets the requirements of the design specification. Quality control
(QC) is a function or element of a quality assurance program, and is
the proving or checking of activities associated with such a program.
It is necessary at the outset of examining a supplier for program
preparation and implementation, to differentiate between these two
terms, and what they imply for the product. Without this understanding, there will continue to be a belief that quality will be
assured by extensive testing and gauging. This is only partially true.
This does not provide complete assurance of quality. The two
aspects considered in examining a supplier program are:
Program preparation. The QA program is defined in the QA
manual, which is a component of the QA program. This definition
provides guidance of what steps the supplier will take throughout the
material procurement manufacture and shipping phase of a supply
contract to ensure the delivered goods meet the specified requirement. This QA program may also, depending upon its level of
sophistication, define any internal engineering review and how
materials and component parts will be manufactured, controlled,
and inspected prior to use.
Such a program can include many other procedures and controls
depending upon the components being supplied.
683
Turbine Steam Path Maintenance and Repair—Volume Two
Program implementation. Once a QA program has been
designed, and documented by means of a manual, it must be implemented within the plant facilities. This implementation can, depending upon the program design, extend from marketing through engineering, to the final phases of manufacture, testing and shipping.
An implementation review requires that the various functions
covered by the program, are adhering to it and follow fully the intention of the total company in achieving a quality product, i.e., one
that complies with design specification.
Company organization. For a quality program to be effective, the
manager responsible for the preparation and implementation of the
program must have adequate authority to administer it. This
requires:
•
that this manager must have direct responsibility for all personnel who are responsible for gauging and reporting quality and nonconforming conditions. This will normally
include QA staff and all inspection and nondestructive testing operators
•
that this manager must report to an authority beyond the
influence of those people and departments of the organization with a responsibility for production. Such reporting
authority can include the works manager, or manager of
design engineering
A well designed and operated quality program can act in an
anticipatory manner, making quality checks early enough within the
total manufacturing cycle, so that errors are detected and/or corrected before they become significant and jeopardize a complete product, or procurement schedule.
684
Quality Assurance for Replacement & Refurbished Steam Turbine Components
Responsibility for product quality
Product quality is a responsibility of design engineering. Design
engineering undertakes the basic product analysis and defines the
characteristics that must be achieved in the final components to
ensure they will operate as expected in the turbine to meet predicted
performance. The design engineer provides to the manufacturing
department a clear definition of the material properties, possibly the
method and sequence of manufacture, the dimensions and tolerances
within which the product must fall to be acceptable. This engineering
definition will also provide details of any special processes that are to
be undertaken, e.g., welding, brazing, heat treating, coating, etc.
Design engineering provides a definition of quality of the component. However, the responsibility for ensuring product compliance with this definition is met by the quality assurance department,
which monitors the total procurement process and all other aspects
of manufacture, and has a responsibility for advising design engineering in the event a nonconforming condition is detected.
Many QA programs place a considerable amount of responsibility for quality with the individual shop floor workers, for “in-process”
inspection, however the responsibility for final acceptance of each
stage of manufacture is placed with the “QC” department for monitoring and design engineering for evaluation and acceptance or
rejection of any noncompliance.
Acceptance does not imply that at each manufacturing step the
design engineering function must become involved. Rather, in the
event a nonconforming condition arises, then only design engineering can evaluate this condition and establish an accept/reject decision, and then develop any corrective action that may be necessary
and acceptable to total product quality.
The implementation of the QA program is one assigned to the
QA department, managed by a QA manager. However, the QA
manager is merely the keeper and enforcer of the program, with
685
Turbine Steam Path Maintenance and Repair—Volume Two
power to monitor its application, and report discrepancies to the
design function.
The inspection department
Many companies have an inspection department located within
the manufacturing department, which reports to the manufacturing
function. By the definitions and requirements of a QA program, and
simple logic, this represents a totally unacceptable situation. This is
because there could be a direct conflict, and pressure placed on
inspectors, urged by their immediate and ultimate superiors to pass
a marginal, or just outside specification piece, rather than declare it
as nonconforming.
The obvious solution to this problem, if it exists, is to have the
QC function report to the QA manager, or alternatively to engineering, which has the final responsibility for product quality.
Companies with no separate QA department often have the QC
department report to design engineering, or other high level management position that is not concerned with manufacture. In this
manner, the quality function is completely divorced from manufacturing, and there is little possibility of conflicts arising.
The final quality
What quality level must be achieved? To produce a component
that is better, produced to tighter dimensional tolerances than specified, or having a surface finish several degrees finer than necessary,
or produced from a better material, may be good for the strength or
appearance of the component. However, if this is beyond designdefined quality, it is wasteful in terms of the costs of manufacturing
the product. For these reasons, the requirements specified by the
design function that represent the minimum acceptable should be
observed, but no effort or costs should be expended in bettering
686
Quality Assurance for Replacement & Refurbished Steam Turbine Components
them. This additional effort will not improve the quality of the product, which meets supplier specifications.
Design engineering will assess quality requirements of the products it designs. It will specify for each of these components what it
considers to be the quality requirements. These requirements are
defined in a number of ways, and presented to the manufacturing
department in a number of drawings or specifications. These specifications can also include many specific requirements and other less
obvious parameters for more complex elements.
The object of the QA program is therefore to direct the attention
of the company, and its employees, to ensuring the products meet
design specification, which is selected to conform with the purchaser specified, or agreed to requirements. These requirements are considered a minimum, but are intended to ensure the product meets a
specified level of performance. It is possible companies are prepared
to provide a product that better meets or exceeds these minimum
requirements. If this is a conscious policy of the company, this
upgraded requirement should be provided by the engineering specification, not left to arbitrary decisions within the manufacturing
function, as this could ultimately lead to confusion, and the ability
to evaluate and accept nonconforming items. Spending time and
money on achieving excess quality characteristics is expensive.
However, this could be the price a supplier is prepared to pay to
achieve a superior product in appearance or performance. This
could be a good business decision, but is not the business of the QA
program until it is made so by engineering definition.
687
Turbine Steam Path Maintenance and Repair—Volume Two
THE INSPECTION AND
TEST PLAN
The supplier’s engineering design function is responsible for
product quality. It has been stated that the necessary information is
supplied to the manufacturing department in the form of drawings,
specifications, etc. These establish the “accept and reject” criteria
that must be met.
In a structured and formal QA program, the engineering function, “the arbiter of quality,” prepares quality instruction to other
functions in an “inspection and test plan” (I&TP). The I&TP is a
responsibility of design engineering. It can, however (and normally
should) have important contributions from other departments within
the company organization. However, engineering is ultimately
responsible for approving and accepting such contributions.
In its simplest form, the I&TP is a chronological listing of the
inspections and tests that must be undertaken at each stage of the
total production process. The I&TP provides a definition of the standards of acceptability at each stage. In the event a product (at any
phase) is found by inspection to be in a nonconforming condition,
this must be reported to engineering, which has the responsibility to
evaluate the condition, and make a decision to “scrap,” “rework,”
“repair,” or “accept-as-is.” The majority of these dispositions are normally quite evident, and “scrap” decisions are often made by manufacturing or inspection, but ultimately are the responsibility of
design engineering.
In some contracts for complex components, nonconforming
conditions may, by agreement, need to be referred to the purchaser,
both for information, and in some instances for acceptance, except
for a “scrap” decision. The possible exception being that if “scrap”
688
Quality Assurance for Replacement & Refurbished Steam Turbine Components
has a serious effect on schedules, even these require purchaser
review and acceptance.
The I&TP, and the need to refer nonconforming conditions to
engineering for resolution, and the purchaser for approval, can be
the most contentious issues in the implementation of a QA program. These issues are often resolved only after protracted negotiations. In these instances, the requirements defined in the contract
must govern.
In the case of nonconforming components, the QA department
is not normally qualified to establish acceptability. However, in
lower-level programs, the QA manager does often have such responsibility. Manufacturing cannot have such authority. Manufacturing is
a department, which by definition, is more concerned with “on
time” delivery. Therefore, who within the supplier’s plant is better
placed to decide the acceptability of a defect than the designer, or in
his or her absence, the QA manager, who monitors to some defined
quality requirement? The design engineer or QA manager normally
has available information on the performance potential of the components, and is more aware of the probability of failure if a situation
is not adequately corrected.
The I&TP should also define those records to be developed for
the manufactured products.
PURCHASER ASSURANCE
OF QUALITY
The objective for an equipment purchaser in undertaking any
form of “surveillance” or “inspection” at a supplier facility is to
achieve product quality. But what is quality? A quality product is
689
Turbine Steam Path Maintenance and Repair—Volume Two
defined as one that conforms in all respects to the requirements of
design. Any effort made by a manufacturer to achieve a product with
compliance, in terms of dimensional conformance, surface finish,
mechanical properties, or other characteristic beyond design engineering specification, cannot be considered necessary to improve
the defined quality level. Certainly, such narrower control may
improve the value of the product. But the assumption in accepting
the design-specified value of these various component characteristics is that the value defined by design is acceptable from technical
considerations. Such a finish may have certain ascetic value, or can
even be used as part of a marketing strategy, but if it is beyond the
design requirements, no quality value can be attached to it.
To obtain assurance that delivered components meet or comply
with the design definition, and can be considered a quality product,
there are certain steps the purchaser can take. These include an
“engineering review” of the design and manufacturing specification, then the undertaking of surveillance of the components during
the total procurement process, from material procurement to shipment, to ensure they comply. Alternatively, it can include only that
part of the total process considered most likely to influence quality.
Product surveillance will help establish that elements are produced
in accordance with the design specification, and that all steps,
processes, and checks included in the manufacturing processes are
undertaken correctly.
PRODUCT SURVEILLANCE
During the total procurement cycle, there are a number of
actions that influence the quality of the product. It is normal for
many purchasers to monitor details of production to help ensure
components are produced to specification, and will perform ade-
690
Quality Assurance for Replacement & Refurbished Steam Turbine Components
quately when installed in the unit. This surveillance normally takes
the form of “monitoring” supplier activities to ensure his or her
actions produce a product that conforms to his or her design and the
purchase specification. It is also normal for the purchaser’s inspection representative to report on any deviations, and to the extent possible verify the error and suggest to the engineer possible corrective
actions.
In this section, the company inspection technicians will be
called inspectors, while the purchaser’s monitoring will be undertaken by the “quality surveyor,” or “surveyor.”
Surveillance involves activities beyond what is normally termed
inspection, although the most common definition for the person
responsible for these activities is normally the surveyor, who directs
his or her activities towards critical characteristics of the products.
The surveillance (or inspection activities) must be carefully
planned to be cost effective. These activities must then be undertaken by a surveyor who has a knowledge of the component, preferably
prior experience with the supplier of the equipment, and a clear
understanding of those characteristics of the component that are
most important to defining its total quality and suitability for use.
Preparation for inspection at the
supplier facilities
Prior to inspection at the supplier facilities, the surveyor must
identify the critical characteristics of the components being supplied. While the individual surveyors often have extensive product
knowledge, it is not their responsibility to identify these quality
requirements. Where then does this responsibility lie, and where will
the information and quality definition come from? There are a number of sources including:
691
Turbine Steam Path Maintenance and Repair—Volume Two
The inspection and test plan (I&TP). It is normal for each component to be manufactured to a test and inspection plan, prepared
by the equipment supplier, and approved by, or negotiated with the
requisitioning engineer as part of the “engineering review meeting.”
This I&TP will be performed to the standards of acceptable quality.
The technical purchase specification. The technical portion of
the purchase specification should define overall quality requirements. This is done in terms of performance requirements and the
ability of the components to be assembled, maintained, and possibly
repaired. However, it is unlikely that such specifications can contain
all the information the surveyor will require. Therefore, initial and
continuing meetings between the engineer and surveyor are normally necessary to provide the final definition of quality characteristics.
The purchasing (requisitioning) engineer. The requisitioning
engineer has the primary responsibility for defining, and as needed,
identifying those characteristics that require special, or closer attention during the total procurement process. This engineer could also
bear the individual responsibility for writing the technical portion of
the technical purchase specification. This specification should
define the quality requirements in terms of performance (efficiency
and reliability), and possibly define individual requirements covered
by supplier material and process specifications.
Previous operating experience. Records of operating experience
with the component, its anticipated life, failure rate and mode, and
overall operating performance are normally available at the power
plant. These experience factors should be considered in selecting the
areas to be given special attention during the surveillance review
and recording process. This information can in fact be among the
most important information available; also, the requisitioning engineer can often be located at the generating facility.
692
Quality Assurance for Replacement & Refurbished Steam Turbine Components
Records of experience with the supplier. When a new supplier
is to be used to provide components, there is a learning process
involved, as the supplier and surveyor become familiar with the total
interaction involved, and with each other’s mode of operation.
When a supplier is used with whom the surveyor is unfamiliar, it is
often advisable that records of this supplier’s past performance are
reviewed for the components involved. This review will help the surveyor to direct his or her attention to areas that have previously
proven to be those where nonconforming conditions can arise.
Anticipated changes in the operating modes of the components.
There are situations where the turbine is to be operated in a changed
mode. This changed mode could be coincident with the replacement of worn and damaged components before it is returned to service. Similarly, older units will often be refurbished for return to service, or will have their capacity or steam conditions modified a little.
If this occurs, the requirements of the new components could be
changed in small, but possibly important ways. The surveyor should
be made aware of such changes and be prepared to review any
changes this will involve in the total supply process.
Minutes of the design review meeting. The design review meeting can be an important part of the award of any contract. The minutes of this meeting will record any technical changes agreed to
between the supplier and purchaser, and can include significant
changes to the processes involved, the records to be produced and
provided by the supplier, and other significant changes to the components to be supplied.
The minutes of this meeting will, after agreement, form part of
the supply contract, and as such will take precedence over the portions of the specification originally issued to solicit bids. The surveyor must review these minutes.
693
Turbine Steam Path Maintenance and Repair—Volume Two
Inspection (surveillance) activities
and responsibilities
When agreement has been reached between the requisitioning
engineer and the surveyor as to the activities representing the most
critical operations in the total supply process, and the acceptance
standards have been defined, the surveyor must plan his total
involvement. This can include, but not necessarily be limited to the
following:
Quality program review. The quality program within which the
components are to be manufactured will have been agreed to at the
design review meeting. However, such approval could have been
given without any “in plant” audit at the supplier’s facilities, and
based on what is perhaps a cursory review of the manual, or on the
understanding that the program will be acceptable from previous
and other client’s acceptance. Therefore, a preliminary activity of the
surveyor can be to undertake a more in-depth review of the quality
program, and then to review supplier in-plant activities and records
to ensure its implementation.
The review and implementation check of the entire program
can represent an expensive undertaking. Therefore, in the case of
the supply of a short delivery, or critical component requiring the
application of only a portion of the total QA program, this initial
program review can be limited to those elements of the program
that will affect the quality of the components being supplied under
the contract.
Inspection records. An integral part of any QA program is the
documentation of the results of tests, examinations, and measurements made of the components at various stages of the procurement
and production process. These documents are made available to the
surveyor for review and acceptance. For any document to be considered complete, there are two necessary requirements. The document
must be signed and dated, and the person undertaking the supplier
694
Quality Assurance for Replacement & Refurbished Steam Turbine Components
inspection must be qualified to complete this work and accept
responsibility for accepting it. The signing inspector must be qualified
to undertake that work, which can be anything from the use of normal measuring instruments, to the most sophisticated nondestructive
examination.
The purchaser’s surveyor therefore has the responsibility of
ensuring the qualification of the supplier inspector before any determination, review, or acceptance of the results themselves.
Inspection instrument calibration. For a supplier inspection to
be valid, and to provide acceptable verification of the product or
process, the instruments employed to undertake tests and measurements must be calibrated. It is a normal part of any quality program
that an internal system exists requiring the periodic recalibration of
all instruments used to verify a product. Depending upon the type of
instrument, the calibration period can range up to as much as a year
for some complex nondestructive testing devices. Instruments
employed to measure physical dimensions are normally calibrated
on a monthly basis.
Each time a device of any type is recalibrated, the calibration
date and the result of the recalibration are recorded in a central log.
Any tests or measurements taken with an uncalibrated instrument
are unacceptable, and must be repeated with a certified instrument.
The use of personal measuring instruments within any supplier
plant is unacceptable, unless these instruments are entered into the
general calibration system and are recalibrated in the same manner
as company-supplied instruments.
The surveyor has the responsibility to ensure that all instruments
used to gauge products are calibrated, and that such calibration is
readily observable by the display of a dated calibration sticker or
other suitable means.
695
Turbine Steam Path Maintenance and Repair—Volume Two
Hold or witness points. Certain critical tests within the total
manufacturing and proving cycle of some components or assemblies
are of sufficient importance; they must be witnessed by the surveyor. These are known as “hold points.” When these stages of manufacture are reached, it is necessary for the manufacturer to notify the
surveyor that such a test is to be performed. It is also necessary to
provide this information at some specified time before the test, e.g.,
72 hours, to provide sufficient time for the surveyor to be present to
witness the test. These are also points beyond which production,
manufacture, or assemble cannot proceed without the agreement or
approval of the surveyor, or in certain instances where a requisition
engineering review of the results are required, without that purchasing engineer’s approval.
When such a test is to be undertaken, the surveyor has a responsibility to become aware of the test procedure, details of the results
to be achieved, and information of the results to be recorded. If special instrumentation is to be used, the surveyor needs to be sure of
its calibration, and that it is in conformance with that identified in
the test specification.
Many such tests contain what is considered proprietary information, and copies (while not provided to the surveyor or requisitioning engineer), will be available for examination at the test. In the
case of this type of procedure, these test specifications will normally have been reviewed by the engineers during the bidding phase,
and were accepted. If the test procedure was not available for examination prior to the witness visit, the supplier must allow sufficient
time for the surveyor to review, and possibly discuss with the requisitioning engineer what he or she considers to be the most critical
steps to monitor.
Nonconforming items. It is almost inevitable that during the
manufacture of complex components that some nonconforming
conditions will occur. When these occur there are certain steps in
providing a disposition of the component that must be considered:
696
Quality Assurance for Replacement & Refurbished Steam Turbine Components
•
The generation of a nonconformance report by the supplier’s
inspector, copies being provided to the surveyor
•
An evaluation of the condition by the supplier’s design engineering function, and the generation of some corrective disposition
•
Provision of a copy of the engineering disposition to the requisitioning engineer for acceptance/rejection
•
Undertake corrective action to an agreed plan
In the case of a nonconformance, the surveyor is to keep the requisitioning engineer advised, initially of the occurrence (which is
often done in parallel with the suppliers notification), then to advise
the engineer of any possible corrective action considered appropriate in the circumstances of the error, and finally to monitor the
agreed upon corrective action.
Special processes. Special processes are those that cannot be
examined without destructive testing when the process is complete.
In these circumstances it is necessary to calibrate and confirm the
process by destructive means, and then monitor its application to
ensure the process is performed in an acceptable, and controlled
manner when applied to the components.
The surveyor should participate in this calibration process and
then monitor to ensure the correct procedure is applied throughout
the total process. This does not require 100% attendance at the
process, but requires, if possible or necessary, repeat calibration
processes are undertaken to ensure there has been no “relaxation” of
the process during its application.
Dimensional and surface conformance. The dimensional checks
and review of dimensional inspection records of components probably comprises the major portion of any surveillance activities
undertaken by the surveyor. In the early phases of a contract, or
697
Turbine Steam Path Maintenance and Repair—Volume Two
when the plant is new to the purchaser, it may be that dimensional
compliance is checked continuously. However, as a purchaser company develops confidence in the supplier, it will become less common for the checks to be repeated by the surveyor; rather reliance
will be placed on the supplier’s “inspection department records,”
and a review of these, together with an examination of any nonconforming material reports.
However, sample and “spot” checks are never abandoned. For
large supply contracts, “patrol inspection” or arbitrary dimensional
checks are still considered a suitable manner of helping ensure quality.
Material compliance. These activities will normally involve the
review of material test certificates and the comparison of these with
supplier material specifications. Such activities should also include
the review of heat treatment charts for special application materials,
and can, in the case of critical materials, involve the witnessing of
destructive and other tests at the plants of material suppliers. For
reverse engineered components, it can also require involvement in
material identification.
Nondestructive testing. An integral part of a supplier’s inspection activities includes the application of nondestructive testing to
ensure the material and final components conform to design-specified requirements. The surveyor should review the results of these
tests, and on a spot basis witness those that are critical, and, if substandard, could lead to poor performance of the delivered components. It is normal for the surveyor to request copies of all critical
material test certificates.
Supplier purchaser control. When the manufacturer of turbine
components employs sub-suppliers for the procurement of critical
components, there is a need for the surveyor to monitor activities
within these plants also. This need should be defined in the purchase
specification. However, the primary responsibility for monitoring
698
Quality Assurance for Replacement & Refurbished Steam Turbine Components
quality remains with the primary supplier, but his checking of subsupplier quality must be documented, and monitored by the surveyor.
Documentation and shipment. Prior to shipment, the surveyor
should review all documentation relative to the supply of the components, and where appropriate ensure copies of these are made
available to the purchaser. The documents supplied to the purchaser are those required for records to allow product quality to be verified at some later time. This will not normally include proprietary
information of a design nature, but depends upon what is negotiated as being in the scope of supply at the time the contract was let.
It is important to recognize that surveillance does not require the
surveyor to undertake double inspection as a normal course (however, it may be more intense in the case of a contract with a new supplier). Rather the surveyor should direct his or her efforts to monitoring the records generated by the supplier inspection and QA staff,
to ensure both the production processes and inspections were
undertaken by qualified staff and adequate records, providing traceability, are generated.
NONCONFORMING
SITUATIONS
During the total procurement cycle of all components, from
material specification to preparation for shipping and packing, it is
possible for a situation to occur where the products do not comply
with the design specified requirements. This is referred to as a nonconforming situation. In this condition the components are not technically in compliance, and should not be shipped.
699
Turbine Steam Path Maintenance and Repair—Volume Two
When the manufacturer or supplier inspection indicates a nonconforming condition exists, this must be evaluated. The logic
process of evaluation for performance potential is shown in Figure
11.13.1. This figure outlines the avenues the supplier engineer will
explore in deciding if corrective action is possible, and from among
several that represent the most appropriate corrective action. With
such an evaluation, one of four different decisions can be reached.
In some circumstances, the decision is relatively simple to make,
and in fact is obvious. In others, options are available, and a decision is made based on the probability of failure or poorer than predicted efficiency, the possible cost of repair, and the ultimate consequences, including the correction of consequential damage, which
is the result of not taking appropriate corrective action. In certain situations where delivery is a critical consideration, this must also be
considered in the total evaluation.
When such a nonconforming condition arises, it is necessary for
the supplier to review the total situation, evaluate various corrective
actions, and make a recommendation for correction. The responsibility for this correction is one placed on the supplier’s design engineering function, since only the design function is qualified to make
such judgment. Once an evaluation has been made and a corrective
action decision reached, it is normal for this to be referred to the purchaser for information or approval. This depends upon the form and
wording of the contract, which in turn depends upon how the purchase specification was prepared and negotiated.
The four basic decisions, and the logic that should be considered
are outlined in Figure 11.13.1. The possible corrective actions, in
terms of their possible effect on performance are:
Scrap and replace. This is a decision that is reached when a
manufacturing, process, or testing error has occurred, and the component is no longer suitable for its intended purpose. This component must be replaced, either because it is impossible to correct the
situation adequately, or because the cost of correcting it is greater
700
Quality Assurance for Replacement & Refurbished Steam Turbine Components
than starting re-manufacture. Often, this is a self-evident decision,
and there is little need for evaluation. At other times, this decision is
reached only after extensive review of the options. In such a situation it is judged the risk associated with accepting the component is
too great.
MANUFACTURING
NONCONFORMANCE
Does it affect
availability?
Does it affect
efficiency?
Yes
Does it affect
maintainability?
No
No
Yes
Does it affect
short term
efficiency?
Does it affect
long term
efficiency?
Does it affect
reliability?
* Accept
as is.
Yes
No
Yes
Cost of operating
inefficient unit.
Interchangeability.
Long term
integrity.
Short term
integrity.
Safety.
Accessibility.
Correctability.
Yes
No
Accept
as is.
Yes
Scrap and
replace.
Cost of making
correction.
Decision to
correct.
Yes
No
Rework.
No
Repair.
Figure
11.13.1
Fig. 11.13.1—The evaluation process for
a manufacturing
nonconformance. This figure
The
evaluationto
process
a manufacturing
figure should be
should be
compared
Figurefor1.8.1
of Chapter 1nonconformance.
for a field foundThis
nonconformance.
compared to figure 1.8.1 of Chapter 1 for a field found nonconformance.
701
Turbine Steam Path Maintenance and Repair—Volume Two
There are occasions when to scrap and replace is prejudicial to
the project schedule. Under these circumstances, the evaluation
becomes more complex, and there are greater constraints placed on
the component supplier. If such a nonconformance occurs, the purchaser must be fully aware of the situation, because to use the component could mean the unit will operate at risk for a period, and until
conforming components can be made available and placed in service.
There are situations where the nonconforming component can
be used for a specified time, but must be replaced at the first opportunity. The replacement part is provided at no cost to the purchaser.
Repair. A repair corrects a nonconforming condition, but does
not re-establish the original design characteristics within the component. It is often possible to make repairs to manufactured or partially manufactured components, sufficient to allow them to be
placed into service. Depending upon the nature of the nonconformance and of the repair, the affected component may or may not
ultimately require replacement.
Recently, there have been significant advances in many repair
and refurbishment techniques (see chapters 7, 8, and 9). This is particularly true in cases involving welding, where new technology has
made available materials and techniques capable of extending the
useful life of many components, which prior to the development of
these fusion techniques would have been scrapped.
The technical requirements for performing such repairs are stringent. However, if they allow a component to be saved, and the unit
placed in service on schedule, rather than require an extended outage,
or until replacement components can be produced from the beginning
of the procurement cycle, then the costs and minimal change in risk
levels associated with such repairs can often be justified.
702
Quality Assurance for Replacement & Refurbished Steam Turbine Components
Again, the repair decision is normally made after a review of the
nonconformance, an evaluation of the possible repair procedures,
and the level of risk involved.
Rework. To rework a component that has some form of nonconforming condition during initial manufacture is to undertake actions
that will return it to a condition in which it is equivalent, in terms of
dimensional requirements and mechanical properties, to those it
would have achieved had the nonconforming condition not
occurred.
When a nonconforming situation occurs in the manufacture of
new parts, if a condition equivalent to those required of the new
component can be achieved, such rework should be undertaken.
However, it is the responsibility of the design engineer to determine
that such equivalence has been achieved.
Under most conditions this is an easier decision to reach, if
proven procedures for the rework are available, particularly when
applied to components in which stress levels are low, and are in a
low risk level or environment.
Accept-as-is. An accept-as-is decision is one that permits a component to continue in manufacture, or to be used with no effort
being made to correct the nonconformance. Two reasons for reaching and deciding upon this course of action are:
•
There is little need to make any corrections. To make them
will add no, or at best marginal improvements to the turbine
performance
•
The cost and time of replacing, repairing, or refurbishing
cannot be justified, either for the turbine or for the degree of
noncompliance in the components. Often to rework or
repair the situation could increase the risk to performance
level
703
Turbine Steam Path Maintenance and Repair—Volume Two
This “accept-as-is” decision is often based on the experience of
the supplier design engineer and can only be made by being aware
of any risks that are introduced. Such a decision should not be made
as a desperation measure. The risks, if any, should be fully evaluated, and the options, from an extended outage before return to service, and the probability of failure, must be fully considered.
The “accept-as-is” decision can often be made, being aware of
the risks, while replacement parts are obtained.
This decision or evaluation process can be complex. Occasionally the solution is self evident, such as when a nonconformance
exists to the extent parts cannot be used, manufacturing must cease,
and the parts replaced. In those instances of a nonconforming condition, when there is no time to correct the situation before the
scheduled date for return to service, mature judgment is required on
the part of the design and operating engineer, and acceptance of the
fact that the unit will operate at risk if the nonconforming components are used.
AVAILABLE QA PROGRAM
When a company installs a QA program, it must represent their
actual “in-plant” procedures and activities. These activities should
be reflected in both the QA manual, and the procedures developed
to define their methodology. There are many national and international standards in existence, many developed to provide guidance
to suppliers of equipment to the military or other complex and
demanding industries. These programs provide a secure basis from
which to develop a company program. However, the supply of turbine components may not be as complex, and while the supply of
such components that do not meet design specification can have
704
Quality Assurance for Replacement & Refurbished Steam Turbine Components
serious and expensive consequences for the purchaser, the installation of a program with complexities beyond what is required to provide a suitable product is wasteful in both time and costs to the supplier and purchaser.
An international standard that is developed and utilized now by
many companies as a basis for their internal programs is the ISO
9000 series. This international series of programs is three tiered, having three levels of complexity.
The most appropriate level in any component supply contract is
dependent upon the complexity of the supply process, and should
reflect the degree of any project that is to be undertaken. The complexity of the program should not be confused with the complexity
of the component form and material. A complex component can be
produced within a relatively simple program if the processes in its
production are understood, and there are no anticipated difficulties
within the total supply process.
THE MACHINING OF
TURBINE COMPONENTS
The components of the steam path are produced to exacting standards, and design specifications are set so the supplier of the equipment can provide a “quality product” at an acceptable cost. This
allows the operator to have an efficient and reliable unit available on
the system. Also, the manufacturer has an income that covers cost,
provides a profit to his shareholders, and has sufficient funds remaining to support further development of the product and its components.
An integral part of establishing the final cost and quality of the
steam turbine is the expertise and control of the manufacturing
705
Turbine Steam Path Maintenance and Repair—Volume Two
process. This process, while very much influenced by the individual
machinists and their attention to detail, is also a product of the engineering specifications, and the machine tools and cutters made
available to the machinists.
The total production of the steam turbine involves using a number of different metal forming processes and cutting techniques. A
number of different techniques are required for different components. Also, the level of finish and tolerances applied can be different, not only from component to component, but within the components themselves. In general there are more stringent requirements placed on the rotating components than those that are stationary, such requirements being related to the effects of stress rather
than expansion and flow efficiency.
In general, the cutting or forming processes selected for the production of each component of the steam turbine are those most suited to provide the tolerance, and produce the surface finish required.
The manufacturer of steam turbines will have available machine
tools suited for each major component. However, there is continual
development in manufacturing these components, and certainly the
introduction of advanced computational fluid dynamic techniques
has allowed a more sophisticated aerodynamic form to be defined.
For these reasons there can be situations where components, while
manufactured with considerable care, are not necessarily produced
by the most suitable and economic means in the early phases of production of the particular components.
The cutting process
Fortunately, the majority of the machining undertaken in the production of turbine components utilizes conventional cutting techniques. Therefore, the necessary machine tools are always available,
and the factors to be considered in defining requirements are as follows:
706
Quality Assurance for Replacement & Refurbished Steam Turbine Components
Surface integrity
The quality of the machined surface is dependent upon a number of factors, the majority of which are related to the quality of the
machine tools and cutters used. It is possible that a surface can conform to dimensional, and apparently conform to surface requirements, yet contain deficiencies that have the potential to degrade the
surface quality. Such deficiencies may have little or no effect on
component efficiency, but can introduce structural weaknesses. Typical of these deficiencies are:
Tears and gouges. The tearing or gouging of a surface can introduce microcracking, which (depending upon its location) can be the
initiation site for some form of rupture. These surface discontinuities
can be caused by a number of independent factors, such as tool
sharpness, cutting speed, and the rate at which material is being
removed from the surface.
Tool chatter. To achieve a consistent conforming surface, the
cutting tool must be mounted correctly, securely fixed, and the material to be removed must be removed at the correct rate, in terms of
cutting speed, feed, and depth. These are necessary so that the tool
is free from any form of tool vibration.
Surface burning. Surface overheating can cause carburization of
the surface. This can be caused by the cutting rate being too high or
by there being insufficient cutting fluid.
Built-up edge retention. The cutting action removes material
from the machined part, forming a series of “chips,” which can form
an apparently continuous strip of material. In fact, this strip can contain a series of chips that are fused together by the heat generated by
cutting. It is also common for the chips to build up on the cutting
edge of the tool, as shown in Figure 11.15.1. This edge can break
loose under the effects of the cutting force, and become embedded
in the main material surface.
707
Turbine Steam Path Maintenance and Repair—Volume Two
Such embedded “built-up” edges, if they exist in the final surface, represent an undetectable, but present surface discontinuity.
Fig. 11.15.1—Showing the “Built-up Edge” (BUE)
which detaches periodically and can be buried
in the surface of the component being machined.
Surface finish
The reasons for specifying a particular surface finish on any component, or part of that component can be for one of five reasons:
708
•
To achieve a condition with minimal efficiency loss (see
chapter 6)
•
To minimize the possibility of stress concentration
•
To allow nondestructive examination from the surface
•
To allow the surface to form a joint that is tight against leakage of some fluid or gas through it
Quality Assurance for Replacement & Refurbished Steam Turbine Components
•
To produce a surface that is finer than the engineering specification will normally cost more in terms of machining and
finishing requirements, will add no technical value to the
product, and cannot be considered to improve the “quality”
of the component
Definition and means of quantifying surface finish are covered in
chapter 6.
Machining rates (speeds and feeds). The machined surfaces of a
turbine component can contain many levels of complexity. This is
particularly true of blading, where complex vane forms are defined,
and where precision is of utmost importance. The rate at which these
surfaces have material removed to form these surfaces will influence
their quality, and since, as dimensional changes occur, so will the
relative cutting speed between the material and cutter. These speed
differences therefore must be accommodated in the total machining
definition.
In producing the final or finish surface on any component, it is
possible to use a number of different cutting tools. These tools must
be properly sharpened and suited to the material being machined.
This is necessary to help ensure the chips that are formed by the cutting process are removed, rather than forced into the surface of the
material where they will exist as material discontinuities, capable of
causing stress concentration, and forming a “rough” surface.
Cutting fluids. Fluids, normally liquid, but occasionally gas, are
used to surround the metal cutting edge to achieve certain objectives, including the removal of heat, the lubrication of the surface,
and the washing away of the cut off material. Therefore, fluids must
be selected to achieve these objectives. However, these fluids must
contain no additives or other constituents that are, or could become
corrosive, if not completely removed from the components before
being assembled in the turbine.
709
Turbine Steam Path Maintenance and Repair—Volume Two
REFERENCES
1. Andrew, D.D., and W.P. Sanders. The Concept of Reverse
Engineering, Turbomachinery Maintenance Congress, Berlin,
October, 1991
2. Surface Texture, published by the American Society of
Mechanical Engineers, United Engineering Center, New
York, New York
3. Sanders, W.P. Select your Supplier of Steam Turbine Blades
Judiciously, Power, April, 1984
4. Kaczmarek, J. Principles of Machining, by Cutting, Abrasion
and Erosion, published by Peter Peregrinus Limited, Stevenage, England, 1976
710
Chapter
12
The Manufacture
and Inspection
Requirements of
Steam Turbine Blades
INTRODUCTION
The components within the steam turbine most susceptible to
damage, most often replaced, and with modern repair techniques
among the most often repaired or refurbished are the rotating blades.
Complex stresses are produced in these components during operation, at both steady loads and operating conditions, and possibly
amplified during transient operation. Because of this, and their sensitivity to vibratory type damage, it is important that when these
components are replaced, the design engineering specification for
each is closely followed.
711
Turbine Steam Path Maintenance and Repair—Volume Two
This chapter considers the characteristics of the blades, which
must be considered when they are manufactured, and the important
aspects of their inspection. The most significant characteristics of any
blade row is dependent in part on their operating environment, the
pressure and temperature of the steam, and the type of loads to which
they are subjected. For this reason there are considerable differences
in the type of information required to be checked, and the inspections that must be applied to blades from different rows. The most
important checks for any row are dependent upon the size of the
elements themselves, and where they will be installed in the turbine.
The various manufacturers have developed methods of making
characteristic checks of their blades, and in many instances have
developed not only manufacturing techniques suited to their products, but have designed inspection devices and techniques for proving compliance with design characteristics.
The important characteristics of the blade rows are not always
apparent, and therefore it is necessary to consider other means of
ensuring that the blades comply with design specifications in terms of
material, dimensional requirements (including surface finish) and (for
the longer elements) that their vibration characteristics are acceptable.
RADIAL ALIGNMENT OF
ROTATING BLADES
An important characteristic that must be maintained, and
becomes more critical as the radial height of the vane increases, is
the radial alignment of the blade vane to the root. The problems
associated with misalignment of the vane include the distortion of
the expansion passage formed between the vanes, worsening along
712
The Manufacture and Inspection of Rotating Blades
the radial height towards the tip, and the centrifugal bending stress
induced in the blade as a consequence of such misalignment.
Shown as Figure 12.2.1 is the root and tip section of a long vane,
and the position of the center of gravity of both “G.” Under normal
circumstances, these two positions of “G” are located above each
other and above the center of gravity of the root platform to the
greatest extent possible. There are however, exceptions to this. On
large radial height blades there can be large steam bending forces
developed on the vanes as a consequence of the change of steam
momentum in flowing through the expansion passages between the
vanes. In an effort to reduce the steam bending effect, the vanes can
be given a tilt in both the axial and tangential directions to counter
a portion of the stress induced by steam bending effects.
Fig. 12.2.1—The root and tip section
of a large rotating blade with the
center of gravity ‘G’ above a common point.
713
Turbine Steam Path Maintenance and Repair—Volume Two
Figure 12.2.2 shows the steam forces developed on the vane.
This steam force “Fsb” can be resolved into two components, one
axially “Fas” and one tangentially “Fts.” These forces will induce
stress “fas” and “fts” in the blade vane. These forces and stresses are
transmitted through the vane to the root, and affect both the blade
fastening, and the rotor or wheel to which it is attached.
Fig. 12.2.2—The steam forces ‘Fsb’
developed on the blade, divided
into tangential ‘Ft’ and axial ‘Fa’
components.
Incorrect radial alignment of rotor blades
During manufacture of the turbine blade, it is essential the vane
be produced in correct radial orientation relative to the root form.
There are various manufacturing errors that can cause the design
value of alignment not to be achieved. Failure to maintain correct
radial positioning of the vane relative to the root can result in the
rotor blade, on assembly, being in an incorrect radial alignment
within the steam path. Such misalignment can result in incorrect
pitching and a maldistribution of the steam flow within the blade
row lowering expansion efficiency.
714
The Manufacture and Inspection of Rotating Blades
However, a significantly more hazardous consequence of this
incorrect alignment is that during operation, the centrifugal force of
the vane will attempt to rotate the blade about its root section to correct, and achieve radial alignment.
As discussed previously, many manufacturers calculate blade
stresses due to steam flow, and then take advantage of the centrifugal
bending effect to counter a portion of these steam-bending effects.
Therefore it is obvious that the magnitude of this centrifugal bending
stress can become significant, and if blade vane location is not adequately controlled during manufacture, this can introduce excessive
stresses on the blade, capable of inducing premature failure.
Consider a blade shown in Figure 12.2.3, in which the locus of
the center of gravity of the vane is shown as line “Gr-Gt” from the
root to tip. Here the center of gravity of the vane is coincident with
the center of gravity of the root platform at the point of attachment
to the platform. This locus can be straight or curved, curved being
more consistent with a varying profile vane. Also shown is the radial line “R-R.” This line passes through the root center of gravity at the
root platform top. With this blade, the vane is set forward by an
amount “da” at the center of mass “m” that occurs at a radius “Rx,”
and increases to a tip movement of “ϕa” in the axial direction.
Similarly, the blade vane requires a tangential adjusting tilt of
“dt” at “Rx” which is equivalent to “ϕt” at the tip.
During operation the blade will have steam forces developed on
it due to the change of momentum as the steam flows or expands
across it. This steam momentum force can be resolved into two
components, one of these in the tangential direction, and the other
in the axial. During the design phase the designer calculates these
forces as a function of blade height, and then selectively adjusts the
tilt of the blade in both the axial and tangential directions to counter a portion of these steam forces. In this manner it is possible to
lower the operating stress levels. The calculated tilt “ϕa” and “ϕt” is
715
Turbine Steam Path Maintenance and Repair—Volume Two
normally small, of the order of 2-4°. Therefore it can be seen that
even marginal differences in vane radial orientation can have a significant effect on the blade stresses.
R
da
Mac
da
R
Mas
Gt Steam
flow
m
dt
Mts
Mtc
Gt
Dt
dt
m
f as
fts
Rx
Gr
Gr
Dr
(a)
R
(b)
R
Fig. 12.2.3—The steam and centrifugal bending
moments on a blade.
The bending moments in both the tangential and axial directions
due to the steam bending effect, and the centrifugal bending effect
are shown in Figure 12.2.3.
Mac
Mas
Mtc
Mts
=
=
=
=
Axial moment due to blade centrifugal load
Axial moment due to blade steam load
Tangential moment due to blade centrifugal load
Tangential moment due to blade steam load
It is recommended, especially for longer radial blades, to undertake gauging (an audit) of the pitch and throat at various blade
716
The Manufacture and Inspection of Rotating Blades
heights. Table 12.2.1 shows the results of such an audit on a partial
row of rotating blades. This audit measured 20 passages on a
14.215" vane, the readings being taken 2.0" below the tip. Figure
12.2.4 shows the results of this audit on a partial row. The variation
in the measured data can be seen, and the level of variation established in the ratio “O/P.”
Passage
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
Throat
Height
Pitch
O/P
’α‘
0.671
0.658
0.672
0.674
0.670
0.676
0.674
0.675
0.673
0.668
0.679
0.685
0.672
0.672
0.668
0.670
0.675
0.667
0.665
0.679
0.675
14.045
14.193
14.234
14.214
14.260
14.219
14.188
14.183
14.193
14.219
14.209
14.239
14.209
14.198
14.183
14.193
14.173
14.204
14.234
14.255
14.222
1.754
1.744
1.749
1.759
1.732
1.776
1.753
1.734
1.771
1.733
1.746
1.756
1.744
1.751
1.760
1.723
1.753
1.748
1.746
1.731
1.759
0.3826
0.3773
0.3842
0.3832
0.3868
0.3806
0.3845
0.3893
0.3800
0.3855
0.3889
0.3901
0.3853
0.3838
0.3795
0.3889
0.3851
0.3816
0.3809
0.3923
0.3837
22.49
22.17
22.60
22.53
22.76
22.37
22.61
22.91
22.33
22.67
22.89
22.96
22.66
22.57
22.31
22.88
22.65
22.43
22.39
23.10
22.57
Table 12.2.1—The audit values taken from a 14.215" radial height rotating blade: 2.0"
below the tip section.
Typical tolerances for these parameters in this rotating blade are:
Pitch
“P” +/- 4%
Throat “O” +/- 4%
Ratio
“O/P” +/- 2%
717
Turbine Steam Path Maintenance and Repair—Volume Two
1.76
0.67
1.75
1.74
0.66
1.73
1.72
0.65
O/P - 2" Below Tip
14.215"
1-2
(b)
1.78
1.77
0.395
0.390
6-7
O/P
Sin β 2
11-12
16-17
Pitch - 2" Below Tip.
Inches
Opening
Pitch
0.68
21-22
23.0°
0.385
22.8°
0.380
22.6°
0.375
22.4°
0.370
22.2°
22.0°
Sin β2 - 2" Below Tip
2.0"
(a)
Opening - 2" Below Tip.
Inches
Note: The actual tolerances are normally set in terms of a percentage of the vane radial height, but as such no industry standard exists.
However, exceeding the above values is considered excessive.
(c)
Vane Height.
Inches
14.30
14.25
14.20
Design
Height
14.15
14.10
14.05
14.00
Fig. 12.2.4—The results of an audit of 20 expansion passages in rotating blades.
These readings were taken at their discharge point 2.0" below the tip.
At batch end positions, these values may be relaxed a little, causing a permissible 25-50% increase in the values. However, this
should not be necessary for quality-manufactured blades.
For the stationary blades the tolerances are discussed in chapter 7,
and the results of an audit are shown. The stationary blades, because
stress levels are significantly lower are capable of being adjusted to
some extent, and the effect of adjusting repaired stationary blades is
discussed as well.
718
The Manufacture and Inspection of Rotating Blades
Blade vane tilt
This bending effect can be introduced into the vane by its geometry, and an offset bending moment induced in the vane due to its
tilt from the true radial position “R-R.” This offset can be in the tangential and/or axial direction. When tilt is specified by design, it can
be selected in both the axial and tangential direction to balance a
portion of the steam bending effects considered earlier. However, the
steam bending effects in the axial direction are often sufficiently
small that no attempt is necessary to make a bending balance.
It is common practice for the design process to specify a degree of
tilt in the larger radial height blades. Where this tilt or offset is calculated and produced in a direction that will induce a bending moment
capable of countering a portion of the steam bending moment.
Figure 12.2.3 shows the effects of offset in the blade vane. In this
figure are shown the effects of the total bending moment in the tangential direction “Mts” due to the steam, and those due to the centrifugal effects “Mtc.” Similarly, in the axial direction the bending
moments are “Mas” due to the steam, and “Mac” due to the centrifugal effects.
In the tangential direction, there is a total bending moment acting
on the vane of “Mts-Mtc,” and in the axial direction there is a total
bending moment of “Mas-Mac.” These bending moments cause a stress
in the axial direction of “fas,” and in the tangential direction of “fts.”
If the vane is then manufactured to be given total tilts of the tip
section center of gravity “Gt” by amounts “ϕt” and “ϕa” in the tangential and axial directions respectively, as shown in Figure 12.2.5,
these tilts will produce bending moments in opposition to the steam
and pressure moments, thereby canceling a portion of the stresses
induced in the vane. In this figure the “tilt shift” from “Gr” to “Gt”
is shown (the tip originally being centered over “Gr”), and the tip
movement in the axial and tangential directions “ϕa” and “ϕt” can
be seen.
719
Turbine Steam Path Maintenance and Repair—Volume Two
Fig. 12.2.5—The tip shift due
to axial ‘ϕa’ and tangential
‘ϕt’ tilt.
Consider a blade in which the steam bending moment in the tangential direction is “Mts,” inducing a bending stress at the blade root
of “fts.” In this same direction, the blade can be given a backward tilt
at the tip, giving a bending moment of “Mtc,” which is equivalent to
a fraction of the steam bending moment, the fraction selected being
designated “-dMts,” if the steam moment is taken as “Mts-dMts.”
Unit load
“M ts”
“M-dMts”
“sMts”
“fstr”
Case 1
100
50
0
100
50
0
100
100
100
0
-50
-100
0
-50
-100
Case 2
100
50
0
100
50
0
50
50
50
+50
0
-50
+50
0
-50
Case 3
100
50
0
100
50
0
0
0
0
+100
+50
0
+100
+50
0
Table 12.2.2—Resultant stresses from tilt variations (alternative tangential
displacements ‘dtg’).
720
The Manufacture and Inspection of Rotating Blades
Note: In this table three different cases of tilt levels have been
considered. In the first, the blade is tilted to balance the steam
moment at full load. However, when the unit operates at no or low
load levels there is a high bending stress induced. In the second
case, the tilt is adjusted to balance 50% of the steam moment, and
in this case the bending stress does not exceed 50% of the potential
maximum. In the third case, the blade is given no tilt, therefore the
steam moment is unbalanced. The resulting stress can in this
instance be a maximum when the unit is running at full load.
In making this tangential tilt adjustment, there is a need to modify the tangential correction tilt to account for the pitch or chordal
enlargement, which occurs as a function of radius. This is done as a
ratio of the diameters “Dr” to “Dt” in Figure 12.2.3.
A similar tilt adjustment can be made in the axial direction, if the
stress levels justify such an adjustment. The extent of the adjustment
in either direction represents relatively small adjustments in the
order of 0.050" to 0.015". For this reason it can be seen that achieving the correct radial alignment when mounting blades on a rotor is
critical, in terms of reducing excessive and unknown levels of bending stress. No radial position adjustment is required in the case of
axial tilt.
Center of gravity shift of short blades
When a centrifugal bending effect is produced on short blades,
no effort is made to account for this as the total effect on the stresses is small. However, as shown in Figure 12.2.6 there are blades that
can have a considerable “offset.” This offset is accounted for in the
design phase, normally from experience. But, if the blades are not
mounted securely (packed) to the wheel, and looseness exists, this
will allow the blades to tilt in operation, resulting in the load-bearing surface in the root being loaded heavily at one end, and this can
result in root damage.
721
Turbine Steam Path Maintenance and Repair—Volume Two
da
Gv
Gr
dt
Bending
moment BMv
χ
Fig. 12.2.6—The bending
moment ‘BMv’ caused by
the displacement of the
vane CofG Gv’ above the
root CofG ‘Gr’.
Blade tilt and the effect on tenons
If blades that have a manufacturing (as opposed to a design) with
a tilt occurring randomly, then there will be some difficulty in
assembling any coverbands required on the stage. For short blades,
on which it is not possible to deflect the vane, it will be necessary to
custom cut the tenon holes in the coverband. However, on long
vanes it is possible that efforts will be made to deflect the blades by
some small amount, sufficient to allow the coverbands to be passed
over the rivet heads. While this allows assembly of the rotating row
components that cannot be detected, it will cause an elastic deformation of the vane, and high contact pressures between the tenon
head and hole. This introduces complex stresses, possibly of low
magnitude in both the coverband and tenons.
722
The Manufacture and Inspection of Rotating Blades
In operation this could eventually cause high-cycle fatigue or
fretting, or some other phenomena at the interface that could develop into mechanical failure.
BLADE MANUFACTURING
TECHNIQUES
Among the many complex components of the steam turbine,
blades are a main cause of performance deterioration, in terms of
both structural integrity/mechanical failure, and degradation of their
efficiency. For this reason it is valuable to consider aspects of the
manufacture of these components, as the majority of the technologies employed in their production and final acceptance are also used
in other critical components. Possibly the one process not used at
this time in the production of steam turbine blades is casting.
The remainder of this chapter considers the manufacture and
quality requirements of blades. This is convenient, as it is recognized
that these components contain some of the most demanding requirements in the production of steam turbines.
The designer of steam turbine blading spends a considerable
amount of time, and applies technological expertise, calculating stage
requirements and establishing suitable vane forms and root geometries for the individual stages. At completion of the design phase, a
specification for the blading is prepared, which is intended to ensure
it is manufactured to acceptable standards so it is able to operate for
a sufficient period, and at a performance level consistent with the
requirements of the system into which the turbine is being installed.
The blade design specification can be considered to be an engineering definition of the technical requirements of the product, which
723
Turbine Steam Path Maintenance and Repair—Volume Two
will also define the standards to which it will be manufactured to
achieve the correct quality. The blade specification will establish the following parameters, which must be complied within the final product:
724
•
Material specifications—this includes material from which
the blade is to be manufactured. This portion of the specification will normally identify the material by specification
number, the specification being either a generic type material, or more likely an “in house” material developed by the
manufacturer to meet his or her particular requirements, for
long-term operation, within the local steam environment
•
Dimensional requirements—the dimensional requirements
will be a specification for the form of the vane, its length, and
any profile variation as a function of height. The root form
will also be defined, normally as a standard root profile for
which cutters and other manufacturing capabilities exist
•
Manufacturing tolerances—the tolerances within which the
vane is to be manufactured, the tolerances of the blade root and
any stage hardware required to mount the blades to the rotor
•
Spatial relationships—this is the spatial relationship between
the vane and the root on which it is to be mounted. This will
include tolerances for position on the root platform, and any
radial lean, which will be required to counter steam-bending
stresses
•
Surface finish requirements—the surface finish requirements
can include not only the degree of finish, but in certain
instances the direction of finish
•
Manufacturing techniques—the manufacturing technique
will be used to produce the blade, and can include the basic
metal forming techniques, such as forging or cutting, and in
certain instances, the machine tools to be used to make the
cutting or forming
The Manufacture and Inspection of Rotating Blades
•
Special processes—the special processes are used to improve
the operating characteristics of the blade. This can include
welding or brazing techniques to be used for the attachment
of stage hardware or erosion shields, or processes used to join
the blades together in groups. It can also include processes
such as hardening, coating, or shot peening. In each case the
engineering definition should establish the standards against
which these processes are to be calibrated
•
Nondestructive examination—the non-destructive, and even
destructive tests are used to establish acceptability of the
blades and stage hardware. Such testing techniques can also
be applied to the results of the special processes that are used
on the blades
•
Blade mounting—the blade mounting to the rotor is an
assembly that is an integral part of the total manufacturing
process. Any special requirements for mounting and alignment for all stages will be defined, as will the arrangement
for any special closing requirements
•
Any post assembly requirements—this defines any machining or finishing requirements after the blades are mounted to
the rotor. This will include the tolerances that are to be
achieved to ensure correct position of the blades and coverbands on the completed rotor to achieve axial and radial
clearances and stage lap
Many of these engineering definitions, or quality criteria are set as
standards of the industry, or manufacturing standards of the manufacturer, and are not redefined for each blade or row. In addition, many of
these requirements are established in agreement with the manufacturing engineers responsible for producing the blades. However, in the
final analysis the quality of the blade is a design engineering responsibility, and it is the designer who must adjudicate in the situation of any
nonconformance that arises during the manufacturing process.
725
Turbine Steam Path Maintenance and Repair—Volume Two
THE BLADE MANUFACTURING
PROCESSES
The turbine manufacturer normally obtains the materials used to
manufacture steam blades from the materials supplier in one of several specified forms. The form of the blade material required for any
stage is defined by the design engineer, and selected on the basis of
material suitability, and to a degree on material availability. The
defined material also considers the most economical method of producing the blade. These selections of material form are always made
with the requirements of unit availability in mind. The material
delivery form is defined by the design engineer, who also coordinates with the manufacturing engineer to define the manufacturing
process, or processes that will be use to produce the final form.
There are obviously different requirements in specifying various
blade materials, either as bar stock, precision, or envelope forgings
or other forms, although the materials will have similar mechanical
properties. However, any specification defining the requirements for
replacement blades should identify the preferred material production requirements. The specification should be sufficiently specific
that if bar stock is used in place of forgings, there is no degradation
of the mechanical properties or microstructure of the material itself.
Blades are manufactured using various processes, or basic metal
forming procedures. Principal among these are:
Metal cutting. Blades produced entirely from bar stock by metal
cutting are normally those that are short, and have no change, or relatively small change in vane profile along their length. The bar stock
is cut and shaped to the final form by metal removal alone.
With modern material production techniques, bar stock can normally be produced to provide the same material properties and quality characteristics as forgings. However, for many blades normally
726
The Manufacture and Inspection of Rotating Blades
produced from forgings there is a considerable twist in the profile, and
to produce these blades from bar stock would require the removal of
substantial quantities of material. This can require an extension of
manufacturing time and increased costs on a per blade basis.
For small quantities of the vortex blades, bar stock is often
acceptable and economical, and allows short period deliveries. For
larger quantities, the cost of the forging dies can be offset by both
material and the production costs, and by the time required to
remove the excess material from the bar.
Even for forged blades the dies do wear, requiring maintenance.
This is particularly true of precision forging dies.
Envelope forging. An “envelope” forging is one that requires the
removal of material from all surfaces to produce the final blade form.
This type of material delivery is normally specified for blades with a
large radial length, which to produce from bar stock would require the
removal of considerable quantities of material. Material is removed
from all surfaces of the envelope forging, and it is normal for the forging to provide sufficient material both for the production of test
pieces, and machining location positions on the main body of the
forging.
To produce these blades from forgings requires the production of
dies, which are expensive, and therefore the blades produced by
forging are normally a standard type, in which the same vane and
root form are used. An exhaust stage blade produced from an envelope forging is shown in Figure 12.4.1.
Precision forging. The precision forging does not require material to be removed from the pressure and suction faces of the vane.
However, some manufacturers will undertake some polishing of
these surfaces, but this can be for cosmetic purposes only on large
elements with low Reynolds numbers. The vane inlet and discharge
edges will require the removal of “flash,” which is the excess volume
727
Brown Boveri, Inc.
Turbine Steam Path Maintenance and Repair—Volume Two
Fig. 12.4.1—An exhaust blade produced from an
envelope forging.
728
The Manufacture and Inspection of Rotating Blades
of material beyond that required to complete the blade. This material is “squeezed” out from between the dies.
Note: With proper die cleaning between the production of individual forgings, the forged surface is normally in a fully acceptable
condition. If, however, the dies are not cleaned between the production of each blade, there is the possibility of scabs being formed
on, and embedded into the surface, so they do require cleaning and
polishing after cooling.
This means of production does not require the removal of metal
from the vane surfaces as in envelope forgings. However, there is a
requirement to machine the root form, and possibly tie wire holes or
snubbers.
Electric discharge machining. Electric discharge machining is
used principally for stationary blade rows, and has been employed
for the manufacture of both first stage nozzle boxes and complete
diaphragms. However, this method has been used successfully by at
least one manufacturer to produce solid machined rotating blades in
discrete groups. These are used in control stages, which are
machined so they can be attached to the rotors by pinning. Figure
12.4.2 shows the portion of such a row, with blades being removed
from the rotor, each group of three blades being held in place with
three axial pins.
Vane extrusion. A method of vane production used on some
older designs, and units that are still in operation in many plants, both
utility and industrial, is that of making the blade vane portions from
an extruded section. In these blades the vanes are held apart by
“spacer pieces.” These spacer pieces are tapered at an angle equal to
ψ, where ψ = 360/Zb, and Zb is the number of rotating blades in the
row. This is the number of pitches and includes the closing blocks if
these are used. The geometry of the spacer piece is shown in Figure
12.4.3. Here the top portion of the piece is produced to a height “H”
to control the discharge height of the passage, and the sum of the
729
Turbine Blading Repair, Inc.
Turbine Steam Path Maintenance and Repair—Volume Two
Fig. 12.4.2—Groups of three blades produced by electric
discharge machining. These blades are in integral
groups of three and are attached to the wheel by three
axial pins.
Fig. 12.4.3—The extruded vane and the spacer piece placed
between the vane segments.
730
The Manufacture and Inspection of Rotating Blades
pitches “Pv” and “Pp” is equal to the design pitch between the vanes.
The vanes normally have a surface finish consistent with that of the
extrusion dies, i.e., with these vanes the manufacturer does not normally make any finishing polish onto the pressure and suction faces.
The root forms are selected to allow the form to be machined into the
base of the blade by cutting in the tangential direction.
The spacer pieces have the same root form machined into them,
and are then assembled to the rotor through an access window. On
some stages with a large centrifugal load, the vane is bent or “upset”
under the spacer piece, as shown in Figure 12.4.4. This helps to
secure the vane, and transfer the centrifugal load from the vane “V”
to the spacer piece “P” at a load transfer surface “a-a,” which then
provides a greater load transfer surface to the rotor.
Fig. 12.4.4—The vane “bent foot”
under the spacer piece.
Pinch rolling. Blades produced by pinch rolling are formed in
the hot condition by pinching a billet between rotating rollers. In this
method of manufacture the vane is hot formed and (depending upon
the rollers) can be rolled to the final shape requiring little or no surface polishing to complete the vane. It may be necessary to dress
“flash” from the inlet, and discharge noses after rolling. The vane is
then clamped and the root form is machined onto it.
731
Turbine Steam Path Maintenance and Repair—Volume Two
The requirements of tolerances are such that while precision
forging, pinch rolling, and other non-metal cutting processes can
produce vanes, the root portions of all blades are normally produced, or their final dimensional requirements achieved, by metal
cutting. This is necessary to ensure the vane has the correct spatial
relationship to the root, and the blade vanes achieve the correct radial alignment when mounted to the rotor.
Note: In many blades with a large radial height vane, the vane
is “tilted” in both the tangential and axial directions to counter a portion of the steam bending stresses induced in it. This tilt is selected
so that at a predetermine steam flow the centrifugal bending effect is
cancelled in both directions.
The extent of tilt is relatively small, which implies that if a blade
vane is not aligned correctly to the root for any element there can be
large bending stresses introduced into the blades. Therefore it is
important that the tip pitch is monitored during manufacture and
mounting to minimize this stress.
Because blade forging and rolling processes represent “hot”
working of the material, there is a need to stress relieve after forming the blade, and such heat treatment can cause the vane to “warp”
and/or “twist” to some degree. For this reason, there could be a need
to undertake some level of bending or straightening after the basic
blade has cooled. This adjustment is a cold working process, and is
necessary to ensure the vane lies with the correct radial orientation,
and setting angle. If the vane has to be adjusted cold, it is a good
practice to subject it to some form of heat treatment, after completion of adjustment. This is necessary to ensure its freedom from plastic deformation, and any residual stresses that may have been
induced by the adjustment.
732
The Manufacture and Inspection of Rotating Blades
Basic form production by material deformation
Material deformation is a means of producing the basic form of
the vane. This is normally completed by deforming a portion of the
material to form the vane, and leaving sufficient material to produce
the root from an integral bulb of material. The material forming
processes include forging, extrusion, and pinch rolling. This is a hot
deformation process, and heating is necessary to ensure the blade
material is in a structural condition that will allow it to be worked
into a suitable form, having no residual stresses locked into it.
Of these processes, only the envelope forging requires significant
amounts of material to be removed from the vane after the basic
forming process. The other processes may involve a finishing operation, but this represents only a small amount of material removal,
producing a finer finish, and will not represent a significant dimensional change to the vane.
Forging of the material
The two methods of forging a blade require the use of dies, a male
and female portion, which are forced together over a hot billet. Figure
12.4.5 shows forged blades being removed from the heat treat furnace.
Cutting metal to form the vane
The processes selected to remove the excess stock from the basic
material, either bar stock or envelope forgings, for the production of
the vane, are a function of its form. The selected processes for any
blade depending upon various factors, including:
•
the complexities of the profile, which include the degree of
twist and taper, and the need to produce integrally those discontinuities such as tie wire stubs, or localized thickening in
733
Turbine Steam Path Maintenance and Repair—Volume Two
the region of tie wire holes, possibly tip thinning or thickening, and the possible production of an integral coverband
•
the machine tools and techniques that are available within
the manufacturing plant. In fact, if large numbers of a particular blade are to be produced, it may be economical for a
manufacturer to invest in new facilities to accommodate the
new requirements. Usually the blade is designed so it can be
produced with the machine tools and facilities available
•
the material from which the blades are to be produced. There
are some materials used for blading, particularly the titanium
alloys, which require a different manufacturing technique, cutting tools, or production sequence from those made from steel.
The form of the material, i.e., bar stock or forgings, could also
influence the cutting procedure for various materials
•
the total economics of the production process. The manufacture of blades is expensive when a new blade form is to
be produced. The design and manufacturing engineers,
together with marketing, will examine the anticipated manufacturing scheduleand predict future requirements, and
based on these estimates and the existing facilities, select a
manufacturing technique best suited to overall costs (material and labor) consistent with blade quality
Until the advent of numerically controlled machining techniques, and the development of multi-spindle copying machines,
milling and planing were the major metal cutting processes used for
blade vane production. The application of multi-spindle, multi-axis
copying machines has allowed far more complex designs to be specified and produced. This has also reduced the per-element cost to the
extent that complex vortex profiles can now be produced at competitive prices. This change has allowed the turbine steam path to
achieve higher levels of efficiency at reasonable costs. However,
there are many turbines still in service that utilize constant section
734
Leibstritz
The Manufacture and Inspection of Rotating Blades
Fig. 12.4.5—Precision forged rotating blades being removed from the heat treating
furnace.
blades, and these operate at levels of efficiency that are acceptable
in most respects, and which to redesign the form of the vane is
uneconomical. In addition, there are stages where the cost of the
vortex design cannot be justified. Consider the normal methods of
manufacturing the various forms of blade:
Cylindrical profiles (vanes of constant cross section). There are
various metal cutting techniques used to produce blades with a vane
of constant profile; the techniques used for any particular stage
depend upon the blade vane length. The selected techniques are
also influenced by the stress levels in the blade as determined by the
design process.
735
Turbine Steam Path Maintenance and Repair—Volume Two
An early form of vane was one in which the pressure face was produced, normally by plunge milling from the tip to the root on a piece
of bar stock. Here the vane form is shown in Figure 12.4.6, where the
milling cutter is passed radially along the pressure face, the milling cutter being profiled to achieve the correct form of the pressure face. The
major disadvantage with this form of design is that the center of gravity of the vane is not coincident with, or near the center of gravity of
the root. Therefore, for large radial height blades, high centrifugal
bending stresses can be set up in both the vane and root. As a result,
this method (while economical in terms of manufacturing costs) was
suited only to the smaller blades with lower calculated levels of stress.
Fig. 12.4.6—Plunge milling along the axial length of a piece of bar stock to
produce one face of the cylindrical profile of the blade vane.
Another method of producing the pressure face was to mill across the
vane in the direction of the width, as shown in Figure 12.4.7(a), and the
suction face as shown in Figure 12.4.7(b). In each case the cutter center
locus is shown as “T-T.” Under these circumstances it was normal to position the vane on the root as near coincident with the root platform center
of gravity as possible, as shown in Figure 12.4.8, for a rectangular shaped
platform, where the vane center of gravity “Gv” and the center of gravity
of the root “Gr” are close to coincident. Here the root platform center was
at the center of the rectangle at the position “W1,” “T1,” and the vane
center of gravity was not coincident by the amount “dx,” “dy.”
736
The Manufacture and Inspection of Rotating Blades
Fig. 12.4.7—Milling a cylindrical profile. In (a) cuts are made on the pressure
face, and in (b) on the suction surface.
T1
T2
W1
dx
Gr
Gv
W2
Gr
dy
Gv
Fig. 12.4.8—Vane placement of the “C of G” from
the root platform “C of G”.
There were often limitations to the removal of material from the
pressure face by across width milling if the radius of curvature of the
pressure face was too small. In these cases it was not possible to produce a milling cutter of sufficiently small radius that it could be
mounted onto the spindle of the milling machine. In such situations
it was necessary to pass along the radial length of the vane, and then
make a final cut at the root section to finish the blending radius.
737
Turbine Steam Path Maintenance and Repair—Volume Two
Many early designs employed blades that were mounted to the
rotor using a vane and spacer piece placed between them. This was
a cost-effective method, as the vane material could be produced by
milling, or from extruded stock, cut to length, and then possibly a
tenon produced on the tip. The required taper in the root portion
was machined into the spacer block, as shown in Figure 12.4.3. The
spacer block was also required to have the correct form on its two
tangential faces to match the shape of the vane.
In each case of milling the constant profile suction face in which
an integral root (rather than a spacer piece) was used, the milling
cutter had to be passed in the width direction across the blade material, in a path that would form the required surface contour. This
process is shown in Figure 12.4.7(b).
Costs associated with the production of this type of constant profile blade escalate considerably if it is required to produce an angled
sidewall at either the inner or outer surface, shown as “α2” and “α1”
in Figure 12.4.9, or if it was required that an integral coverband
R
S
S
α1
T
T
α2
R
Fig. 12.4.9—A rotating blade with
an integral coverband and angled
sidewalls at “α1” outer and “α2”
inner.
738
The Manufacture and Inspection of Rotating Blades
should be produced on the blade, at the outer surface with an angled
side wall “α1.” In these cases, it required a separate set-up of the
machine tools to remove the material adjacent to the walls, shown as
triangles “RST” in Figure 12.4.9, and to produce the remaining vane
portion and the inner and outer surface fillet radius. It would also
require some hand polish finishing, and radius blending to complete
the vane.
Fig. 12.4.10—Gang milling a tenon.
The milling techniques used with this form of blade were sufficiently controlled and the vanes did not require extensive polishing
after the vanes were complete. The only possible exception was the
tapered sidewalls, and their point of joining either the root platform
or the integral coverband. With this type of profile it was normal to
produce a tenon by passing a gang of milling cutters across the tip
of the blade, as shown in Figure 12.4.10. Here the milling gang is
shown in (a) and the resulting tenon form in (b). An isometric of the
739
Turbine Steam Path Maintenance and Repair—Volume Two
tenon is shown in (c). Either one or two tenons could be produced
in this manner. The one disadvantage with the production of tenons
by gang miller cutting was that there was no platform around the
tenon, as is achieved with more modern machining techniques.
General Electric
Long blades are often produced by cutting the vane in the radial
direction, as seen in Figure 12.4.11. This method of production will
use either planing or milling cutters to remove material from the
envelope forgings. A master is followed for the shape of the vane and
can achieve close tolerances. The root fillet radii are produced as a
separate operation (shown in Fig. 12.4.12). The blades are finally
hand polished to close tolerances (see Fig. 12.4.13). These methods
are able to produce high quality blades conforming closely to vortex
requirements. However, this complex method of production could
only be justified for long exhaust stage blades where the performance
of the individual stages was critical to the performance of the turbine.
Fig. 12.4.11—Machining the vane portion of long exhaust blades. The cutting direction
is radial.
740
General Electric
The Manufacture and Inspection of Rotating Blades
General Electric
Fig. 12.4.12—Machining the vane/root platform fillet radius.
Fig. 12.4.13—The final hand polishing operation on a large exhaust stage blade.
741
Turbine Steam Path Maintenance and Repair—Volume Two
Vortex profiles (vanes of twisted section). For earlier generation
units to produce vanes that accord with the vortex requirements of a
row, required the use of compromise profiles constructed from arcs
of circles and straight lines. Because of machining limitations, i.e.,
the need to produce the vanes by milling and other similar procedures, the true vortex requirements could not be met. However,
close approximation could be achieved by a combination of axial
length milling and planing. But, because of costs, this method of
manufacture could only be justified for the longer blades, and other
methods such as the “straight generated” profile were used to
approximate vortex requirements for other shorter stages.
The principle of producing the “straight generated” vane is
shown in Figure 12.4.14, where the material billet is placed on the
milling machine table, inclined at an angle “ψ,” and then the milling
cutter is passed over this block on the line “N-N,” using a different
part of the cutter at each radial position, and therefore producing a
varying profile, along the radial height.
The introduction of the multi-spindle profile copying machine,
shown in Figure 12.4.15, made the production of vortex profiles on
blades of varying section at economical costs a reality. In this figure,
five blades are being produced, following the master profile on the
left. Modern high efficiency turbines almost exclusively utilize
blades manufactured by such methods if there is enough change of
Root
Mean
Tip
T
M
ψ
N
R
N
T
M
Root
Mean
Tip
Dr
Profile Milling
Cutter
Dt
Dm
R
Fig. 12.4.14—Passing a formed milling cutter over bar stock material at an angle
“ψ” to produce the straight generated face on the material.
742
General Electric
The Manufacture and Inspection of Rotating Blades
Fig. 12.4.15—The multi-axis milling machine producing a number of
vortex vanes copied from a master on the far left.
section to warrant such manufacture. In some manufacturing facilities blades of even relatively short radial height are produced by this
method, since to use a single manufacturing technique has eliminated the need for several production lines, and the cost differences are
sufficiently low that with the volume involved the twisted vane can
be economically justified.
The cutting process develops forces on the vane, which are not
(because of the amount of material being removed) equal. Figure
12.4.16 shows a plot of these forces as a function of the position
being cut. When the profile cutting is complete, the surface is then
given a final polish. This surface requires hand polishing to establish
the design specified finish. In fact only a small amount of material is
required to be removed, and this can be accomplished relatively
easily. The blade is polished to the form of a “guillotine” or “shutter”
gauge as shown in Figure 2.13.1 of chapter 2.
743
Turbine Steam Path Maintenance and Repair—Volume Two
Fig. 12.4.16—Showing the variation of cutting force around
the vane with profile copy milling.
The forging process
The forging process uses mechanical pressure to work a hot billet into a form that is either a final form, requiring no material be
removed from the vane portion (precision forged), or it requires a
small amount of material removal to produce the final form (envelope forging). In either case the process requires the use of dies to
produce the final form. Figure 12.4.17 shows a large blade being
removed from the press. The dies can be seen on the bed and upper
press of the forge. In the forging process, it is normal to undertake
some preforming, by hammering to an approximate form, where the
initial billet is being extended and partially formed prior to forging.
After preforming, the billets are heated to the forging temperature
and then pressed into their final correct form. After forming the
blade, it is heat treated to achieve the required mechanical properties and relieve stress (see Fig. 12.4.18).
744
Leibstritz
The Manufacture and Inspection of Rotating Blades
Leibstritz
Fig. 12.4.17—A precision forged blade being
removed from the press at completion of the
forming process.
Fig. 12.4.18—Precision forged blades at completion of their initial stress relief.
745
Turbine Steam Path Maintenance and Repair—Volume Two
Leibstritz
After the initial cooling of the precision forging, the “flash” is
removed, as shown in Figure 12.4.19, and the blade is heat-treated.
In Figure 12.4.20 the forgings are being loaded into a vacuum furnace
for stress relief.
Leibstritz
Fig. 12.4.19—Removal of the ‘flash’ after
forging.
Fig. 12.4.20—Blades placed in the vacuum
furnace for stress relief.
746
The Manufacture and Inspection of Rotating Blades
The extrusion process
The extrusion process is not used now to produce high performance blades for large rating units. However, the process can still be
used for smaller, industrial type turbines where the costs available
from selling these turbines does not support the cost of high performance blades, and where the engineering design exists. In these
cases, blades are produced as extruded stock, the extruded pieces
having been formed by being drawn through a die that has the airfoil form. Therefore, these blades are of constant section, and require
a spacer piece between them to achieve the required pitch, as
shown in Figure 12.4.3. The vane pieces are cut to a length that is
the sum of the root depth, the vane radial height, and the height of
any tenons that must be produced integrally.
Pinch rolling
Pinch rolling is a forming process, which utilizes two rollers
between which is forced a heated billet. The rollers have produced
on them the profile of the suction and pressure faces. As the billet
is pulled between the rollers, it is formed into the required airfoil
form. This process is shown in Figure 12.4.21, where the rollers in
vice slots “E-E” capture the root block, and the billet is pulled into
the rollers to start the forming process. Once the rolling has started, the remainder of the billet is pulled through, forming the profile. The centerline of a typical profile is shown in the detail, with
the centerlines from the root position “R-R” to the tip “T-T” stacked
above the center of gravity “G.” The pinch rolling process is shown
in Figure 12.4.22.
The rollers normally require lubrication, and there is judgment
required to ensure the billet does not slip once rolling has started, as
this will cause deformation of the profile form. This process, like forging, will often employ an initial hammer preforming to make the
rolling easier. The rolling is also sometimes undertaken in two stages,
747
Leibstritz
Turbine Steam Path Maintenance and Repair—Volume Two
Fig. 12.4.21—Pinch rolling billets into
profile stock.
Fig. 12.4.22—The concept of ‘pinch rolling’, where a preheated billet
is pulled between rollers, one with the pressure face profile, and the
other with the suction face profile.
748
The Manufacture and Inspection of Rotating Blades
with a final rolling undertaken to achieve a higher quality surface finish. Again this is an older process, and while it can produce highquality blades, it is tending to be replaced by more modern, and ultimately more cost effective methods. At the vane inlet and discharge
positions, the rollers do not meet, and there can be a certain amount
of “flash,” which will require dressing after the blade has cooled.
Those processes that form the blade vane using heat and pressure will normally result in a vane with a degree of twist or warpage.
The vane will therefore require straightening after cooling. This
straightening will be completed after the blade has been formed at
its inlet and discharge edges, and has cooled, but before the root is
machined. It is normal to adjust the vane using cold bend and twist
methods, using a guillotine gauge to ensure the vane is both of the
correct form, and adjusted to the correct setting angle. This is to
ensure the expansion passage formed between the vanes is of the
correct form. Depending upon the material and the extent of the correction required, the blade should be stress relieved after adjustment
to remove the residual stresses that result from plastic deformation.
PROFILE AND
CASCADE TOLERANCES
The vane profile is fundamental to stage performance (efficiency
and reliability), and therefore its form must be controlled within
close tolerances throughout the manufacturing cycle. During the initial manufacture, repair, and replacement of turbine steam path
components there are certain spatial relationships in the axial, tangential and radial directions that must be met, and others that (by
preference) should be met to assist in optimizing the performance of
the unit. The spatial requirements of what must be achieved in terms
749
Turbine Steam Path Maintenance and Repair—Volume Two
of alignment are discussed in chapter 2. To achieve these alignment
criteria, it is necessary that the individual components, particularly
the blades, are produced within design limits to ensure that adequate
adjustment of the components on final assembly is possible. The
individual profiles, and their correct position relative to each other
are essential to achieve acceptable stage efficiency.
During the initial production phase, the turbine supplier will
manufacture and assemble the component parts of the steam path to
ensure their design requirements are achieved within specified tolerances. When a unit is removed from service for a maintenance
outage, it is often necessary to perform some remedial work to correct damage and any deterioration that is found. These remedial
actions should aim to return the components, and their arrangement,
to as close to the original conditions as possible, consistent with preserving the performance of the unit. Therefore, it is necessary for
plant maintenance staff to have sufficient knowledge of the component parts and their assembled requirements, so they can determine
the most appropriate course of action in any repair situation.
Dimensional conformance during the production phase, for both
metal shaping and component assembly, is essential to the satisfactory performance of the unit. Similarly, when repairs are undertaken,
safeguards should be employed to retain this dimensional conformance. The turbine stationary and rotating blades are designed,
manufactured, and assembled, so they are able to interact with other
blades, both within their own row, and with rows that precede and
follow them.
Engineering tolerances are selected by the unit designer of the
component to ensure performance requirements are met and, where
appropriate, components can be disassembled for repair or replacement. Engineering tolerances should also be selected so components
can be interchanged, both within similar units in a station, and
between stations. This engineering requirement also allows for a
minimum of replacement parts to be carried in inventory.
750
The Manufacture and Inspection of Rotating Blades
Tolerances must be evaluated and defined during the design
phase to first achieve the requirements of assembly, and secondly
they must support the performance requirements of the unit. It is not,
from the manufacturer’s perspective, a good policy to make tolerances tighter than necessary or required to achieve these objectives.
To make tolerances tighter adds to the cost of production, and would
do nothing to improve the quality of the product.
Blade profile and cascade quality tolerances
For a blade row to operate with a maximum efficiency, and for
those stresses induced in each blade to be shared and equal, it is
necessary to achieve to concurrent objectives:
•
The individual blades, both stationary and rotating, must be
identical, or as close as possible within the tolerances
defined by design engineering
•
The individual blades within the rows must be spaced so that
the passages formed between them are consistent and form
passages of the type defined by design engineering
Each of these requirements is fundamental to blade quality, and
the manufacturing process selected to meet these requirements must
be suitable, and able to meet requirements consistently.
The vane profile is fundamental to stage performance. Therefore,
its form, finish, and the tolerances applied to its manufacture must
be set and controlled within limits, throughout the entire manufacturing cycle, enough to ensure it meets design requirements.
While the requirements of the individual profiles and the vanes
they form either as a constant section along the radial height, or the
rate at which they modify from root to tip, are of considerable importance in establishing the efficiency of the row, the more critical considerations are those associated with the form of the expansion pas-
751
Turbine Steam Path Maintenance and Repair—Volume Two
sage formed between them. The requirements of the cascade, and
the form of errors were discussed in chapter 2, and will not be considered further. These requirements of positional tolerances, etc. are
applicable to both stationary and rotating rows.
Blade profile definitions and tolerances
Before considering the requirements of the profile, it is necessary to
establish a nomenclature to be used in discussing it. Figure 12.5.1 shows
a single profile that is labeled to indicate its principal characteristics.
A primary concern of any blade is the form of the profiles. If the
profile has any of a number of possible errors, it is unlikely it can fulfill its function entirely satisfactorily. There are certain dimensional
characteristics of the profile that help to define its quality and ability to meet performance requirements. While these were considered
in chapter 2, these requirements are summarized here for completeness. The more significant of these being:
Inlet
nose
β10
Inlet
edge
Pressure
face
B
Suction
face
T
ξ
θο
C
b
Discharge
edge
β20
Discharge
tail
Fig. 12.5.1—The nomenclature and definition of a
profile.
752
The Manufacture and Inspection of Rotating Blades
Inlet nose. The “inlet nose” exists at the point of steam entry to
the blade row. This nose is at the inlet, which is the juncture of two
surfaces, and defines the form of the inlet to the expansion passage.
If the profile has a design specified inlet nose radius “r,” this radius
must be maintained; it must also blend without discontinuity to both
the pressure or suction faces.
The inlet nose of the profile must be shaped to accept the working fluid over a small range of inlet angles while incurring only a
minimum energy loss. This nose must guide, or divert the flow into
the two passages it helps to form. This must be done without causing excessive flow disturbance, turbulence, or separation of the
boundary layer. The suitability of a profile to work over a small range
of steam inlet angles is necessary because steam angles could experience minor change of direction under certain conditions, changes
in steam properties or quantities, or distortion or damage to the discharge area of the previous row.
If there is a distortion of area for steam flow in the previous row,
this will influence the pressure at inlet to that row, modifying the
enthalpy drop across it, and therefore causing a change in the discharge velocity, which will alter the required inlet angle in the row,
into which it is directing its steam flow.
The discharge nose (or tail). The discharge nose, as shown in
Figure 12.5.1, has a defined thickness “b,” which for the rotating elements must be maintained at the design values, as this represents a
region on the profile where stresses both direct and cyclic, are high
during operation. It is particularly necessary near the root section,
where an undersized discharge tail thickness can be a source of
crack initiation in the event of any form of mechanical damage.
On stationary blade profiles it is desirable to keep this tail as thin
as practicable, to minimize the generation of “wakes.” However, this
must be done within the bounds of the bending stresses that are
present. For these stationary profiles, this tail region of the profile is
753
Turbine Steam Path Maintenance and Repair—Volume Two
that portion that experiences material loss if there is solid-particle
erosion. Therefore, there is some justification for making these thicker (increased “b”) in those stages.
Discharge tail. The “discharge tail,” or discharge portion of the
profile, defines the shape of the throat and controls the exhaust area
of the stage. Discharge tail curvature is a characteristic of the profile
that is carefully considered by the designer, and the radius of curvature maintained at a large value, which in conjunction with the tail
thickness, is used to minimize the width of the “wake region” downstream of the passage discharge.
The metal section inlet angle (“β1o”). Figure 12.5.1 shows a
vane section in which a mean inlet angle is shown as “β1o.” This
angle is a function of the profile skeleton line, and is not influenced
by steam flow angles, but is established by the shape of the profile
and the angle at which it is mounted on the blade platform, at the
setting angle “ξ.”
The metal section discharge angle (“β2o”). The metal section
discharge angle is shown as “β2o” in Figure 12.5.1, and is the angle
between the tangent to the skeleton line at the discharge point on the
tail, and a tangential line in the direction of rotation when the profile is at a setting angle “ξ.”
The metal section turning angle (“θo”). The metal section has a
turning angle “θo,” which in terms of the metal section inlet and discharge angles “β1o” and “β2o” provide a metal section turning
angle “θo.”
θo = 180° - (β1o + β2o)°
The radius of curvature of the profile surfaces. Pressure and suction surface radii of curvature should not change abruptly. On the
suction face losses result from sudden reductions in the radius of curvature, as these will promote separation, thereby introducing losses,
since once detached the boundary layer will not reattach. Similarly,
754
The Manufacture and Inspection of Rotating Blades
on the pressure surface a sudden increase in the radius of curvature
could induce losses. For further discussion see chapter 2.
Pressure and suction faces. The profile must be formed to produce the pressure face of one steam passage, and a suction face in
the adjacent one. It must do this, and at the same time be able to
form a passage so the desired degree of reaction or pressure drop
occurs. This is achieved by controlling the throat at any radial location, and therefore the discharge area over the entire discharge of the
blade row.
The “pressure face” is the concave face against which the steam
exerts a positive pressure in being deflected through its turning angle
“θo.” The “suction face” is the convex face of the profile.
Profile chord and thickness. These are the two major characteristics that establish the mechanical strength of the profile, helping
establish its ability to carry load and its natural frequency. In Figure
12.5.1 the maximum thickness is shown as “T,” and the chord as “C”
(not shown). There are different definitions of profile chord.
Profile sectional area and bending modulus. The profile must
possess sufficient mechanical strength to be able to withstand the
forces and loads that are developed on, and within it. These include
those which are the direct, and due to the mass of the vane, and the
mass of other stage elements it must support. The profile must also
be able to withstand alternating bending stress from vibratory loads
and various stimuli induced in the blade row during operation.
The blade must continue to operate under these loads, and the
stresses that are induced during normal and transient operation.
These characteristics are used to define the profile. However, for
purposes of manufacture the designer provides an “envelope of tolerances,” as shown in chapter 2, Figure 2.13.2; within this envelope,
the profile is considered fully acceptable.
755
Turbine Steam Path Maintenance and Repair—Volume Two
Blade cascade definitions and tolerances
It is clear from the previous discussion that producing a profile
within tolerances is not in itself a guarantee the stage will perform as
anticipated. The profile position in relation to adjacent elements
must also be examined. The more important considerations affecting
cascade acceptability are:
Blade pitch, “P.” The pitch “P” between any two profiles (see
Fig. 12.5.2) is a function of the stage diameter “D” and the number
of blades “Zb” in the row, and can be determined from:
Inlet edge
β1
θ
W
ξ
O
β2
Discharge edge
P
Fig. 12.5.2—The definition of a bladeFigure
cascade.
12.5.2
The definition of a blade cascade.
Throat opening “O.” The throat or opening “O” is formed usually at the discharge point of the flow passage (see Fig. 12.5.2), and
is the minimum distance between the pressure and suction faces of
adjacent profiles.
The throat is fundamental to determining the discharge area for
any expansion passage, and in total establishes the discharge area
756
The Manufacture and Inspection of Rotating Blades
from the row, and therefore the pressure at discharge and the
enthalpy drop across the row. For passages of mean throat “Oe,”
radial height “H,” and with “Zb” blade elements in the row, the individual throat area is “a” and the total row discharge area is “A,” both
defined by the following equation:
The steam inlet angle, “β1.” The required steam relative inlet
angle “β1,” is determined from the velocity triangles, and is dependent upon the steam and blade velocities. It also requires a blade profile that will admit the steam without excessive incidence. When
selecting the profile to use in any stage application, the design
process selects, or designs, a profile that has a metal section angle
most nearly suited to meet the requirements of “β1.” The inlet nose
has, for the more modern profiles, a rounded form, helping the steam
to enter the blade passage without undue shock and flow disruption.
The steam discharge angle, “β2.” The discharge angle “β2” is a
function of the shape of the profile tail, and is affected by the ratio of
throat opening to pitch, both of which parameters will vary along the
length of the blade vane. Therefore, it is normal for the discharge
angle to vary also. Manufacturing tolerances must be applied to this
value, and the ratio of throat opening “O,” to pitch “P” to achieve this.
The influence of the discharge tail shape can be seen in Figure
12.5.3. Here the throat “O” is formed on the discharge tail, which
has a degree of curvature. The throat, or minimum opening, occurs
across “g-g.” At that position on the tail, the steam, if it were to separate and flow at the same angle would deviate from the discharge
angle of the tail, at its discharge point by an angle “Γ.” Another significant angle at the discharge point is the metal section angle “β2o,”
which is set at this angle relative to the tangential direction.
757
Turbine Steam Path Maintenance and Repair—Volume Two
Fig. 12.5.3—Details of the discharge tail.
As discussed in chapter 7, the effective design angle “β2” is:
An acceptable angle of deviation “Γ” is dependent upon the
designer. This angle is kept as small as possible as its thickness will
cause an increase in the “wake” thickness, and introduce efficiency
losses in the flow. The larger this angle “Γ” becomes, the greater the
chance of boundary layer separation from the discharge edge, and
the formation of vortices that could be carried forward into the following row. Many manufacturers attempt to minimize this angle,
and set a limit of seven degrees. In doing so, they attempt to maintain a maximum value at the mean diameter. However, it is possible
this value will be exceeded at the tip section of constant profile
blades, where the pitch has increased, and the throat will have
moved to, and be formed on, a smaller radius portion of the profile.
The steam turning angle “θ.” The profiles must be able to divert
the working fluid through the desired turning angle “θ,” and in
changing its momentum have a net thrust or force developed. This
force must be able to be transmitted to the rotor and cause it to
rotate. An equation was provided earlier for the turning angle of the
758
The Manufacture and Inspection of Rotating Blades
metal section “θo.” However, since both the steam inlet angles, due
to the possibility of incidence, and the discharge angle, due to the
form of the discharge tail will be different, the turning angle of the
steam “θ” is given by:
θ = 180° - (β1 + β2)°
For an impulse stage the value of “θ” is usually in the range 110160 degrees. As the degree of reaction increases, this turning angle
reduces, so that for the long last stage blades with a considerable
variation of the degree of reaction along the profile, the turning angle
requirement is changing also. For a large blade, the turning angle
can change from about 160 degrees at the root to 10-20 degrees at
the tip section. Therefore there is a considerable range of shapes of
profiles that can be encountered in any turbine, or stage of a turbine.
For the higher-pressure stage of a unit designed for about 50% reaction in the high and intermediate pressure section, the turning angles
will be in the range 60-80 degrees.
The profile setting angle, “ξ.” The profile setting angle “ξ,”
shown in Figure 12.5.2, is fundamental to the shape of the expansion
passage. By being correctly set, it helps ensure the expansion passage
is the correct form, and the discharge area of the passage, and therefore the discharge flow from it, is at or near the design value.
The expansion passage form. The profile must ensure the passage formed between each pair of surfaces, from adjacent profiles,
forms a passage that is of the form required by the pressure ratio
across the row. The manufacturer will apply a level of tolerances.
With this level being at its extremes, reconvergence should not occur
between adjacent elements. (Reconvergence occurs when after convergence, there is a change to divergence in the passage and then a
further convergence.)
In the pair of profiles shown in Figure 2.14.17, of chapter 2, the
passage effective width decreases from “Oe” at entry to the row, to
759
Turbine Steam Path Maintenance and Repair—Volume Two
“Oe” at discharge. This shows a normal rate of convergence for a
passage designed for subsonic flow.
Inlet and discharge edges. The inlet and discharge edge define
the theoretical extremities of the profile at “inlet to” and “discharge
from” the stage (see Fig. 12.5.2). The manufacturer must specify the
extent to which a blade can be “proud of” or “recessed from,” the
inlet or discharge edge. For larger blade elements, this value (acceptable at any radial height) could be a function of the blade radial
position, or distance from the attachment (root) end of the profile.
Cascade width, “W.” This major dimensional characteristics of
the vane width “W” are dependent for any profile upon the setting
angle. When a vane has been defined as one suitable for application
within the steam path, it is not normal for the setting angle “ξ” to be
changed. However, there are profiles that can be used at various setting angles, and these have been checked to establish that the shape
of the expansion passage is acceptable at each.
It is normal for the manufacturer to have established tolerances
for each of these characteristics above, and then to monitor the manufacturing and assembly processes to ensure they are achieved. In
blade manufacture and assembly, the existence of a nonconforming
condition indicates either the dimensional requirements, as outlined
in this section, have not been achieved, or there is evidence of structural distress in the form of component distortion.
In establishing a maintenance strategy, it is advisable to have
available established tolerances to which the components should be
returned by any repair/refurbishment procedure that is used. These
tolerances should be observed and monitored by the maintenance
engineer during the corrective procedures that are selected as the
result of an evaluation of the initial nonconforming condition.
760
The Manufacture and Inspection of Rotating Blades
Gauging the profile
Because the profile is so fundamental to performance, there is a
need to be able to gauge the final form to ensure it falls within the
design-specified envelope of tolerances. There are various methods
available for making these checks. These include:
The guillotine gauge. The most common method of gauging the
compliance of a blade vane is to use a guillotine gauge. Such a
gauge, shown as Chapter 2 Figure 2.13.1, locates the blade in a radial direction from its root, and then a series of shutters are offered to
the pressure and suction faces of the vane. These gauges have stops,
which allow them to travel to the correct position. Figure 12.5.4
shows a blade in a guillotine gauge before the shutters are closed to
measure for dimensional conformance.
Leibstritz
These gauges are normally produced so that with the guillotines
in their fully closed position, there is a gap between the knife-edge
of the gauge to both sides of the profile, which is equal to the sum
of the plus (+) and minus (-) tolerances, and then a “go/no-go” gauge
is used to establish if the profile tolerances have been met.
Fig. 12.5.4—The vane prior to profile gauging.
761
Turbine Steam Path Maintenance and Repair—Volume Two
Projection methods. Another method of gauging the profile is to
measure and project the form onto a screen. To undertake this method,
it may be necessary to remove a slice of blade from a vane. This shadowgraph, at a suitable magnification, is then transferred to a sheet of
paper or vellum for comparison with the design requirements.
Westinghouse Electric
Eye lashing. Eye lashing is a method of reproducing the vane
profile at any radial location on a piece of paper at a suitable magnification. This consists of preparing a series of arcs from the vane
using a follower to monitor the form of the vane. The process of “eye
lashing” the output of such an operation is shown in Figure 12.5.5,
and the output of such a procedure is shown in Figure 12.5.6 This
drawing is then compared with design requirements to establish if
compliance exists. This method has been essentially superseded by
computer methods.
Fig. 12.5.5—Eyelashing a profile.
762
The Manufacture and Inspection of Rotating Blades
Fig. 12.5.6—The output of an eyelash examination of a vane section.
Computer traces. Modern computer techniques allow a blade
profile to be gauged by coordinate measuring machines, as shown
in Figure 12.5.7. This method allows a profile to be gauged in three
axes, and then compare the manufactured form with the design
specification. This can be a relatively slow process, but is of considerable use in setting up and quantifying an initial cut. As shown in
Figure 12.5.8, the computer can also provide a plot of the profile
shape and its conformance with the design specification.
763
Next Page
Turbine Steam Path Maintenance and Repair—Volume Two
Fig. 12.5.7—A three axis measuring machine
being used to check a blade.
Fig. 12.5.8—The output from a computer measurement of a profile
showing the deviation from the design profile. This is a graphic
comparison between a master and an in-production blade. The
actual data are within the tolerance band of 10µm (0.004").
764
List of Acronyms
LIST OF ACRONYMS
AA
AISI
ASME
BHN
BWR
CLA
Dcsp
EDT
EOH
EPRI
ESV
F
FEA
FofS
FOMIS
GTAW
HAZ
HCF
LCF
HP
HP/IP
I&TP
LCF
LF
LP
NDE
NDT
NOH
NPF
OEM
ppb
psi
QA
QC
RMS
SCC
SHR
SMAW
SPE
SPI
T-G
TIG
TTD
UTS
UT
arithmetic average
American Iron and Steel Institute
American Society of Mechanical Engineers
Brinell hardness number
boiling water reaction
centerline average
direct current straight polarity
enthalpy drop test
equivalent operating hours
Electric Power Research Institute
emergency stop valve
Farenheit
finite element analysis
factor of safety
Fossil Operations and Maintenance Information System
gas tungsten arc welding
heat-affected zone
high-cycle fatigue
low-cycle fatigue
high pressure
high pressure / intermediate pressure
inspection and test plan
low-cycle fatigue
load factor
low pressure
nondestructive examination
nondestructive testing
normal operating hours
nozzle passing frequency
original equipment manufacturer
parts per billion
pounds per square inch
quality assurance
quality control
root mean square
stress corrosion cracking
station heat rate
shielded metal arc welding
solid-particle erosion
solid-partical impact
turbine generator
tungsten inert gas
terminal temperature difference
ultimate tensile stress
ultrasonic testing
ix
Appendix
Thermodynamics
and the Mollier
Enthalpy-Entropy
Diagram for
Water/Steam
INTRODUCTION
Thermodynamics is a science that considers those relationships
existing between thermal energy and work, and the conversion from
one form to another. It does this by considering and evaluating the
behavior of gasses and vapors, and their physical variations under
the action of changes in environmental pressure and temperature.
In engineering terms gas is a substance that has completely evaporated from the liquid state, and is dry, i.e., gas contains none of the
liquid phase substance. By definition a vapor is a gas that contains
815
Turbine Steam Path Maintenance and Repair—Volume Two
some of the liquid phase in suspension; and as such will not obey the
laws of gases, because with the changes in the thermodynamic properties defining the gas phase, some proportion of the liquid phase will
change state and cause a flow of heat from the liquid to gaseous
phases and vise versa. Therefore vapors obey different laws defining
their change of state.
This chapter covers the principles required to apply to steam and
a steam/water mixture as it is expanded in the steam turbine.
Water is a naturally occurring substance. It exists on, below, and
above the surface of the earth in vast quantities, and is readily handled.
It is also a substance very suited for use in a power cycle, and contains
many properties making it suitable for application within these different cycle configurations. Water/steam will accept and reject heat in a
manner making it a suitable fluid for such application. At all levels of
thermodynamic conditions it is non-toxic, and does not break down
easily under the actions of pressure and temperature. However, at high
pressures, and temperatures above 1,200°F steam will dissociate into
hydrogen and oxygen, and can even support combustion within the
steam path.
The study of the flow, or expansion of steam is complicated by the
condensation process, which exists within many fluids as they give up
their thermal energy. This process occurs in the turbine steam path, as
the steam expands through the blade rows to produce mechanical
power and portions of the working fluid are converted back to the liquid phase.
Steam, when heated so that no liquid phase particles are present,
can be considered for all practical purposes to behave as a perfect gas,
and the perfect gas laws can be applied to its expansion when it has
been heated so that no moisture particles remain. However, once
water is formed within this steam, the two-phase flow behaves in a
somewhat different manner, and it becomes necessary to account for
these differences during the design process. This is because of the
816
Thermodynamics & the Mollier Enthalpy-Entropy Diagram for Water/Steam
expansion and transportation of the moisture particles that are
entrained within the steam, and the effects this water will have upon
the steam path components once it is deposited upon, and flows across
their internal surfaces.
In any discussion of the steam turbine and the need for its maintenance and care, it is essential to recognize that it is a thermal machine,
designed specifically for energy conversion from one form to another,
i.e., steam from a fossil or nuclear boiler is released into the turbine at
high levels of potential energy, and this energy causes the turbine rotor
to rotate and either generate electrical power in a generator, or drive
some piece of mechanical equipment.
The internals of the unit are arranged so the energy of the steam is
released in a controlled manner that maximizes the efficiency of the
energy conversion process, and therefore produces a maximum of
power from the minimum amount of steam energy.
For power generation installations, the driven machine is a generator that produces electric power, which can then be utilized in a number of different applications. However, even within power generating
facilities there can be other smaller units used for driving feed pumps.
These are variable speed units and will require certain considerations
that are similar to the main constant speed units, but present other conditions that must be evaluated.
The process of designing the turbine is complex. The engineer
responsible for this work is required to select various blade and other
elements for the unit so that while they can be arranged to optimize
the energy conversion process, the elements can be arranged and
aligned, so the conversion process is as efficient as possible, and the
unit will operate with a high degree of reliability. The amount of coal,
oil, gas, or other fuel consumed per unit of electrical power generated
must be kept to a minimum. The design process therefore will utilize
thermodynamic and aerodynamic principles in order to ensure that the
total energy conversion is achieved as economically as possible.
817
Turbine Steam Path Maintenance and Repair—Volume Two
THE PHYSICAL PROPERTIES
OF WATER/STEAM
There are certain properties or physical characteristics that
define the energy level of steam and/or a mixture of water as it exists
within the turbine. It is important to be aware of their significance,
as they are fundamental to the design process. There are other properties that define important characteristics, as they will apply to
steam flowing through a turbine steam path.
Pressure—”P or p”
Pressure is a measure of the force that is exerted on any surface
by the kinetic action of the steam against that surface. In imperial
units this pressure is measured in pounds per square inch (psi).
An important pressure is that of 1 standard atmosphere, which is
the pressure exerted by the atmosphere on a surface at sea level. This
pressure is equal to 14.7 (14.696) psi. It is also convenient to relate
this atmospheric pressure to the column of mercury (in a Bourdon
gauge). This means of representing pressure is that 14.7 psi will support a column of mercury, which is 30 (29.9) inches at an ambient
temperature of 62°F.
For all practical purposes, it is sufficiently accurate to convert
these quantities, and say that 1 psi is equal to 2 inches of mercury,
which is a convenient means of defining sub-atmospheric pressure
in the power cycle.
Specific volume—”Vs or v”
The specific volume of steam is a measure of the volume that will be
occupied by a given weight of the steam at the local environmental conditions. In imperial units this is in cubic feet per pound (cu ft/lb).
818
Thermodynamics & the Mollier Enthalpy-Entropy Diagram for Water/Steam
The specific volumes of water and saturated steam are determined by experimentation. These results can then be reduced to
empirical formulae, which allow their computation for all values of
pressure and temperature. The normal units of specific density are
lb/cu ft, and density is equal to the reciprocal of specific volume.
For water and saturated steam the specific volume is a known
quantity at any pressure and temperature level. For partially evaporated water, or partially condensed steam, the total volume can be
found from the sum of the volume of the liquid (water) and the
gaseous phases at that pressure.
If the mixture has a dryness fraction “q,” or moisture content of “x.”
q = 1-x
then the total volume of the mixture “Vm” can be found from:
Vm = q. Vs + (1 - q) . Vw
or Vm = (1 - x) . Vs + x . Vw
where:
Vw is the specific volume of 1 lb of water
Vs is the specific volume of 1 lb of dry saturated steam
It can be seen that at “q = 1.0” the water term disappears, and the
specific volume is then equal to “Vs.” In fact, for all practical engineering purposes the water term “(1 - q). Vw” can be ignored, as the
specific volume of water is particularly small compared to the saturated steam term, so its inclusion will add very little to the accuracy
of the value determined in the form of steam engines encountered.
Figure A2.1 shows a curve of the specific volume of dry, saturated steam from 1,100 psia to 100 psia shown as a function of saturation pressure and temperature. A similar curve for water covering the
same pressure range is shown in Figure A2.2. The pressure and temperature values are the saturation values. Similar curves for a pressure
range from 100 psia to 10 psia are shown in Figures A2.3 and A2.4.
819
Turbine Steam Path Maintenance and Repair—Volume Two
Fig. A2.1—The specific volume of steam from 100 psia to 1,100 psia.
Fig. A2.2—The specific volume of water from 100 psia to 1,100 psia.
820
Thermodynamics & the Mollier Enthalpy-Entropy Diagram for Water/Steam
Fig. A2.3—The specific volume of steam from 10 psia to 100 psia.
Fig. A2.4—The specific volume of steam from 10 psia to 100 psia.
821
Turbine Steam Path Maintenance and Repair—Volume Two
For superheated steam specifically, the best method of establishing
an approximate value of the specific volume for any particular degree
of superheat and at a pressure “p” is to apply the general gas equation.
where:
p
Vs
Ts
Vt
Tt
is the pressure of the steam
is the specific volume at saturated conditions
is the saturation temperature
is the specific volume after superheating
is the superheated final temperature
At constant pressure, the previous equation reduces to:
This is a convenient and reasonably accurate method of computing these specific volumes.
Temperature—”T”
Steam temperature is a measure of the degree of hotness or energy level of the steam. In imperial units this is measured in degrees
Fahrenheit (F).
In order to quantify the degree of heat, it is necessary to establish accurate and repeatable scales to which comparisons with bodies of unknown temperature can be compared. Therefore, a temperature scale can be constructed by establishing these two limits, one
upper and one lower. This range is known as the fundamental interval. This fundamental interval is then divided into a number of equal
subdivisions, each division then being known as one degree.
822
Thermodynamics & the Mollier Enthalpy-Entropy Diagram for Water/Steam
In engineering work there are two basic scales available and in
current use:
•
Centigrade scale—this scale takes as its upper and lower
intervals the boiling and freezing point of water, when measured at a pressure of one standard atmosphere. This scale is
divided into 100 subdivisions or degrees
•
Fahrenheit scale—this scale establishes its lower and upper
limits as the freezing point of alcohol, and the temperature of
pig’s blood, both at a pressure of one standard atmosphere.
This scale is divided into 100 subdivisions or degrees. A
more general, and convenient definition of the Fahrenheit
scale is that the fundamental interval between the freezing
and boiling point of water is divided into 180 degrees, with
the freezing point being set at 32°F
Because of the difference in the fundamental intervals, these two
scales are not equal, and 100 degrees Centigrade is equal to 180
Fahrenheit degrees.
In addition the zero point of the Centigrade scale is equivalent to
the 32-degree point on the Fahrenheit scale. Therefore:
Degree F = (Degrees C x 180/100) + 32
Degree C = (Degrees F - 32) x 100/180
Absolute temperatures. Considering a gas that conforms to
Charles Law, the equation for volumetric change for change in temperature is:
Vt = Vo [1+ αt]
where:
Vt is the gas volume at temperature “t”
Vo is the gas volume at absolute zero temperature
αt is the coefficient of increase in volume at constant pressure
823
Turbine Steam Path Maintenance and Repair—Volume Two
The implication of this equation is that as temperature reduces,
the specific volume decreases until at absolute temperature the volume of the mass of gas is zero. It also implies that the lowest temperature possible is the absolute temperature, which is equal to
-273.12 degrees Centigrade, or -460 degrees Fahrenheit.
Temperatures quoted in the absolute scale are normally referred
to as the Kelvin scale in degrees Centigrade, and are designated as
degrees “K,” or degrees Rankine “R” in the Fahrenheit scale.
The internal energy—”E”
The internal energy of a substance is the total energy stored in
that substance by virtue of its thermal and other energy forms. It is
impossible to quantify the total internal energy level in any gas or
vapor. However, from engineering considerations the significance of
internal energy is the change in energy levels that occurs as a consequence of changes to the substance as a result of its giving out or
taking in energy, and doing or having work done upon it. Should a
substance take in energy and do no work, then its level of internal
energy will increase. Similarly if a substance does work and takes in
no thermal energy, then this work is done at the expense of its internal energy, and its level will decrease.
The internal energy of a given quantity of gas is dependent upon
its temperature only, and unless this gas does work in expanding, or
takes in or gives out heat, its internal energy will remain unchanged.
This statement is normally referred to as Joules Law. This internal
energy represents the energy stored in the gas, and can be made
available to perform external work.
The relationships existing between thermal energy, work, and
internal energy are discussed below:
824
Thermodynamics & the Mollier Enthalpy-Entropy Diagram for Water/Steam
The enthalpy of a Gas—”H”
Enthalpy is a relatively simple concept used to define the total
energy of a substance including gasses and vapors, as they exist at
some specific set of thermal conditions. The total heat (enthalpy) of
any substance is the sum of the internal energy “E” and the pressure
energy that exists as a consequence of the potential to do work or
change state. The quantity enthalpy is normally designated by the
symbol “H.”
“J” is the Joules mechanical equivalent of work. In imperial units,
the equivalent of “J” is 1 Btu is equal to 778 ft lbs of work.
When a substance is expanded at constant pressure “p,” and
there are temperature reductions from “T1” to “T2,” its condition or
internal energy is reduced from “E1” to “E2” by the removal of a
quantity of heat. If during this expansion the volume of the gas
changes from “V1” to “V2,” then the work done “W = p (V2-V1),”
which is equal to the quantity of heat removed.
Heat removed = E2 – E1 + p(V2-V1)
This equation indicates that the quality of heat removed is equal
to the pressure work done by the gas in expanding. If the quantity of
heat converted to work during the expansion is given the symbol
“∆H,” then:
That is:
Giving:
825
Turbine Steam Path Maintenance and Repair—Volume Two
This equation indicates that for any expansion, the work done
(neglecting losses that might occur as a consequence of the expansion) is equal to the change in total heat or enthalpy of the substance
from initial to final conditions.
This heat energy is quite distinct from the temperature. Heat is a
measure of the energy that is contained in the steam. The unit of heat
energy is a measure of the amount of heat in any given weight quantity. In imperial units this is British thermal unit per pound (Btu/lb)
where a British thermal unit is the amount of heat energy required to
raise one pound of water one Fahrenheit degree.
The total heat in a body is a quantitative measure of the amount
of thermal energy contained in that body. There are various definitions of heat used in engineering work. Among these, the most commonly used in engineering work are the following:
•
Calorie (Cal)—the calorie is the amount of heat required to
raise one gram (gm) of water one Centigrade degree at one
standard atmosphere
•
British thermal unit (Btu)—The Btu is the amount of heat
required to raise one pound of water one Fahrenheit degree
at one standard atmosphere. A normal manner of seeing this
unit quoted in relation to the discussions of power plants is
in Btu/pound, or abbreviated to Btu/lb, indicating that each
pound of steam at any location within the cycle contains a
certain number of Btus of thermal energy
•
Centigrade heat unit (CHU)—this unit has now become
largely discarded. It is the amount of heat required to raise
one pound of water one Centigrade degree. This unit will not
be considered further
In fact, in these units the amount of heat required is also dependent upon the initial temperature. An international agreement has
been reached in defining both the calorie and the Btu. For the calo-
826
Thermodynamics & the Mollier Enthalpy-Entropy Diagram for Water/Steam
rie, it is defined as the amount of heat required to raise 1 gram of
water from a temperature of 14.5 to 15.5°C. This is known as the
15°C calorie. Similarly, the Btu is defined as the amount of heat
required to raise one pound of water from a temperature of 59.5 to
60.5°F.
The total heat of steam is the sum of those heats required to raise
its temperature or internal energy to the condition at which it exists.
The most convenient manner of defining total heat is to establish the
amount of thermal energy above some arbitrary level. The level chosen for engineering work is the point at which one pound of ice has
been converted to one pound of water at a temperature of 32°F
(492°K).
Therefore, this total heat or enthalpy of the steam includes the
sensible heat required to heat the water to 212°F, the latent heat of
evaporation required to convert the water to dry saturated steam,
plus any superheat used to raise the temperature of the steam to a
condition above the saturation point. These heat quantities all
assume the pressure at which the heat is transferred is held constant
at all times.
Therefore, the enthalpy “H” of steam at some pressure “p” can
be found from:
H = h + L + Hs
where:
H
h
L
Hs
is the total heat of the steam
is the sensible heat required to convert 0°F water to
boiling water at pressure “p”
is the latent heat required to evaporate the water to
dry saturated steam
is the superheat
The total heat “H” of saturated steam at some pressure “p” can
be found from:
827
Turbine Steam Path Maintenance and Repair—Volume Two
H=h+L
For a quantity of water in which only a portion of the total latent
heat has been added, and therefore is only partially evaporated, the
total heat is less than that given by the above equation.
If only a percent of the total latent heat required to evaporate the
water has been supplied, then the total heat of the two-phase mixture will be:
H=h+q.L
H = h + (1-x) . L
where “q” is the dryness fraction, and is defined as the ratio weight
of the dry steam contained in the mixture to the total weight of the
mixture. If “Ws” is the weight of the dry steam, and “Ww” is the
weight of the water, then:
In most work related to the steam turbine, interest is usually
related to the amount of water that has been formed in the steam as
a consequence of its expansion through the steam path, and the
reduced pressure at various locations that result from this expansion.
Such expansion is normally from an initial superheat condition, and
its conversion of thermal potential energy to mechanical work. It is
common to refer to and define this presence of water as the moisture
content of the working fluid and the presence of water is defined and
given the symbol “x” representing a percentage content.
Therefore, if 10% of the saturated steam has condensed, and
exists either in the steam as transported droplets, or has been
deposited upon the internal surfaces of the steam path components,
then the steam is said to have a moisture content of 10%, and the
dryness fraction is 90%.
828
Thermodynamics & the Mollier Enthalpy-Entropy Diagram for Water/Steam
The viscosity of water and steam
The viscosity of any fluid is a measure of the shearing stresses
that exist within that fluid when it is flowing. These stresses will be
present in the fluid whenever there is a difference in velocity from
one stream element to another. Such stresses will also reduce in
magnitude as the differences in velocity diminish.
Consider the viscous flow velocity profile shown in Figure A2.5.
Here the velocity of the fluid (gas or liquid) increases from zero or
stationary at the solid wall to some mean velocity. This velocity is
constant at a value of “C” along any stream tube. The stream filaments are shown as lines parallel to the “x” axis. The derivative
“dV/dy” is a measure of the rate of change of velocity at a distance
from the solid surface at which the velocity is “0.”
y
Velocity
profile
x
Fig. A2.5—Viscous flow velocity profile,
adjacent to a surface.
The shearing stress “τ” that exists from element to element can
be found from:
Where “µ” is defined as the coefficient of viscosity.
The values of kinematic viscosity of water and steam are shown
in Figure A.2.6.
829
Turbine Steam Path Maintenance and Repair—Volume Two
Fig. A2.56—The ‘kinematic viscosity’ of steam and of water.
830
Thermodynamics & the Mollier Enthalpy-Entropy Diagram for Water/Steam
THE GAS EQUATIONS
Superheated steam can be considered a prefect gas, and one that
obeys the laws of such gases. There are also certain physical parameters or characteristics of any gas that must be understood. These are
well defined in engineering terms:
Boyles law
Boyles law states that the volume of a fixed quantity of gas,
maintained at a constant temperature, depends on the pressure to
which it is subjected.
What this means is that for a given quantity of gas (or superheated steam), the pressure it exerts on its surrounding surfaces is
dependent upon the volume that the gas occupies if the temperature
remains constant.
Pressure multiplied by volume is a constant, or:
P x Vs = k
where:
P is the steam pressure
Vs is the gas volume
k is a constant
Charles law
Charles law states that for a fixed quantity of gas heated at a constant pressure, the volume increases by a constant fraction of the volume at 0º for each degree increase in temperature.
This means that for a given quantity of gas (or superheated
steam), the volume occupied at constant pressure is inversely proportional to the temperature. Volume divided by temperature is a
constant, or:
831
Turbine Steam Path Maintenance and Repair—Volume Two
V/T = k
where:
V
T
k
is the gas volume
is the gas temperature
is a constant
The general gas equation
The equations of Boyle and Charles can be combined into a
single equation known as the general gas equation. This equation
relates these properties of pressure, temperature, and specific volume, and can be applied to a gas, and to steam when it is in the
superheated condition. If the suffix “1” applies to an initial condition, and suffix “2” applies to a second set of conditions at completion of some thermal process, then the properties of the gas or steam
before and after their change of state, assuming at both conditions
the gases are still a perfect gas, are related by the equation:
This equation can also be written in the general form:
or: P .V = m . R . T
The mass of gas employed is “m,” and the constant “R” is
known as the general gas constant, and will remain the same for all
conditions.
Note: In applying this general gas equation to any set of conditions, the units must be compatible and the temperature used must
be in the absolute scale.
832
Thermodynamics & the Mollier Enthalpy-Entropy Diagram for Water/Steam
Normal temperature and pressure
The properties of gases as they are considered to exist in a standard atmosphere are always given at standard conditions that in
Fahrenheit units are found from substituting into the General Gas
Equation for one pound of gas:
Temperature
Pressure
Specific volume
32 + 460 = 492 degrees R
14.669 psia
12.39 cu ft/lb
Therefore, applying the general gas equation to one lb of steam
at atmospheric pressure gives:
THE HEATING AND
EXPANSION OF STEAM
As heat is added to or removed from steam there is a change in
its physical characteristics. It is necessary to consider these energy
levels changes, and how they will influence the resulting steam conditions at the end of the process. There are a number of interchange
processes that can affect the steam turbine, and it is necessary to
develop an understanding of each.
The addition of heat to water/steam
In order to understand the factors that affect the design and
dimensioning of the steam path, and also those factors that influence
the unit after it goes into operation, it is necessary to have an under-
833
Turbine Steam Path Maintenance and Repair—Volume Two
standing of the working fluid, which is either steam or a mixture of
steam and water.
The chemical compound H2O can exist in three basic forms or
states—solid, liquid, and gaseous. The solid phase is commonly
known as ice, the liquid phase as water, and the gaseous phase as
steam. This compound will readily change from one of these states
to the other by the addition or removal of heat. The amount of heat
required to achieve these state changes will vary, and is dependent
upon the pressure at which the heat transfer occurs.
The addition of heat and the change of states. If a block of one
pound of ice is at, or near absolute zero, and subjected to atmospheric pressure that is heated, it will expand by some small amount
under the influence of this heat addition. The internal temperature of
this ice will also increase. This process of heating, expansion, and
temperature rise can continue until a temperature of 32°F is reached.
Up to this temperature the ice will remain in the solid form. Then
with the addition of a small amount of heat, the ice will begin to
convert to water at the same temperature. If further heat is added the
ice will melt, converting completely to the liquid phase. During this
addition of heat there will be no increase in the temperature of the
ice/water, which will remain at 32°F (492°K).
The amount of heat required to complete this phase transformation from the solid to liquid states is termed the latent heat of fusion,
and the quantity of heat required per pound of ice to complete this
phase change is dependent upon the pressure at which the process
is being undertaken.
The formation of steam (heat addition to water). If one pound
of water, which was obtained from the addition of the heat of fusion
to the one pound of ice, existing at a temperature of 32°F and a pressure of one atmosphere continues to receive heat, its temperature
will again begin to rise. Therefore, the temperature of the liquid
phase substance (water) will increase, but the state or form of the
834
Thermodynamics & the Mollier Enthalpy-Entropy Diagram for Water/Steam
water will not change. The heat being added is called sensible heat.
Its effect can be measured and quantified by measurement of the liquid phase temperature.
Again a point will be reached when the temperature has risen at
atmospheric pressure, by 180°F to 212°F (672°K). At this point a portion of the substance will be driven off from the surface of the water
as steam. From this heat, or energy level, the addition of further heat
will not increase the water temperature, but will continue to convert
the water (liquid phase) to the gaseous phase steam, at a constant
temperature of 212°F.
The heat that is added to the water, increasing its energy level but
introducing no temperature change, is known as the latent heat of
evaporation. This latent heat is defined as the amount of heat
required to evaporate the water completely forming one pound of
steam at the same temperature and pressure. If the pressure at which
the heat is added changes, so will the quantity of heat required to
complete this phase change.
The latent heat of evaporation. The quantity of heat required to
completely evaporate the one pound of water and convert it to one
pound of steam at the same temperature and pressure is known as
the latent heat of evaporation.
The quantity of heat required to evaporate this one pound of
water is again dependent upon the pressure at which the substances
exist, which is the pressure of evaporation. When the one pound of
water has been completely evaporated, and before any further addition of heat to increase its temperature, the steam is said to be saturated. Saturated steam is by definition dry, i.e., no moisture exists
within it.
Moisture content and dryness fraction. An important parameter
of water, particularly when applied to heat engines such as the steam
turbine, is the dryness fraction, or moisture content of the working
835
Turbine Steam Path Maintenance and Repair—Volume Two
fluid. When water is evaporated, it is conceivable that until high levels of evaporation have occurred, there will be moisture present in
the two distinctly recognizable states of water and steam (liquid and
gaseous). However, when steam expands from a higher to a lower
pressure, giving up its internal energy, the temperature falls until a
point is reached when the enthalpy of the steam is such that moisture must exist in the working fluid. Under these circumstances,
water droplets are normally formed by nucleation, and exist within
the steam, and are transported by it through the steam path. As the
expansion continues and the internal energy level of the two-phase
flow reduces, the steam (gaseous phase) gives up further energy, and
a further portion of the working fluid condenses and converts to the
water (liquid phase).
For the engineer a definition of the moisture content is essential
as this helps define what portion of the latent heat exists in the steam
and is available for conversion to some other form of energy.
Superheated steam. Once the ice has been converted to dry saturated steam, if heat is continued to be added to the one pound of
saturated steam, this further heat cannot change the state, but will
cause an increase in the temperature of the steam. The steam is then
defined as superheated. The degree of superheat is dependent upon
the temperature to which the steam is raised. Again, the amount of
heat required to achieve a certain degree of superheat is dependent
upon the pressure at which the steam exists.
The conversion of ice to superheated steam. The various
processes and the effect on phase changes for the heating of ice to
superheated steam have been discussed earlier. These processes are
shown diagrammatically in Figure A4.1, where the changes of state
and the increases in temperature are shown. The addition of heat
from “A” to “B” raises the temperature of ice to the point where the
836
Thermodynamics & the Mollier Enthalpy-Entropy Diagram for Water/Steam
further addition of heat from “B” to “C” will cause the ice to melt
forming the liquid phase (water). This water will be at the same temperature as the ice at condition “B,” i.e., the latent heat of fusion has
been added causing no increase in temperature.
Fig. A4.1—The effect on phase changes and conditions of adding heat to a quantity
of water.
From “C” to “D” the sensible heat has been added to the water,
raising its temperature to the point where the addition of further heat
will cause evaporation. From “D” to “E” the water is fully evaporated converting the liquid/gaseous phase to dry saturated steam at “E,”
during which process there has been no temperature increase in
either of the phases. This represents the addition of the latent heat of
evaporation. The addition of further heat from “E” to “F” will cause
the steam to become superheated.
The heating and expansion relationships
If steam is heated and expands at constant pressure “P,” the volume of the gas will increase from “V1” to “V2” during this heating
process, then the work done “W” by the steam in expanding is:
W = P . (V2 - V1)
837
Turbine Steam Path Maintenance and Repair—Volume Two
However, if the pressure does not remain constant during the
heating process, but changes, then the work done by the expanding
steam can be found from:
From this definition it can be seen the work done by the steam
during such an expansion, or done on it as it is being compressed, is
represented by the area bounded by the “p-v” expansion curve, and
the upper and lower pressure limits, i.e., the area “ABCD” of the
curve shown as Figure A4.2. This form of diagram is commonly
known as an indicator diagram.
Fig. A4.2—The ‘indicator diagram’.
If the expansion of the steam is at constant temperature, it is an
isothermal expansion, then “p1.V1 =p2.V2,” and the work done in
expanding is equal to:
838
Thermodynamics & the Mollier Enthalpy-Entropy Diagram for Water/Steam
The specific heats of steam
The specific heat of steam is defined as the amount of heat energy required to raise the temperature of unit weight of the substance
through a temperature rise of one degree. The heat that is supplied
to the steam is used to increase its intrinsic energy.
When heat is added to steam and it expands, the work that is
done is given by the previous equations. However, the amount of
work that is done is subject to considerable variation, depending
upon the amount of expansion that occurs, and the consequent
change in pressure levels. In the case of an ideal gas (including
superheated steam), it is most appropriate to consider two specific
heats for different modes of parameter change. These are those
occurring when the gas is heated at constant pressure, “Cp” and
those when the gas is heated at constant volume, “Cv.”
Consider a weight of steam “m” and that its temperature is raised
by the addition of heat from some external source by an amount
“dT;” the heat supplied to it, if its volume is maintained constant is
“m.Cv.dT,” and no work is done. Then if this same steam is heated
at constant pressure, heat equal to “p.δv/J” is added where “δv” is the
increase in volume due to the addition of the heat at constant pressure “P,” causing a further temperature increase of “dT,” and the heat
supplied is equal to “m.Cp.dT.”
Therefore, a consequence of this combined heating process in
two stages is:
839
Turbine Steam Path Maintenance and Repair—Volume Two
This reduces to:
But P.V = m. R.T, therefore:
Therefore:
The heating of steam
Another method of considering this relationship that exists
between the specific heats of steam at constant pressure, and then
at constant volume can be developed as follows. Consider steam
with initial conditions of “p, V1, and T1” and with final conditions
after heating at constant pressure to “p, V2, and T2.” If one pound
of this gas is heated at this constant pressure the work done can be
found from:
W = P.(V1 - V2)
But P.V2 = R.T2, and P.V1 = R.T1
Therefore, substituting these values in the equation for W gives
W = R (T2 - T1)
= R/J[T2 - T1] thermal units
Also H = Cp [T2 – T1]
And E = Cv [T2 – T1)
840
Thermodynamics & the Mollier Enthalpy-Entropy Diagram for Water/Steam
From the law of conservation of energy the following general
equation can be stated:
Total Heat Supplied = External Work Done + Increase in Internal Energy
That is: H = W + E
From which it can be seen that
In its more familiar form:
The ratio of the specific heats
When steam is heated at constant volume it does no work as
there is no expansion, and therefore no change in volume. With this
form of heating the specific heat is known as the specific heat at constant volume, and is given the symbol “Cv.” When gas is heated at
constant pressure it will increase in volume, there is expansion, and
therefore, work will be done. This method of heating gives a higher
value of the specific heat at constant pressure “Cp.”
The specific heat of a gas at constant pressure will be numerically greater than the specific heat at constant volume. This difference is the consequence of the gas having to do work on the internal pressure in expanding.
The ratio of these specific heats is an important consideration in
the study of thermodynamic processes and is given the symbol “γ.”
841
Turbine Steam Path Maintenance and Repair—Volume Two
The expansion of steam
Steam enters the turbine at a high pressure and temperature, and
expands through the steam path to some lower set of conditions. It is
normal for this to occur in a number of series stages, arranged along
the axial length of the unit. In order to examine the requirements of
the steam path it is necessary to consider certain thermal concepts.
Using the general gas equation, curves can be drawn for the variation of steam condition as the steam expands giving up thermal
potential energy; both the pressure and temperature reduce with such
expansion. The thermodynamic properties accompanying an adiabatic expansion are shown in Figure A4.3, which illustrates the change
in pressure, temperature, and specific volume from inlet to discharge
conditions. These curves are plotted to a base of total heat, the initial
steam condition being 900 psia, and 900°F and the exhaust pressure
is 2.0" Hga (1psia). At a pressure of about 78 psia, and an enthalpy of
about 1,183 Btu/lb, the total superheat has been expended and the
conditions pass into the saturated region. Moisture then begins to form
in the parent steam. A curve of moisture content for continued expansion to the exhaust pressure is shown. The specific volume curve is
corrected for the reduction in effective steam quantity.
From these curves it can be seen that with this expansion the
pressure and temperature fall relatively even. However, if the condition curve for the specific volume is examined, there is a dramatic
increase in the volume as the steam approaches exhaust conditions.
From these curves, consider one pound of steam (1 lb) at a pressure
of 900 psia and a temperature of 900°F. This 1 lb will occupy a volume of about 0.85 cubic feet. Then after this 1 lb of steam has
expanded through the steam path to a condenser pressure of 1 psia
(2.0" Hga), the 1 lb will now occupy a volume of 334.1 cu ft/lb. Had
there been no condensation, and with about 20% converted to moisture particles, this volume of steam would be about 267 cu ft/lb. At
this pressure the specific volume of water is 0.0161 cu ft/lb.
842
Thermodynamics & the Mollier Enthalpy-Entropy Diagram for Water/Steam
Fig. A4.3—The variation of steam conditions as a function of enthalpy for an
adiabatic expansion.
However, the moisture content at completion of this adiabatic
expansion is at about 20%, giving it a dryness fraction of 0.790.
Therefore, the total volume (assuming none of the moisture has been
deposited on the sidewalls and steam path components) would be
Vm, given by:
Vm = 334 x 0.80 + 0.0161 x (1.00 - 0.20) = 267.40 cu ft/lb
From these numbers it can be seen that the steam phase occupies 267 cu ft/lb, and the water portion 0.01 cu ft/lb. This demonstrates that there is (in most cases) sufficient justification in ignoring
the contribution of the water to the total volume of the mixture.
In Figure A4.3 an adiabatic expansion that is at constant entropy
was assumed. If steam at this same condition, but with an expansion
efficiency in the range 80% is assumed, then the curves in Figure A4.4
843
Turbine Steam Path Maintenance and Repair—Volume Two
could be drawn for the expansion of the working fluid in the turbine
steam path—these conditions recognizing the inefficiencies, or losses,
which occur in the steam path expansion. Again, from these curves it
can be seen for this non-reheat unit that the pressure falls relatively
evenly as does the temperature. However, if the condition curve for
the specific volume is examined, it can be seen there is again the dramatic increase in volume as the steam nears exhaust conditions.
Fig. A4.4—The variation of steam conditions as a function of enthalpy for
an 80% efficient expansion.
This increase in volumetric flow in the lower pressure regions of
an expansion accounts for the dramatic increase in blade length of
a steam turbine as the steam expands from stage to stage. It also
emphasizes the need to often provide multiple flows in the low-pressure section, so there can be made available sufficient annulus area
844
Thermodynamics & the Mollier Enthalpy-Entropy Diagram for Water/Steam
to allow the steam to exit to the condenser at realistic velocities. If
instead of 2.0" Hg the condenser had produced an exhaust pressure
of 1.0" Hg, then the steam volume would have been 656 cu ft/lb.
From these volumetric values it can be seen that at the exhaust from
the low-pressure section, the volume flow of the steam is very sensitive to pressure. In fact, throughout the low-pressure sections relatively small changes in steam pressure will have a significant impact
on the volume flow, and therefore exhaust velocities.
The work done by steam in expanding
Steam can expand from a higher to a lower pressure in a variety
of ways with different combinations or variations in the defining
parameters of pressure, volume, and temperature. The amount of
work done during these different expansions depends upon the variation of the heat content of the steam from initiation to completion
of the expansion.
The expansion of superheated steam follows closely the laws of
a perfect gas. However, once moisture begins to form, there is a need
to modify the predictions of the final conditions that will be achieved
to account for the effects of this moisture.
Expansion at constant volume. The case of steam expanding at
constant volume is the same as for a perfect gas, but in fact has little
or no application in heat engines, as the purpose of an engine is to
do work, and for any constant volume expansion the work done will
be zero.
Expansion at constant pressure. Heat engines convert thermal
potential energy to work by allowing the pressure and enthalpy level
of the steam to degrade in a controlled manner sufficient to generate
mechanical work. For this reason the constant pressure expansion
has little or no application in heat engines.
845
Turbine Steam Path Maintenance and Repair—Volume Two
General case, an expansion according to the law, pVn = C (the
polytropic process). Consider the expansion of steam from condition “p1,V1” to conditions “p2,V2,” as shown in Figure A4.5. The
total work done “W” during this expansion is equal to the area under
the “pv curve” ABCD, from the following equation:
But P . Vn = P1 . V1n, and
Therefore:
Which can be simplified to:
1
Pressure - ’p’
A
p.v n = C
2
D
B
Volume - ’v’
C
Fig. A4.5—The indicator diagram for an
expansion according to the law pvn = C.
846
Thermodynamics & the Mollier Enthalpy-Entropy Diagram for Water/Steam
An adiabatic expansion, (pVγ = constant). An adiabatic expansion is one that is carried out reversibly, and in which no heat is
allowed to enter or leave the system during the process, i.e., at completion of expansion (or compression) the total heat of the gas is
unchanged. This implies there is no loss in the gas due to turbulence,
eddies, or frictional heating (no internal energy losses), and as a
result, the process must be carried out slowly.
Therefore, any work done during an adiabatic expansion is done
at the expense of the internal energy of the gas. Similarly, during an
adiabatic compression all the work that is done in compressing the
gas is converted to internal energy within that gas.
Applying the law of conservation of energy to an adiabatic
expansion, in which no heat is accepted to or rejected from that
process, gives:
But dH = 0
Therefore, for the adiabatic expansion
For the adiabatic expansion of a perfect gas therefore:
In a perfect gas “dE = Cv.dT,” therefore the expression for the adiabatic expansion of a perfect gas can be written:
But P = RTV
847
Turbine Steam Path Maintenance and Repair—Volume Two
Therefore :
Which on integration gives:
Substituting “Cp - Cv = R/J,” and dividing by “Cv” gives:
Writing:
γ = Cp/Cv
γ Loge V - Loge V + Loge T = A constant
But:
P . V/T = A constant
Therefore:
Loge P + Loge V - Loge T = A constant
Adding these last two equations gives:
Loge P + γ . Loge V = A constant
or P . Vγ = A constant
Therefore, the expression for the work done in an adiabatic
expansion is similar to the general expansion considered previously,
except that the index of expansion is defined as “γ,” which is equal
to “Cp/Cv.”
The change of temperature during an adiabatic expansion or compression can be found from consideration of the two relationships:
848
Thermodynamics & the Mollier Enthalpy-Entropy Diagram for Water/Steam
Combining these two equations gives:
Therefore, the work done “W” in an adiabatic expansion can be
shown to be equal to:
An isothermal expansion (at constant temperature)
An isothermal expansion is one that occurs at constant temperature, and during expansion the gas satisfies the relationship:
P.V = R.T
The work done in such an expansion again can be found from:
In this equation “T” is the temperature at which the expansion
takes place.
849
Turbine Steam Path Maintenance and Repair—Volume Two
During an isothermal expansion or compression there is no
change in the internal energy of the gas since there is no change in
temperature, (the internal energy of the gas is a function only of temperature). Therefore, the heat taken in or given out during an isothermal process is equal to the work done or expended.
Throttling. Throttling or wire drawing is the name given to any
process in which steam passes, or escapes through a constricted
opening, causing a reduction in pressure. In this case, any energy
expended by the movement or expansion of the steam is converted
to velocity energy, which is reconverted to heat energy as the velocity is destroyed.
Consider the leakage through a small gap between two surfaces
shown as Figure A4.6. Here the steam is accelerated through the
opening, and achieves a velocity “Ac” at discharge.
Small
clearance
Ea
A
pa
Vsa
Eb
pb
Cx Vsb
B
Fig. A4.6—The small
gap for throttling.
Figure
6A4.
The small gap for throttlin
g
If the steam conditions on the higher pressure side “A” are “pa”
and “Vsa,” and the steam has an internal energy “Ea,” and on the
lower pressure side “B,” the steam conditions have changed to “pb,”
and “Vsb” with internal energy to “Eb.”
Now work is done on the steam in forcing it to flow from side
“A” to side “B,” and this work is equal to “pa.Vsa/J;” after the steam
has passed through the constriction it does work on the steam on
that side by an amount equal to “pb.Vsb/J.”
850
Thermodynamics & the Mollier Enthalpy-Entropy Diagram for Water/Steam
Equating the energy on the two sides, after the velocity energy
has reconverted to heat energy gives:
Because the sum of the work “P.Vs” term and the internal energy at any point is equal to the enthalpy, then:
Ha = Hb
and therefore the work done in a throttling expansion “W = 0.”
Metastable conditions during expansion
(supersaturation)
In discussing the expansion of a gas, it has been assumed that,
at every condition during this expansion, equilibrium conditions
are maintained, and that as conditions are reached at which moisture is to form, there is an immediate heat transfer from the
gaseous to the liquid phase and that moisture particles appear, and
are suspended in and then transported by the steam as it continues its expansion. In fact, this heat transfer will take a finite period of time to occur, and there is therefore a delay in the formation
of these moisture particles or droplets. Under these conditions the
rapidly expanding steam becomes temporarily supercooled or
supersaturated.
For steam expanding in a turbine within the saturated region, the
total expansion (because of the velocity with which the steam is
moving through the blade rows) is always unstable, and there is
some degree of delay in transferring heat from the gaseous to the liquid phase. Therefore, there is always some supersaturation present
when steam is expanding in the saturated region.
851
Turbine Steam Path Maintenance and Repair—Volume Two
When in the supersaturated state, the density of the vapor is
higher than the equilibrium density of the saturated vapor at the
local environmental pressure. Therefore, this supersaturated condition is not stable, i.e., it cannot correct itself and disappear completely as soon as there has been time for the heat transfer from the
gaseous to liquid phase to occur, because during this time there has
been a further expansion and a further portion of the vapor has not
had time to condense.
There is, in fact, a limit to the degree of supersaturation that can
be achieved in any expansion of steam. The degree of supersaturation depends upon a number of factors, including the chemical purity of the steam itself.
THE ENTROPY OF STEAM
An important thermodynamic concept or property in understanding and quantifying thermodynamic processes and energy conversion
is entropy. Entropy is best, and most simply defined as a relationship
that exists between heat and temperature. To provide a suitable definition of entropy, consider a quantity of steam that is heated
reversibly, or simply has heat added to it without its doing any work.
By definition, in adding a small quantity of thermal heat “δH” to
a quantity of steam, when its temperature is “T,” will cause an
increase in its entropy of “δH/T.” Starting from some suitable zero
point if a quantity of heat “δH” is added to a substance at a temperature “T,” then the change in entropy “δs” is:
852
Thermodynamics & the Mollier Enthalpy-Entropy Diagram for Water/Steam
If any substance is subjected to changes in state in a reversible
manner from condition “1” to condition “2” then the change in
entropy associated with this change is:
Therefore:
It is of value to consider the level of entropy changes that occur
when steam is changing its basic conditions. The following examples illustrate these entropy level changes:
Entropy changes for a reversible cycle in which
the temperature is continually changing.
Shown in Figure A5.1 is the “p-V” diagram for a quantity of steam
cycling through a reversible process, in which the temperature is
continually changing, and returning or recycling to the original condition. To determine the entropy change during each complete cycle,
consider this cycle to consist of a large number of individual, and
small, Carnot cycles, each individual cycle comprising two adiabatic and two isothermal processes. Each adiabatic shown (“a-a,” “b-b,”
“c-c” etc.) and the isothermals (shown as the small lines between the
adiabatic expansions and compressions at the upper and lower conditions “a-b,” “b-c,” “c-d,” etc.) is traversed twice, once in each
direction, canceling each other.
It can be seen that in following the elementary cycles that these
represent, in total, the change of conditions in the base cycle. If
“dH” is the amount of heat transferred in an elementary cycle, at
temperature “T,” it can be seen that the entropy of each individual
elementary cycle is “dH/T,” and for the complete set of elementary
cycles:
853
Pressure
Turbine Steam Path Maintenance and Repair—Volume Two
a
c
e
a
b
e
c
d
f
b
f
d
Volume
Fig. A5.1—A reversible process with constantly changing temperature.
Therefore, it can be seen that for any cycle comprising only
reversible processes in which the original conditions are repeated,
the change in entropy “ds” is zero.
Entropy changes with an adiabatic and
isothermal expansion
The adiabatic expansion occurs without change in the heat content of the steam. Therefore, “dH = 0.”
dh = 0
Therefore: ds = 0
During an adiabatic expansion or compression there is no
change in the entropy level of the gas, and the expansion is “isen-
854
Thermodynamics & the Mollier Enthalpy-Entropy Diagram for Water/Steam
tropic.” An isentropic expansion occurs at constant temperature “T.”
On a temperature-entropy diagram this expansion is represented as
a horizontal line.
Entropy changes when expanding a gas at
constant volume
When a gas is heated and maintained at a constant volume, having an initial temperature “T1,” and a final temperature “T2,” then
the quantity of heat “H” added can be found from:
H = Cv . (T2 - T1)
For a small change in the total heat “dH” there will be a change
in entropy of “Cv.dT.” Therefore, dH/T = Cv.(dT/T), and ds = Cv.
(dT/T). If during this change of state the temperature changes from
“T1” to “T2” then:
Giving:
Entropy changes when expanding a gas at
constant pressure
When a gas is heated at constant pressure, and has an initial temperature “T1,” and a final temperature “T2,” then the quantity of heat
“H” added can be found from:
H = Cp.(T2 – T1)
855
Turbine Steam Path Maintenance and Repair—Volume Two
The change in entropy when heating at constant pressure can be
found in the same manner as used for constant volume. The equation of state reduces to:
dT
ds = Cp . T
Which integrating between the temperature limits “T1” and
“T2”gives:
Reducing to:
T2
s2 - s1 = Cp .loge T1
The cases of steam expanding at constant volume and constant
pressure have no practical application in the steam turbine, as all
expansions within the steam path occur with a change of all three
states—pressure, volume, and temperature. Therefore, it is necessary
to consider the application of all three property changes.
General expression for change in entropy
When steam is heated, or allowed to expand in the steam turbine, there will be a change in the physical conditions, changing
them from an initial state of “P1,” “V1,” and “T1” to a final state of
“P2,” “V2,” and “T2.” Therefore, in the general case the change of
steam conditions can be found from the general equation “H = W +
E” represented by:
dH = dW + dE = T.ds = P.dV/J + m . Cv . dT
By recognizing that steam obeys the law “P.V = R . m . T,” and
that “Cv” remains sensibly constant, the processes can be considered reversible (see next section). Therefore, these variables can be
856
Thermodynamics & the Mollier Enthalpy-Entropy Diagram for Water/Steam
eliminated in turn, and the changes in entropy can be determined in
terms of “V,T,” “T,P,” and “V,P.”
For 1 pound of gas, that is “m = 1.0.”
By differentiating and substituting
“dT = (P.dV + V.dP)/R”
REVERSIBILITY
The concept of reversibility is that the change of state that occurs
during a thermodynamic process of expanding or contracting can be
reversed. During reversal the substance will pass back through all
the phases (conditions) that it passed through during its expansion or
857
Turbine Steam Path Maintenance and Repair—Volume Two
contraction, and at completion of the reversal process the gases will
have the same final state they started with.
The implication of reversibility is that the expansion or compression must occur without any energy being lost due to friction,
and without causing (or introducing) any aerodynamic losses. For
reversibility to exist, the process must normally be slow. During such
a process the gas must remain in mechanical and thermodynamic
equilibrium with its surroundings.
The heat engine. A heat engine is designed to convert thermal
potential energy to work. This conversion is achieved by lowering
the energy level of the supplied fluid, and converting the extracted
energy. To achieve this conversion the inlet energy level is raised to
as high a level as is economically achievable by the combustion of
fuel, or the disintegration of nuclear energy. The lower level is
achieved using whatever methods are available. In the majority of
steam applications, this is by the use of a condenser that produces a
sub-atmospheric pressure to increase the range of available energy.
The efficiency of the heat engine is defined as:
Work Done
Energy Used
Work Done
Engine Efficiency =
Energy Supplied - Energy Rejected
Engine Efficiency =
Reversibility of heat engines. For an engine to be reversible,
each process in its operating cycle must be reversible, which is
obviously not possible, as the combustion of fuel is itself an irreversible process. However, the comparison of a cycle or piece of
equipment to one that is reversible (such as the Carnot cycle) provides a measure against which other engines and cycles can be
compared.
858
Thermodynamics & the Mollier Enthalpy-Entropy Diagram for Water/Steam
STEAM PROPERTIES
AND DIAGRAMMATIC
REPRESENTATION
There are a number of diagrams or charts that can be constructed showing the relationships existing between various parameters,
defining the thermodynamic properties of water and steam. Two of
these, the temperature-entropy (T-s) and the total heat-entropy (H-s)
are the most useful, and therefore the ones most frequently used in
engineering work. However, others can be just as important in the
design process, and are referred to by the engineer to ensure the best
possible results and accuracy are obtained from their calculations.
These two most frequently used diagrams are capable of allowing accurate predictions to be made of the results of many processes, and will allow a diagnosis of many situations that can occur within the cycle and the components that it comprises.
The steam tables
The steam tables are a listing of the various properties of steam at
a range of conditions. They were determined by experimental formulations established from a considerable amount of research into
the individual properties and their variation under certain conditions.
The physical properties of water and steam are determined
experimentally, and have been published as steam tables. There has
been a history of the production of these tables; each generation represents a greater accuracy based on improved experimentation, as
testing equipment and instrumentation becomes more accurate.
Associated with this experimentation is the determination or definition of formulae that describe these steam parameters with greater
accuracy for any specific set of conditions. The most recent, and
859
Turbine Steam Path Maintenance and Repair—Volume Two
those referred to in this work, are those produced by the American
Society of Mechanical Engineers, published in 1967. Table A7.1
shows a portion of the steam property tables.
Table A7.1—A portion of the Steam Tables, (ref. A.1) showing the properties of
saturated steam and water
860
Thermodynamics & the Mollier Enthalpy-Entropy Diagram for Water/Steam
Fig. A7.1—The “temperature - entropy’ diagram for steam/water.
The temperature-entropy (T-s) diagram
This diagram shows the relationship between the gaseous and
liquid phases of water and their variation for different temperatures
861
Turbine Steam Path Maintenance and Repair—Volume Two
to a base of entropy. Locus curves are constructed to represent the
variation or change of entropy from water to steam, from the partially wet to saturated and the superheated condition. The outline of
this diagram for water/steam (from ASME 1967 tables) is shown in
Figure A7.1. On this diagram are certain characteristics that define
the relationships existing between the properties of water, a
water/steam mixture, saturated and superheated steam.
Consider the following characteristics as drawn on the diagrammatic representation of the “T-s” diagram in Figure A7.2:
c
m1 m2
T
Pq
q
u
r
t
v
n1 n2
a
d
s
Fig. A7.2—The ‘T-s’ diagram for water/steam.
The saturated water line. The saturated water line defines the
entropy at the boiling point of water at various pressure levels. If the
entropy of water at its boiling point, when all the “sensible heat” has
been added, were plotted for a number of pressures, a line known as
the saturated water line would be obtained.
This saturated water line is shown in Figure A7.2, as existing
from points “a” to “c.” This water has an enthalpy content equivalent to condition “D” of Figure A4.1. The water contains the total
862
Thermodynamics & the Mollier Enthalpy-Entropy Diagram for Water/Steam
heat required to convert the ice from absolute zero condition “A” of
Figure A4.1 to water at its boiling point condition “D.”
However, as previously stated, for engineering work the reference (or point of zero enthalpy) is shown as energy level “C” of Figure A4.1.
The saturated steam line. Similarly, if the locus of dry saturated
steam were constructed for a number of pressures, the line “c” to “d”
would be obtained. This line then defines the entropy level of the
steam at completion of the addition of latent heat, sufficient to convert all the water to dry saturated steam. This is represented by condition “E” of Figure A4.1, and condition “r” of Figure A7.2 for a
steam pressure “Pq.”
The latent heat. From the discussion of the saturated water and
saturated steam lines, it is clear the horizontal distance between
these two curves represents the change in entropy that occurs as the
latent heat of evaporation is added to the water. This is represented
by the heat addition from “D” to “E” of Figure A4.1, and “q” to “r”
of Figure A7.2.
Consider the line “q - r” on Figure A7.2. This line represents the
heating of water from condition “q” to condition “r,” and the evaporative effect, which occurs during this process. It is clear the addition of latent heat at constant pressure represents or provides no
change in the temperature of the mixture.
The critical point. From an examination of the “T-s” diagram of
Figure A7.2, it is clear there is a convergence of the water and steam
lines, and that these would meet at some condition of temperature,
pressure, and entropy. This meeting or convergence point is shown
as “c.” At this condition it would not require the addition of any
latent heat to convert the saturated water to saturated steam.
The conditions at which this “no latent heat requirement”
occurs, is known as the critical point. At this critical point the steam
863
Turbine Steam Path Maintenance and Repair—Volume Two
has a pressure of 3,208 psia, and the corresponding temperature is
705.4°F. At all fluid pressures lower than the critical pressure the
fluid can exist as a liquid. Above this pressure the fluid can exist in
the gaseous form only.
If water exists below the critical temperature and has a higher
pressure than the saturation pressure (corresponding to that temperature), it is considered to be compressed or sub-cooled water.
The moisture content. In considering the addition of heat at a
constant pressure, raising the enthalpy levels from a condition at
some pressure “Pq” from the saturated water condition “q” to dry
saturated steam condition “r” (Fig. A7.2), the entropy level was
increased and the water was fully evaporated. The increase in
enthalpy due to the addition of heat is the distance “q - r.”
When only a portion of this latent heat had been added, the
entropy level will have been raised to some intermediate condition
represented by point “t.” At this condition, represented by point “t,”
the fluid is a two-phase mixture containing (q - t)% dry saturated
steam, and (t - r)% water. This quantity still requires evaporation to
be converted to dry, saturated steam by the addition of sufficient heat
to increase the entropy level by an amount “t - r.” Therefore, the
moisture content “x” of any partially evaporated mixture can be
found from the ratio:
(q - t)
x =
(q - r)
. 100 %
If a constant moisture level were established for different pressure levels, then the lines connecting this constant content at various
pressure levels would look like the line “c-t-v” in Figure A7.2.
The lines of constant superheat. At completion of the evaporative process at constant pressure, if further heat is added (again at
864
Thermodynamics & the Mollier Enthalpy-Entropy Diagram for Water/Steam
constant pressure), the entropy level will increase. Also, since the
internal energy of the gas will increase by the addition of this heat,
the gas temperature will increase also. This temperature rise is
known as the process of adding superheat. If a curve is drawn from
point “r,” at pressure “Pq” in Figure A7.2, the consequential change
in entropy and temperature (heating at this constant pressure), is the
locus “r - u.” This line indicates the line of superheating at a constant
pressure.
If lines of constant degrees of superheat are constructed for various pressures, these will appear as a series of lines shown as “m n” in Figure A7.2.
The actual “T-s” diagram (Fig. A7.1), is prepared from data contained in the 1967 ASME steam tables. The “T-s” diagram has considerable potential for examining the effects of various processes and
condition changes in a thermal power cycle. However, it has limited application to the actual steam turbine engine, and the Mollier
enthalpy-entropy diagram allows for a more complete analysis of the
effects and influences within the turbine unit itself.
The enthalpy-entropy (Mollier) diagram
To begin to examine operating and other problems arising in the
steam turbine, any investigation is aided considerably by a working
understanding of the Mollier “enthalpy-entropy” diagram for
water/steam. This diagram (with a little interpretation) provides a
clear assessment of what is occurring within the unit as there are
changes in steam conditions, or as various forms of expansion occur.
The Mollier diagram allows an understanding of the phase
change from superheated to saturated steam, and allows a prediction
of where water will form and the possible effects of its presence
within the unit. Steam as a working fluid is ideal for the turbine; in
fact had it not occurred naturally, it is ideal and would probably
865
Turbine Steam Path Maintenance and Repair—Volume Two
Fig. A7.3—The Mollier ‘Enthalpy-Entropy’ diagram for steam.
866
Thermodynamics & the Mollier Enthalpy-Entropy Diagram for Water/Steam
have been invented for this use. Steam lends itself to expansion, can
be easily transported through the steam path, and in its “pure” form
causes no chemical reaction with the materials of construction. The
one unfortunate characteristic of steam is that it forms water, which
can (under the influence of high pressures and at high velocities)
cause various forms of damage and deterioration within the unit.
Within the steam path of the turbine, steam expands, i.e., it passes through the various blade rows, reducing or giving up its pressure,
increasing the volumetric flow, but in the process driving the blades
and generating power to drive the rotor, which in turn can be
applied to any number of mechanical or electrical devices.
The Mollier “enthalpy-entropy” diagram provides the locus of
the various steam parameters as a function of the steam enthalpy to
a base of the entropy of the water/steam fluid. This diagram is of considerable use in identifying the effect of various actions on and within the steam turbine, and it allows analysis of the expansion, leakage, and other factors and limitations that occur within the unit itself.
Figure A7.3 shows the Mollier diagram for steam as determined
from the ASME steam tables. The Mollier diagram is a complex graph
of the variation of main steam parameters, shown as a function of total
heat (enthalpy) and entropy (a function of the heat and temperature).
Adiabatic expansion. If the steam were to expand through the
steam path without incurring any losses due to friction, leakage etc.,
the expansion would be said to be adiabatic or isentropic, i.e., there
would be no change of entropy, and the total process would be
reversible.
Under these conditions the constant entropy expansion would, on
the Mollier diagram, be a vertical line. While the pressure and temperature would fall, and the specific volume would increase, the
expansion would be as efficient as possible. Such an expansion would
be as shown on the representative Mollier diagram in Figure A7.4.
867
Turbine Steam Path Maintenance and Repair—Volume Two
H
Pi
H1
Pd
H2
s1
s
Fig. A7.4—A constant entropy
(isothermal) expansion between
pressures ‘Pi’ and ‘Pd’.
Here the inlet pressure is “Pi,” at an enthalpy of “Hi” and the discharge pressure would be “Pd,” at an enthalpy of “Hd.” The expansion would be at constant entropy “s1.”
Throttling or a non-expansive expansion. If steam expands and
does no work, such as a leakage flow, then the total heat of the steam
would be unchanged as there is no expenditure of the steam’s internal energy. In this case, the expansion on the Mollier diagram would
be a horizontal line; i.e., a line at constant enthalpy. A non-expansive, or “throttling” expansion is shown in Figure A7.5, where the
steam pressure falls from “Pi” at inlet to “Pd” at discharge. During
this expansion the enthalpy remains constant at “H1,” but the
entropy increases to “s2” by an amount “ds.”
This assumes that the enthalpy that is converted to velocity is
recovered at the end of the expansion.
868
Thermodynamics & the Mollier Enthalpy-Entropy Diagram for Water/Steam
A steam path expansion. In the practical steam path the expansion
is not without loss, and the line therefore is not vertical. In fact, the
line is displaced to the right on the Mollier diagram, indicating that as
energy is released in the steam, there are frictional and other losses,
and these tend to increase the entropy of the expanding steam.
H
H1
Pi
A
s1
Pd
ds
B
s
s2
Fig. A7.5—A throttling or constant
entropy expansion between pressures
‘Pi’ and ‘Pd’.
A typical expansion line of steam moving through a steam path
is shown as Figure A7.6. The actual steam conditions of the steam in
the path can be determined from this diagram, as the expansion line
represents a locus of the steam parameter variations. At inlet to the
steam path, the steam is at a pressure “Pi,” enthalpy “H1” and
entropy “s1.” At discharge the pressure has reduced to “Pd.” There
will also be a reduction in the enthalpy to “H2,” and an increase in
the entropy to “s2” as shown by “ds.” This type of expansion and the
form of the expansion locus introduced by these losses will be considered in greater detail later in this appendix.
869
Turbine Steam Path Maintenance and Repair—Volume Two
H
H1
A
Pi
dH
H2
B
C
Pd
ds
s1 s2
Fig. A7.6—The expansion of steam in
the turbine steam path, showing the
enthalpy drop, from pressure ‘Pi’
accompanied by an increase in entropy,
at the same discharge pressure ‘Pd’.
The representation of basic power cycles on the
thermodynamic charts
The steam power cycle is an arrangement of processes that allow
the steam or water/steam mixture to be continually cycled through a
number of series processes in converting energy from one form to
another. The selection and arrangement of these processes are considered in detail later. Also, these processes are conveniently represented on both the “T-s” and “H-s” diagrams.
The Carnot cycle
The principle of Carnot states that no cycle can be more efficient
than one comprising only reversible processes. The Carnot cycle
consists of two adiabatic and two isothermal processes working
between an upper temperature level of “T1” and a lower temperature of “T3.” This cycle is shown on the “pv” diagram Figure A7.7(a)
and on the “T-s” diagram in Figure A7.7(b).
870
Pressure
Thermodynamics & the Mollier Enthalpy-Entropy Diagram for Water/Steam
Isothermals at
constant
temperature
p1,v1,T1
1
2
p2,v2, T1
γ
pv = c
p3,v3, T3
4
p4,v4,T3
3
Volume
Fig. A7.7(a)—The ‘Carnot cycle’ on the ‘pv’ diagram.
T
1
2
4
3
Temperature
T1
T2
s1
Entropy
s2
s
Fig. A7.7(b)—The ‘Carnot cycle’ on the ‘T-s’
diagram.
If “P1, V1, and T1” represent the conditions at the beginning of
the cycle, condition “1” (the highest or inlet conditions), and the gas
is then allowed to expand isothermally to condition “2” (P2, V2, and
T1), the expansion is then completed adiabatically to condition “3”
871
Turbine Steam Path Maintenance and Repair—Volume Two
(P3, V3, and T3). At this condition the gas is then compressed
isothermally to condition “4” (P4, V4, and T3), the compression is
then completed adiabatically to return the condition to “1” (P1, V1,
and T1), and the cycle is complete. The highest gas conditions are
P1, V1, and T1; the lowest conditions are P3, V3, and T3.
The ratio of pressure expansion during the isothermal expansion
1-2 is “re,” and equals P1/P2. Similarly, the ratio of compression “rc”
during compression is P4/P3. For the cycle to close, these two ratios
must be equal, and will be designated as “r.”
The heat supplied = P1.V1 loge r
= R.T1. loge r
The heat rejected
= P2.V2 loge r
= R.T3. loge r
Work Done = Heat Supplied - Heat Rejected
= R.T1. loge r - R. T3. loge r
= R.loge r . (T1 - Te)
Cycle Efficiency =
=
Work Done
Heat Supplied
R.loger. (T1 - T3)
R.logeR . T1
=
T1 -T3
T1
The Rankine cycle
In the Rankine cycle the steam is completely condensed from
condition “3” to condition “4a” of Figure A7.8. From this condition
the water is heated along the saturated water line “4a” to “1.” The
cycle is then repeated.
872
Thermodynamics & the Mollier Enthalpy-Entropy Diagram for Water/Steam
1
2
Temperature
T1
T3
4a
s1
3
Entropy
s2
s
Fig. A7.8—The ‘Rankine cycle’ on the ‘T-s’ diagram.
THE BASIC POWER CYCLES
The “T-s” diagram is the most suitable means for analyzing
power cycles, while the “Mollier diagram” is the most convenient
way of examining the expansion of steam in the turbine. Figure A7.8
shows the basic Rankine cycle in which steam was heated to the
saturation line, and then allowed to expand in the turbine. This is
close to representing the conditions that occur in the nuclear and
geothermal cycles, but is not characteristic of the cycles employed
in most units where superheat is added, and then possibly reheated
after partial expansion. Figure A8.1 shows the more common basic
cycles represented on the “T-s” diagram.
The representation of the power cycles
Figure A8.1(a) shows this simple “Rankine cycle” as described in
Figure A7.8, where steam has its temperature raised from vacuum
873
Turbine Steam Path Maintenance and Repair—Volume Two
saturation temperature “Tc,” condition “A,” to temperature “Ts” (the
saturation temperature at which the steam is to be delivered to the
turbine). At this condition “B,” the steam then has heat added to it at
constant temperature “Ts,” raising its energy level to condition “C.”
The steam is then expanded in the turbine to a pressure with a saturation temperature “Tc,” condition “D” where it is condensed to condition “A.” This basic cycle is complete.
Figure A8.1(b) shows the “T-s” diagram for a cycle, but in this
application, while having the same initial and exhaust pressures,
there is an addition of further heat after the saturated condition “C”
is reached. This heat increases the temperature of the steam, from
“C” to “S.” The steam is superheated to a temperature “Tr.” The
steam is then expanded in the turbine again to condition “D.” However, the effect of superheating the steam before admission to the turbine results in a lower quantity of water in the exhaust. As the steam
expands it crosses the saturation line at condition “N,” at which condition the steam has a temperature “Tn.” Moisture will begin to form
soon after crossing this line (see chapter 3).
Figure A8.1(c) shows the effect of a single reheat. After being
heated to temperature “Tr” in the boiler superheater, and then partially expanded in the turbine to condition “E,” the steam is removed
from the turbine, returned to the boiler, and has its temperature
raised again to “Tr,” condition “R.” The steam is then returned to the
turbine and allowed to expand to condenser pressure and temperature condition “D.” Again, the effect of the reheat can be seen in the
final moisture content at condition “D.”
In the final cycle is shown the effect of double reheat [Fig.
A8.1(d)], where the steam is removed from the turbine for a second
stage of reheating, again completing its expansion to condition “D.”
In this cycle it has been shown that higher initial steam temperatures
and pressures were used.
874
Thermodynamics & the Mollier Enthalpy-Entropy Diagram for Water/Steam
S
Tr
B
Ts
C
B
Ts
C
N
Tc
A
G
D
s1
Tc
A
s2
(b)
(a)
S R1
Tr
S
Tr
B
Ts
Tc
C
R
B
Ts
C
R2
E F
E
A
D
s1
s2
(c)
G
D
s1
s2
Tn
G
Tc
G
A
s1
D
s2
(d)
Fig. A8.1—The basic ‘power cycles’ on the ‘T-s’ diagram. In (a) is shown the basic
cycle; in (b) superheat has been added to the steam; in (c) there is steam reheat;
and in (d) there is double reheat.
When considering the steam conditions within the turbine, the
Mollier diagram is concerned only with the variation of steam parameters from conditions “C” to “D.” Figure A8.2 shows the same cycles
illustrated in Figure A8.1. The enthalpy is shown on these diagrams.
In Figure A8.2(a) is the simple cycle, with the expansion line in
the steam path represented by the expansion line “A-Da.” The
increase in entropy from “s2” to “s2” represents the loss of available
875
Turbine Steam Path Maintenance and Repair—Volume Two
H
H
S
Hi
Pi
Hi
Tr
Ts
Pi
Tn
Hn
Ts
C
N
G Tc
G
Tc
Pe
Pe
He
Da
He
D
Da
D
s
S2’
S2
S2
(a)
s
S2’
(b)
H
Hr 2
H
R
Hr
Hi
S
Pi
Ts
Tr = Ti
R2
R1
Hr1
Hi
Hc2
Hc1
Tr = Ti
S
Pi
E
E
F
Ts
C
C
Tn
Tn
N
N
G
Pe
He
D
He
Tc
G
D
Da
Da
S2
(c)
Tc
S2’
S2
s
S2’
s
(d)
Figure
A8. on the Mollier ‘enthalpyFig. A8.2—The cycles shown in figure
A8.1 but2drawn
The cycles
entropy’
diagram.shown in figure A8.1 but drawn on the Mollier ‘Enthalpy-
energy due to inefficiencies in the steam path. In each of the following figures the entropy “s2” is shown as the condition at the end of
steam expansion. The effect of superheat and reheat is shown in Figures A8.2(b) and (c), and the effects of reduction in the final moisture
content in the cycles (a), (b), and (c), with the same initial pressures
can be seen, as the final steam/water mixture is passed to the condenser from condition “Da.”
876
Thermodynamics & the Mollier Enthalpy-Entropy Diagram for Water/Steam
The actual expansion and mixing in
the steam path
In previous sections three various expansions (as they would be
represented on the Mollier diagram) were considered, and in Figure
A8.2 various cycle configurations were examined. However, the
actual expansion that occurs within the steam turbine is not a simple process, and there are a number of locations where there are
unequal expansions between the same pressures, and other situations where there is a mixing of steam, or removal of water at different energy levels. A brief description of what occurs at these points
will help the engineer responsible for turbines, to make maintenance
and replacement decision a little easier to justify. Consider the following conditions:
The leakage of steam under a stationary blade row. Steam that
leaks past seals and between the stationary blade row and the rotor
will re-enter the main steam flow. This leakage of steam upon reentering the main steam flow has two deteriorating effects on performance.
These are:
•
The steam will have throttled past the blade row and have a
higher energy level. Figure A8.3 shows this total effect (refer to
the Fanno curve, chapter 10, Fig. 10.4.4)
•
The steam in reentering the main steam flow will disrupt the
orderly flow of the steam from the stationary to rotating row
causing some level of turbulence
Figure A8.3 shows the expansion through a single stage. Steam at
inlet to the stage has conditions indicated by subscript “i,” in the axial
gap between the stationary and rotating rows by subscript “q,” and at
discharge from the stage by subscript “e.” If the quantity of steam leaking past the stationary blades is “ms” #/sec, and the main steam flow
through the stationary row is “Ms” #/sec, then a quantity of “ms” #/sec,
with an energy content “Hm” Btu/# will mix with a quantity “Ms” #/sec
with an energy content “Hq” Btu/#, in the axial gap between the rows.
877
Turbine Steam Path Maintenance and Repair—Volume Two
Ti
Hi
ms
Pi
Hm
M
Hq
dHs
Tq
Pq
Te
He
Pe
Fig. A8.3—The effect on the expansion line of
stationary blade row leakage.
This mixture “Ms+ms” of the two flows will have an energy level
between “Hq” and “Hm.” The actual energy level is determined by the
ratio of “ms/Ms.” Normally the quantity “ms” is small compared to
“Ms,” but as leakage quantities increase, so the mean energy level
between the stages worsens.
Therefore this mixing of the two quantities causes a reduction in
the energy that is available to be converted in the stationary blade row.
This now unavailable energy is represented by the enthalpy “dHs.”
The steam then entering the rotating blade row will have an inlet
enthalpy “Hq+dHs.”
The leakage of steam over a rotating blade row. A similar situation
exists with leakage over the rotating blade tip. Figure A8.4 shows the
conditions around a typical rotating blade row. By a similar logic, the
leakage quantity over the rotating blades is “mr” and the main steam
flow is “Mr.” This will again cause a reduction in the energy available
for conversion in the blade row and an enthalpy reduction of “dHr.”
878
Thermodynamics & the Mollier Enthalpy-Entropy Diagram for Water/Steam
Mr
dHs
Hq
Tq
Pq
Mt
mr
He
Te
Pe
dHr
s
Fig. A8.4—The effects of rotating blade row tip
leakage.
dHs
Hq
Pe
∆Ho
∆Ha
Pq
He
ds
s
Sa Sb
Fig. A8.5—Showing the effect of increasing the
entropy of the steam entering the rotating blade row.
879
Turbine Steam Path Maintenance and Repair—Volume Two
Note: If the Mollier diagram in Figure A8.4 is examined, it will be
seen that the apparent energy on the rotating blade has not been
reduced significantly by the reducing in energy levels “dHs” in the
stationary row. There has been a small increase in entropy, from
“dsa” to “dsb,” as shown in Figure A8.5. In this figure, the row still
operates between the pressures “Pq” and “Pe,” with an original available energy of “∆Ho.” When the effect of stationary blade leakage is
taken into account, there is an increase in the entropy level of the
steam from “sa” to “sb” causing an increase of “ds.” However, the
available energy is now “∆Ha.” In fact, because of the reheating
effect this is slightly larger than “∆Ho,” because some of the thermal
energy from the leaking steam has been returned to the working main
steam flow making a minor increase in the energy that is available.
Steam leakage at an intermediate seal location (N2). If turbine
sections (high pressure and reheat) are combined in a single shell,
there are situations where extensive losses can be experienced,
because steam will leak from one section to another, bypassing complete sections of the steam path.
Mh
Mr
v
v
mi
(b)
(a)
(c)
(d)
Fig. A8.6—A combined high pressure and reheat section.
In this design the ‘hot’ sections of both expansions are at
the center of the casing.
880
Thermodynamics & the Mollier Enthalpy-Entropy Diagram for Water/Steam
Consider the form of unit shown in Figure A8.6, which has a
combined high pressure and reheat section in a single casing. With
this design the hot sections are located at the center of the rotor, with
the high-pressure steam entering the unit on plane “a,” and exiting
on plane “b.” The steam reenters the unit, after being reheated on
plane “c” and exits on plane “d.” Therefore, while the hot sections
are located at the center of the rotor, there is a considerable pressure
difference at that location, and steam quantity “mi” will leak along
the seals that exist there.
The effect of this leakage can be seen on the Mollier diagram, Figure A8.7. Here high-pressure section steam, after partial expansion
(normally through the first stage nozzle block) will leak along the
rotor, and enter the reheat section at condition “K,” and at a pressure
“Pr” and with an energy level “Hl,” which is lower than the reheat
section steam enthalpy “Hr” with which it is mixing, either directly
or after the reheat steam has expanded through the first stage stationary row. The point of reentry depends upon the detail of design. This
leakage quantity “mi” mixes with the reheat section steam quantity
“Mr” and lowers the mean reheat section enthalpy. This leakage
causes a reduction of unit output and heat rate for two reasons:
•
The leakage steam quantity “mi” bypasses the high-pressure
section blading, and does no work there
•
The leakage steam quantity “mi” causes some level of turbulence or disturbance in the reheat section, again causing losses
On the positive side, it can be seen that the energy of the leakage steam “mi” does become available, and does some work in the
reheat section, but this is not sufficient to replace that lost in the
high-pressure section.
The removal of moisture. The removal of deposited moisture
from the steam path is discussed in chapter 3, and the effect on the
Mollier diagram shown in Figure 3.6.8. Here again the removal of
881
Turbine Steam Path Maintenance and Repair—Volume Two
moisture causes a small increase in the entropy, and it would appear
from first examination that the energy available would cause a
reduction in output. However, because of the reheat effect as discussed earlier, and the water level is reduced, there is less drag on
the steam from alternately accelerating and retarding the transported
water particles, that the output is increased.
Hr
Pr
Pi
T
Hi
Hl
K
HP
Rht
Hch
Hcr
LP
He
Pe
Fig. A8.7—The expansion lines for the
unit shown in figure A8.6. The effect
of internal leakage can be seen with
steam leaking from the high pressure
section mixing with lower pressure
steam in the reheat section.
882
Thermodynamics & the Mollier Enthalpy-Entropy Diagram for Water/Steam
REFERENCES
A1. ASME Steam Tables, published by the American Society of
Mechanical Engineers, New York, NY, 1967
A2. Yellot, J.I., and C.K. Holland. The Condensation of Flowing
Steam, Condensation in Diverging Nozzles, Engineering,
1937
A3. Gyamarthy, G. Grundlagen einer Theorie der Nassdampfturbine, Juris Verlag, Zurich, 1960
883
INDEX
Index Terms
Links
A
Accept-as-is decision
703-704
Acid washing (boiler)
261
Acronyms
Adiabatic expansion
ix
847-850
entropy
854-855
Mollier diagram
867-868
Adjustment computation
96-121
adjustment effects
106-110
horizontal-joint blades
111-115
mismatch errors
115-120
raw data
vane projection across half joint
Adjustment effects
final discharge area
opening
opening-pitch ratio
re-measurement of passages
steam-discharge angle
Adjustment (stage-discharge area/angle)
854-855
96-106
120-121
106-110
109-110
107
107-108
110
108-109
93-96
material removal
95-96
vane bending
95-96
weld buildup
95-96
Administration (quality-assurance program)
683-687
Advanced planning (reverse engineering)
670-671
This page has been reformatted by Knovel to provide easier navigation.
867-868
Index Terms
Aging
Alignment
rotating blades
specification
tie wires
Alignment (rotating blades)
Links
292
411-412
665
411-412
712-723
722-723
blade-vane tilt
719-723
center of gravity shift
721-722
incorrect
714-718
Alloy-steel materials
643
Assembly/alignment specification
665
Axial-entry blade roots
502-503
Axial error
767-770
Axial-gap increase
263
Axial-pressure deformation
122
Axial rubs
370
corrective actions
462-463
coverband
391-392
rotor
459-463
783-788
391-392
459-463
corrective actions
462-463
location
460-461
B
Base-material buildup
224-230
Base-metal buildup
234-236
Basic-form production
712-723
712-723
blade-tilt effect on tenons
Axial rubs (rotor)
665
733
This page has been reformatted by Knovel to provide easier navigation.
459-463
Index Terms
Links
Basic power cycles
873-882
actual expansion and mixing (steam path)
877-882
representation
873-877
Batching/connection (coverband)
390
Batch stress
413
discontinuity
413
twist
413
Below-shield erosion
239-240
Bend-damage classification (rotor)
469-470
Bending
275
blade
275
rotor
464-479
Bending (blade)
275
Bending (rotor)
464-479
causes
464-466
damage classification
469-470
hardness checks
474-475
permanent
467-468
run-out checks
470-474
straightening options
475-479
stress relief
temporary
479
466-467
Bend straightening (rotor)
476
Between-shield erosion
240
Beyond-shield erosion
239
Blade access
Blade bending
Blade-cascade definitions
blade pitch
expansion-passage form
464-479
295-296
275
756-760
756
759-760
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Blade-cascade definitions (Cont.)
cascade width
760
inlet/discharge edges
760
profile-setting angle
759
steam-discharge angle
steam-inlet angle
757-758
757
steam turning angle
758-759
throat opening
756-757
Blade dressing
275
Blade-fillet radii
517-519
Blade inlet-edge erosion damage/refurbishment
206-241
braze-attached shield
230-234
braze-attached shield/base-metal buildup
234-236
braze-attached shield detachment
208
braze-material cracking
211-212
dressing an eroded surface
238-239
local erosion penetration
208
off-shield erosion repair
239-241
raw-weld deposit
236-238
thermally-hardened inlet edges
212-213
weld-attached inlet edge/base-material buildup
224-230
weld-attached shield cracking
209-210
weld-attaching solid-bar inlet nose
213-224
Blade listing
Blade-manufacturing processes
basic-form production
247
726-749
733
cutting metal
733-744
electric-discharge machining
729-730
envelope forging
727
extrusion process
747
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Blade-manufacturing processes (Cont.)
forging of material
733
forging process
744-746
metal cutting
726-727
pinch rolling
731-732
precision forging
727-729
vane extrusion
729-731
Blade modifying steam-flow distribution
260
Blade moment
243
Blade mounting
725
Blade numbering
243-246
blade moment
243
blade weighing
243
data analysis
Blade opening/throat (O)
243-246
29-30
Blade pitch
756
errors
803-811
Blade-pitch errors (factors)
803-811
803-811
blade-root platform pitch
803-805
blade-vane inclination
806-807
root-wedge angle error
808-811
vane placement on root platform
805-806
vane twist/shape variations
747-749
808
Blade placement
248
Blade-production techniques
724
Blade-profile definitions
752-755
discharge nose/tail
753-754
inlet nose
753
metal-section discharge angle
754
metal-section inlet angle
754
766-770
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Blade-profile definitions (Cont.)
metal-section turning angle
754
pressure/suction faces
755
profile chord/thickness
755
profile sectional area/bending modulus
755
radius of curvature (profile surfaces)
Blade-profile/cascade tolerances
754-755
749-764
blade-cascade definitions
756-760
blade-profile/cascade quality
751-752
blade-profile definitions
752-755
gauging the profile
761-764
Blade-profile/placement errors
766-770
Blade removal
275-277
access
295-296
Blade replacement
501-502
Blade reuse
499-501
Blade-root corrosion
483-487
295-296
fretting
486-487
pitting
483-486
Blade-root entry
502-503
783-796
axial
502-503
783-788
tangential
788-796
Blade-root platform pitch
803-805
Blade-root region (rotor)
516-517
Blade-root steeples/wheel rim
480-497
closing-window damage
496
corrosion
483-487
fretting
486-487
high-cycle fatigue
481-482
impact damage
496-497
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Blade-root steeples/wheel rim (Cont.)
low-cycle fatigue
493-494
pitting
483-486
side-grip damage
488-492
stress-concentration centers
481-482
washing erosion
wheel gouging
Blade-root tolerances
497
494-496
775-802
axial-entry direction
783-788
blade-vane positioning (root platform)
776-783
load-bearing surface radius
796-797
radial-entry direction
797-801
root surface-finish requirements
tangential-entry direction
802
788-796
Blade selection
247-248
Blade-setting error
768-770
Blade trailing-edge erosion
255-257
Blade vanes
213-216
384-385
422-424
719-723
776-783
806-807
crack (tie wires)
422-424
inclination
806-807
positioning (root platform)
776-783
preparation
213-216
shortening
384-385
tilt
719-723
Blade-vane tilt
719-723
effect on tenons
722-723
Blade weighing
243
Blast cleaning
205
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Boiler tube
261-262
materials
261
temperature
262
Borasonic examination (rotor forgings)
Borehole machining
342
542-543
Boyles law
831
Braze-attached shield
208
base-metal buildup
234-236
cleaning
231-232
detachment
joint geometry
post-braze inspection
post-brazing activities
preheating
230-236
208
232-233
234
233-234
232
Braze attachment
208
230-236
shield
208
230-236
Braze-material cracking
Brazing
427-431
211-212
208
211-212
230-236
283-284
388-389
427-431
attachment
208
230-236
427-431
coverband
388-389
material cracking
211-212
post-brazing activities
233-234
Built-up rotor
331-332
Burns (rotor)
459-463
corrective actions
462-463
location
460-461
Butt clearances (gland ring)
Buttering (weld material)
620-621
539
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
C
Carbon rings (gland ring)
619-620
Carnot cycle
870-872
Carrier geometry (gland ring)
622-623
Cascade and profile tolerances (blade)
749-764
blade-cascade definitions
756-760
blade-profile/cascade quality
751-752
blade-profile definitions
752-755
gauging the profile
762-764
width
Casing components
760
13-15
casing-exhaust geometries
171-173
explosion/relief diaphragm
168-169
high-pressure packing heads
170-171
high-pressure/high-temperature sections
162-166
low-pressure/low-temperature sections
166-168
Casing-creep cracking
Casing definition
162-173
184-186
13-15
diffuser at exhaust
15
explosion diaphragm
15
inlet section
15
shaft-end packing head
15
shells
15
Casing distortion
590-592
Casing-exhaust geometries
171-173
Casing operating problems/repair methods
174-198
casing-creep cracking
184-186
casing ovality
186-193
low-cycle fatigue cracking
176-184
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Casing operating problems/repair methods (Cont.)
horizontal joint leakage
197-198
humping/hogging
193-196
predictable phenomena
repair methods
unpredictable phenomena
Casing ovality
175
180-184
176
186-193
correction
190-193
inward and binding
187-188
outward
188-189
Casing repair methods
Casings
components
creep cracking
definition
diffuser at exhaust
190-196
174-198
2
4-5
162-198
590-592
13-15
162-173
184-186
13-15
15
distortion
590-592
exhaust geometries
171-173
explosion diaphragm
15
inlet section
15
operating problems/repair methods
174-198
ovality
186-193
shaft-end packing head
15
shells
13
15
13
15
Casing shells
Cast construction
60-61
Caulked strips (gland/seal strip)
636-637
Center of gravity shift
721-722
Centrifugal stress
641
411
This page has been reformatted by Knovel to provide easier navigation.
13-15
Index Terms
Centrifuged-moisture erosion
Charles law
Chemical deposition
Chemical-vapor deposition
Classes of error (blade profile/placement)
Links
54-56
831-832
36-39
266
766-770
axial
767-770
blade setting
768-770
pitch
767-768
770
205-206
231-232
Cleaning
blast
205
hand
206
shield
Closing-window damage
Coating processes
231-232
496
264-270
chemical vapor
266
diffusion/alloying
265
gas phase
268
overlay
265
pack cementation
266-269
plasma
266-270
physical vapor (electron beam)
Coating protection (corrosion)
Company organization (quality assurance)
Component evaluation/testing (reverse engineering)
266
289-291
684
676-677
Component installation (reverse engineering)
677
Component operating modes (changes)
693
Components (rotating)
309-362
coverbands
313-317
rotor
327-342
single-tie connections
323-326
326
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Components (rotating) (Cont.)
tie wires
317-326
turbine-rotor discs
342-362
Components (stationary)
casing definition
diaphragms
stationary-blade definition
Compressive stress
Computer traces (gauging)
Concentricity/balance machining
Constant-pressure gas expansion (entropy)
Constant-volume gas expansion (entropy)
Continuous connection repair (tie wires)
blade-vane crack
fretting
mid-span crack
Continuous connection (tie wires)
repair
Control-stage massive-particle damage
1-15
13-15
5-12
5-6
411
763-764
542
855-856
855
418-424
422-424
424
419-421
414-416
418-424
418-424
277-280
Copper plugs/backing strips
70-73
Correction/repair
59-87
124-128
132-148
153-162
174-198
239-241
291-303
418-426
462-463
59-87
124-128
133-148
erosion
239-241
291-292
horizontal joint
153-154
702-703
axial rubs (rotor)
462-463
casing
174-198
decision
702-703
diaphragm
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Correction/repair (Cont.)
stationary blade
154-162
tie wires
418-426
Corrective options (diaphragm)
124-128
132-133
141-148
281-291
304
398-399
424
439-454
483-487
diametral change
132-133
dishing
124-128
heat and pressure
124-125
inner web
141-148
machining
125-128
permanence
Corrodents
133
450-454
hydrogen sulphide
453-454
oxygen
452-453
sodium hydroxide
451-452
Corrosion
blade root
483-487
corrodents
450-454
coverband
283-287
cracking
398-399
287
disc/spindle interface
445-450
effects
281-291
fatigue
289
fretting
304
398-399
424
486-487
pitting
287-289
483-486
rotating blades
281-291
304
rotor
439-454
steam throttling
442-443
stress cracking
287
This page has been reformatted by Knovel to provide easier navigation.
442
Index Terms
Links
Corrosion (Cont.)
susceptibility
281-287
tenons
283-287
tie wires
Corrosion (blade root)
424
483-487
fretting
486-487
pitting
483-486
Corrosion (coverband)
corrosive attack
283-287
398-399
tenons
283-287
Corrosion effects
398-399
398
fretting
Corrosion cracking
398-399
398-399
287
281-291
coating protection
289-291
regions susceptible to damage
281-287
rotating-blade corrosion forms
287-289
Corrosion fatigue
289
Corrosion (fretting)
304
398-399
486-487
tie wires
424
Corrosion pitting
287-289
Corrosion (rotor)
439-454
blade-tip seal leakage
483-486
444
corrodents
450-454
diaphragm leakage
443-444
disc/spindle interface
445-450
hydrogen sulphide
453-454
oxygen
452-453
rotating-blade rows
442
shaft-end sealing positions
445
This page has been reformatted by Knovel to provide easier navigation.
424
Index Terms
Links
Corrosion (rotor) (Cont.)
sodium hydroxide
451-452
steam throttling
442-443
Coverband damage/repair/refurbishment
axial rubs
batching/connection of groups
283-287
391-392
390
blade-vane shortening
384-385
brazing
388-389
corrosion
283-287
impact damage
363-371
over-riveting of tenons
395-396
refurbishment techniques
396-397
riveting
402-410
screw attachment
382-383
surface rubs
385-388
tenon and hole requirements
400-401
tenon failures
362-410
398-399
395
tenon-hole form
393-394
400-401
tenons
371-382
393-396
400-410
Coverbands
283-287
313-317
323-326
362-410
attached
314
batching with tie wires
323-326
cracks at hole
370-371
damage/repair/refurbishment
283-287
forms
314-317
functions
313-314
integral
314
multi-layer
315
types/shapes
362-410
315-317
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Crack filling
529-530
Crack grinding
504-505
Cracking (corrosion)
Cracking (cross-section)
Creep deflection
Critical point (temperature-entropy)
Cross-section irregularities (vane)
Cutting metal
cutting forces
287
535-537
122
863-864
32-35
726-727
735-741
vortex profiles
742-743
Cylindrical profiles
733-744
743
cylindrical profiles
Cutting process (machining)
145-147
706
735-741
D
Damage and refurbishment (rotating blades)
201-307
SEE ALSO Rotating blades
(damage and refurbishment)
Damage and refurbishment (rotating components)
309-548
SEE ALSO Rotating components
(damage and refurbishment)
Damage and refurbishment (stationary components)
1-199
SEE ALSO Stationary components
(damage and refurbishment)
Damage examination (weld repair)
531-538
axial-surface rebuild
531
cross-section cracks
535-537
stub shaft
537-538
wheel rebuild
531-535
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Damage, deterioration, and failure
mechanisms (overview)
x-xxiv
aging
xiv
external factors
xv
failure definition
xvi
leakage (seal/gland)
xviii
maintenance actions
xxi-xxiv
maintenance options
xx-xxiv
maintenance problems
xiv
operating changes
xiv
quality and inspection
repair/refurbishment options
xviii
xix-xx
steam environment
xv
283-287
root-fastening area
281-282
tie-wire hole in vane
282-283
Data analysis (blade numbering)
243-246
individual opening-discharge areas
opening (O)
pitch (P)
xvii-xviii
281-287
coverband at tenons
Data (raw)
xvii
xviii-xvix
situation evaluation
Damage susceptibility (corrosion)
xvii
96-106
103-106
97-99
98
100
106
radial height
100-103
ratio O/P
103-104
Deep-weld repair (rotor cracks)
cracks
106
508-527
509
522-527
stub attachment crack
508-509
522
surfaces
508-514
This page has been reformatted by Knovel to provide easier navigation.
103-104
Index Terms
Links
Deep-weld repair (rotor cracks) (Cont.)
weld attachment of forgings
522
524-527
wheel
508
514-522
Deflection
121-122
145-147
creep
122
145-147
elastic
121
Design and delivery assumptions
xiii
Design-review meeting minutes
693
Design specification
xiii
assembly/alignment
665
material
664
nondestructive tests
665
physical dimensions
664
special processes
665
surface finish
665
Deterioration (tie wires)
426
Diametral change (diaphragm)
129-133
corrective options
132-133
inner web
147-148
Diaphragm inner-web correction
145-147
diametral changes
147-148
dishing
142-145
Diaphragm leakage
443-444
diaphragm vane
147-148
141-148
creep deflection
Diaphragm repair
663-665
59-87
124-128
59-87
inner web
141-148
sidewall
133-141
thermal distortion
124-128
This page has been reformatted by Knovel to provide easier navigation.
133-148
Index Terms
Diaphragms
explosion
inner ring/web
leakage
nozzle plate/box
outer ring
repair
stationary blades
thermal distortion
vane
Diaphragm-sidewall repair methods
Links
2-12
15
121-148
443-444
15
8-9
9-12
5-7
59-87
121-133
7-8
133-141
welding (cast iron)
139-141
weld rebuild
135-137
Diaphragm thermal distortion
121-133
corrective options
124-128
diametral change
129-133
dishing
121-128
59-87
60-63
metallic inserts
84-86
suction-face finishing
86-87
vane repair methods
64-87
weld buildup
64-84
Diffuser at exhaust
15
Dimensional requirements (reverse engineering)
132-133
133
manufacturing tolerances
Diffusion coatings/alloying
124-128
9
137-138
Diaphragm-vane repair (methods)
141-148
443-444
metallic inserts
permanence of correction
59-87
265
671-672
This page has been reformatted by Knovel to provide easier navigation.
133-148
Index Terms
Dimensional requirements/specification
before adjustment
reverse engineering
Links
28-35
83-84
664
671-672
697-698
724
56-58
753-754
83-84
671-672
stationary-blade row
28-35
surface conformance
697-698
Dimensional requirements (stationary-blade row)
28-35
blade opening/throat (O)
29-30
ratio O/P
29-30
setback (inlet/trailing edge)
31-32
sidewall-discharge diameters
32
vane cross-section irregularities
32-35
vane pitch (P)
29-30
vane-setting angle
30-31
vane-tilt angle
31
Discharge area (adjustment)
109-110
Discharge edge (size)
118-120
Discharge/inlet edges
760
Discharge nose
753-754
Discharge tail
27
cracking
56-58
Discs
342-362
corrosion
445-450
keyways and securing
358-362
removal
356-359
rim
348-349
spindle interface
445-450
turbine rotor
342-362
Disc-spindle assembly
352-356
Disc/spindle interface
445-450
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Discs (turbine rotor)
342-362
disc removal
356-359
disc rim
348-349
disc-spindle assembly
352-356
disc-to-hub fillet radii
350
forms
functions
343-350
343
interstage seals
346-348
keyways and securing
358-362
machining for shrink fits
350-351
pressure-balance holes
349-350
shrink-fit design
351-352
spindle interface
445-450
Disc-to-hub fillet radii
350
Dishing correction
124-128
heat and pressure
124-125
machining
125-128
Dishing (diaphragm)
121-128
axial pressure deformation
corrective options
124-128
122
elastic deflection
121
rubbing
Documentation and shipment
Dressing
blade
eroded surface
142-145
122
creep deflection
inner web
445-450
142-145
122
699
238-239
275
275
238-239
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
E
Elastic deflection (dishing)
Electric-discharge machining
Electron-beam physical-vapor deposition
Engineering review
121
729-730
266
680-682
access to sub-supplier plants
682
drawings
681
hold points
682
inspection and test plan
681
manufacturing processes
681
nonconforming components
681
quality-assurance program
quality records
Enthalpy and entropy
680-681
682
813-883
enthalpy of gas
825-828
entropy changes of steam
852-857
Mollier diagram
865-870
Enthalpy-entropy (Mollier) diagram
adiabatic expansion
steam-path expansion
throttling/non-expansive expansion
Enthalpy of gas
865-870
867-868
869
868-869
825-828
British thermal unit
824
calorie
824
Centigrade heat unit
824
Entropy changes of steam
852-857
adiabatic/isothermal expansion
854-855
constant-pressure gas expansion
855-856
constant-volume gas expansion
855
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Entropy changes of steam (Cont.)
general expression for change
856-857
temperature-changing cycle
853-854
Erosion damage/repair/control
repair
46-50
52-56
153
206-241
254-280
291-292
365-366
454-456
497
239-241
291-292
259
263-270
SEE ALSO Solid-particle erosion (SPE)
Solid-particle impact (SPI),
AND Water-induced damage
Erosion-resistant coatings
chemical-vapor deposition
266
diffusion coatings/alloying
265
gas-phase coatings
268
overlay coatings
265
pack cementation
physical-vapor deposition (electron beam)
plasma-coating process
Erosion-shield damage
cracks
Error types (blade profile/placement)
266-269
266
266-270
208-210
254-255
254-255
766-770
axial
767-770
blade setting
768-770
pitch
767-768
770
Examples
246-247
249-252
597-603
kW loss prediction
600-603
Martin method
585-589
moment weighing
246-247
seal maintenance economics
597-599
249-252
This page has been reformatted by Knovel to provide easier navigation.
585-589
Index Terms
Links
Expansion at constant pressure
845
855-856
Expansion at constant volume
845
855
Expansion-passage form
759-760
Experience with supplier
693
Explosion diaphragm
Extended vanes
15
168-169
33-34
Extrusion (vane)
729-731
Eye lashing (gauging)
762-763
747
F
Fabricated construction
60-61
Failure (definition)
xvi
Fatigue (corrosion)
289
Ferrules loss
424
Forging
337-342
522
524-527
727-729
733
744-746
524-527
envelope
727
material
733
precision
727-729
process
744-746
rotor
337-342
522
522
524-527
337-342
522
rotor attachment
Forgings (rotor)
basic production
borasonic examination
337-338
342
central-inspection boreholes
340-341
inspection during manufacture
338-340
rotor attachment
522
524-527
This page has been reformatted by Knovel to provide easier navigation.
524-527
Index Terms
Links
Forms of seal knife-edge discharge
604-618
knife-edge form
605-607
multi-strip seal configurations
612-618
seal-location geometry
608-612
Fretting corrosion
304
398-399
486-487
tie wires
424
Full-arc admission mode
260-262
Fusion techniques (rotating blades/hardware)
427-433
braze attachment
427-431
weld attachment
431-433
G
Gas equations
831-833
Boyles law
831
Charles law
831-832
general gas equation
832
normal temperature and pressure
833
Gas-phase coatings
Gauging (profile)
268
761-764
computer traces
763-764
eye lashing
762-763
guillotine gauge
761
projection methods
762
General case (steam expansion work)
General expression (entropy)
846
856-857
General gas equation
832
Geometry-change machining
542
This page has been reformatted by Knovel to provide easier navigation.
424
Index Terms
Links
Gland-ring form
618-627
butt clearances and tangential location
620-621
carbon rings
619-620
carrier geometry
622-623
seal and radial/axial steam forces
624-627
spring loading
621
steam seal
623
Gland-ring operating problems
643-650
Gland-ring/seal-strip materials
641-643
alloy steel
sand cast
643
642-643
Gland/seal-strip assembly
634-643
caulked strips
636-637
inserted segments
637-638
materials
641-643
staked strips
635-636
Grinding repair
180
Guillotine gauge
761
641
504-505
H
Hand cleaning
206
Hardness checks (rotor bending)
474-475
Heat addition to water/steam
833-837
change of states
834
conversion of ice to superheated steam
836-837
formation of steam
834-835
latent heat of evaporation
moisture content/dryness fraction
superheated steam
Heat-affected zone (HAZ)
835
835-836
836
541
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Heat engine
Heating and expansion of steam
Links
858
833-852
addition of heat to water/steam
833-837
expansion of steam
842-845
heating and expansion relationships
837-839
heating of steam
840-841
metastable conditions during expansion
(supersaturation)
ratio of specific heats
851-852
841
specific heats of steam
839-840
work done by steam expanding
845-851
Heating and expansion relationships (steam)
Heat of evaporation
Heavy damage (SPI)
837-839
835
44-46
High-cycle fatigue
481-482
High-pressure packing heads
170-171
High-pressure/high-temperature sections
162-166
Hi-lo staggered seal configuration
History (steam turbine)
615
x
Hogging. SEE Humping/hogging
Hold or witness points
Hole plugging/re-drilling
Horizontal half-joint
Horizontal joint
696
296-299
589
111-115
153-154
589
adjustment
blade adjustment; half joint
weld repair
Humping/hogging
correction
111-115
589
153-154
193-196
194-196
This page has been reformatted by Knovel to provide easier navigation.
197-198
Index Terms
Links
Hybrid rotor
336-337
Hydrogen sulphide
453-454
I
Ice conversion to superheated steam
836-837
Impact damage
363-371
coverband
363-371
Impact damage (coverband)
axial rubs
363-371
370
cracks at coverband hole
370-371
excessive overspeed
368-369
heavy radial rubs
368-369
increased thickness
368
reduced thickness
366-368
solid-particle erosion
365-366
solid-particle impact
363-364
water-impact damage
364-365
Impulse-unit seals
578-581
Induced bends (rotor)
464-479
bend-damage classification
469-470
causes
464-466
hardness checks
474-475
permanent
467-468
run-out checks
470-474
straightening options
475-479
stress relief
temporary
Inlet/discharge edges
496-497
479
466-467
760
Inlet-edge erosion (blade)
206-241
Inlet-edge profile finishing
222-223
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Inlet nose
definition
Inlet portion size
Inlet section
Links
26
753
753
117-118
15
Inlet-stage nozzle geometry
260
263
Inner ring/web
8-9
141-148
correction
141-148
Inserted segments (gland/seal strip)
637-638
Inspection and test plan (I&TP)
688-689
692
Inspection (components)
233-234
338-342
blades
brazing
rotor forgings
welding
711-813
234
338-342
233
Inspection department
686
Inspection-instrument calibration
695
Inspection (quality control)
694-695
instrument calibration
695
records
694-695
Inspection/surveillance
(activities and responsibilities)
dimensional/surface conformance
xviii-xvix
697-698
documentation and shipment
699
hold or witness points
696
inspection-instrument calibration
695
inspection records
material compliance
694-695
698
nonconforming items
696-697
nondestructive testing
698
quality-program review
694-699
694-695
This page has been reformatted by Knovel to provide easier navigation.
711-813
Index Terms
Links
Inspection/surveillance (Cont.)
special processes
supplier-purchaser control
697
698-699
Integral connection repair (tie wires)
425
Integral snubbers
299
Internal energy
824
Interstage seals (disc)
Irregularities (vane cross-section)
346-348
32-35
J
Joint geometry
232-233
K
Keyways and securing (disc)
358-362
Knife-edge form
605-607
KW loss prediction
599-604
examples
600-603
L
Labyrinth leakage
559-594
calculation
565-567
gland-ring form
618-627
gland-ring operating problems
643-650
Martin method
567-594
references
651
seal knife-edge discharge forms
604-618
seal-maintenance economics
595-604
seal-strip forms
627-630
seal-strip/gland-ring materials
641-643
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Labyrinth leakage (Cont.)
seal-strip insertion and securing
630-641
steam seal
559-567
Last-stage massive-particle damage
Latent heat
evaporation
Layout/measurements (welding)
Leakage
277-278
863
835
65-66
69
559-594
877-882
labyrinth
559-594
loss calculation
567-594
seal
559-594
steam
877-882
Leakage loss (Martin method)
567-594
example
585-589
incremental loss
593-594
radial-seal clearance
589-592
reaction-turbine dummy pistons
571-575
seal axial clearance
592-593
shaft-end glands
568-571
steam-path seals
575-585
Light damage (SPI)
835
41-42
Lines of constant superheat
864-865
Load-bearing surface radius (blade root)
796-797
Long-shank blades
505-507
Low-cycle fatigue
176-184
457-458
522-524
cracking
176-184
rotor
457-458
This page has been reformatted by Knovel to provide easier navigation.
493-494
Index Terms
Links
Low-cycle fatigue cracking
176-184
grinding
180
stitching
181-184
welding
180-181
Low-pressure/low-temperature sections
166-168
M
Machined from solid (construction)
61
Machining
61
350-351
477-478
541-543
705-709
729-730
components
705-709
finish
541-543
from solid
61
remachining rotor
477-478
shrink fit
350-351
Machining (components)
cutting process
705-709
706
surface finish
708-709
surface integrity
707-708
Machining finish
boreholes
541-543
707-709
542-543
concentricity/balance
542
geometry changes
542
pressure-balance holes
543
surface
542
Machining (shrink fit)
707-709
350-351
Maintenance
xiv
actions
xxi-xxiv
economics
595-604
xx-xxiv
This page has been reformatted by Knovel to provide easier navigation.
595-604
Index Terms
Links
Maintenance (Cont.)
options
problems
xx-xxiv
xiv
Manufacture/inspection requirements
(steam-turbine blades)
711-813
blade-manufacturing processes
726-749
blade-pitch errors (factors)
803-811
blade-root tolerances
775-802
manufacturing techniques
723-726
passage swallowing capacity
770-774
profile and cascade tolerances
749-764
profile and placement errors
765-770
radial alignment (rotating blades)
712-723
references
813
special processes (vane)
774
stage-hardware requirements
812
Manufacturing processes (blade)
basic-form production
726-749
733
cutting metal
733-744
electric-discharge machining
729-730
envelope forging
727
extrusion process
747
forging of material
733
forging process
744-746
metal cutting
726-727
pinch rolling
739
precision forging
727-729
vane extrusion
729-731
747-749
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Manufacturing techniques (blade)
723-725
blade mounting
725
blade production
724
dimensional requirements
724
material specifications
724
nondestructive examination
725
post-assembly requirements
725
spatial relationships
724
special processes
725
surface-finish requirements
724
tolerances
724
blade
724
Manufacturing tolerances
60-63
cast construction
60-61
fabricated construction
60-61
machined from solid
61
pinned construction
60-61
welded construction
60-61
vane adjustment
62-63
Martin method (leakage quantification)
724
567-594
blade tip
586-587
diaphragm packing
586-588
examples
585-589
incremental leakage loss
593-594
radial seal clearance
589-592
reaction turbine dummy pistons
571-575
seal axial clearance
592-593
shaft-end glands
568-571
steam-path seals
575-585
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Massive-particle damage
273-280
all stages
bending and dressing
blade removal
278-280
275
275-277
continue to operate
275
control stages
277
last-stage blades
Materials
277-278
69
95-96
539
664
672-675
698
724
733
compliance
698
deformation (form)
733
forging
733
removal
69
reverse engineering requirements/substitution
specification
95-96
672-675
664
724
Measurements required
65-66
69
Medium damage (SPI)
43-44
Metal section angles (blade)
Metallic inserts
754
84-86
Metastable conditions (steam expansion)
851-852
Mid-span crack (tie wires)
419-421
Misalignment
411-412
rotating blades
712-723
tie wires
411-412
Mismatch errors (vane)
137-138
712-723
115-120
small discharge edge
118-120
small inlet portion
117-118
vane mismatch
115-117
Moisture content/dryness fraction
539
835-836
864
This page has been reformatted by Knovel to provide easier navigation.
152-153
Index Terms
Moisture-impact damage
Links
54-56
364-365
SEE ALSO Water-induced damage.
Moisture removal (steam)
881-882
Mollier diagram (enthalpy-entropy)
865-870
adiabatic expansion
steam-path expansion
throttling/non-expansive expansion
Moment weighing
alternate means
867-868
869
868-869
224
253
blade numbering
243-246
examples
246-247
moment-weighing process
tangential-position determination
242-253
249-252
243
247-248
Monobloc rotor
330-331
Multi-strip seal configurations
612-618
alternative stationary/rotating
614
hi-lo staggered
615
multi-high strip
615-618
multi-high strip staggered
615-617
straight-through
613-614
N
New units (refurbishment)
292
Nonconforming items
696-697
Nonconforming situations
699-704
accept-as-is
703-704
repair
702-703
rework
scrap and replace
Nondestructive examination (NDE)
703
700-702
725
This page has been reformatted by Knovel to provide easier navigation.
454-456
Index Terms
Nondestructive testing (NDT)
Non-expansive expansion (throttling)
Normal temperature and pressure (gas)
Nozzle plate/box
Nuclear unit steam-path seals
Links
665
698
868-869
833
9-12
32-33
557
O
Off-shield erosion repair
239-241
below shield
239-240
between shield
240
beyond shield
239
pressure surface
240
tenon
241
Older units (refurbishment)
Opening pitch-ratio adjustment
292
107-108
Operate/shutdown decision
275
Operating changes
xiv
xvii
Operating damage mechanisms
(rotating components)
309-548
SEE ALSO Rotating components
(damage and refurbishment)
Operating damage mechanisms
(stationary components)
1-199
SEE ALSO Stationary components
(damage and refurbishment)
Operating experience (supplier)
Operating phenomena (stationary components)
692
36-59
chemical deposition
36-39
solid-particle erosion
46-50
solid-particle impact
40-46
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Operating phenomena (stationary components) (Cont.)
thermal transients
58-59
vane-discharge tail cracking
56-58
vane-discharge thumbnail crack
water-induced damage
Outer ring
58
50-56
5-7
damage
148-153
Outer-ring damage
148-153
solid-particle erosion
steam-seal face damage
Overlay coatings
153
149-153
265
Over-riveting (tenons)
395-396
Overspeed damage
368-369
coverband
tie wires
Oxide scale erosion
preventive/corrective actions
426
257-264
259-264
257
rotating-blade tip-section suction face
258
Oxygen (corrodent)
426
368-369
rotating-blade inlet edge
tenons attaching coverband
148-153
258-259
452-453
P
Pack cementation
266-269
Particle hardness
263
Particle size
263
Passage swallowing capacity
770-774
Peening
271-273
Rotor
273-280
477
477
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Performance definitions
660-663
performance factors
661-663
Performance factors (quality)
661-663
availability
efficiency
Permanence of correction
Physical properties (water and steam)
662
662-663
133
816-830
enthalpy of gas
825-828
internal energy
824
pressure
816
specific volume
816-822
temperature
822-824
viscosity
829-830
Physical-vapor deposition
Pinch rolling
266
731-732
Pinned construction
60-61
Pitch (vane)
29-30
747-749
767-768
770
767-768
770
803-811
Pitting (corrosion)
287-289
483-486
Plasma-coating process
266-270
803-811
errors
Post-assembly requirements
Post-brazing activities
inspection
725
233-234
234
Post-weld heat treatment
220-222
Power cycles
873-882
actual expansion and mixing (steam path)
877-882
representation
873-877
Predictable phenomena (casing)
175
Preheating
232
540
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Pressure-balance holes
349-350
disc
rotor
349-350
517
Pressure property (water and steam)
816
Pressure/suction faces (blade)
755
Pressure surface erosion
240
Preventive/corrective actions (oxide scale erosion)
259-264
acid washing of boiler
261
axial-gap increase
263
blades to modify steam-flow distribution
260
boiler-tube materials
261
boiler-tube temperature
262
erosion-resistant coatings
259
full-arc admission mode
263
260-262
inlet-stage nozzle geometry
260
scale-particle hardness
263
scale-particle size
263
vane material
262
Product surveillance (quality)
517
685-686
263
690-699
inspection/surveillance activities and
responsibilities
694-699
responsibility
685-686
supplier facility inspection (preparation)
691-693
Profile and cascade tolerances (blade)
749-764
blade-cascade definitions
756-760
blade-profile/cascade quality
751-752
blade-profile definitions
752-755
gauging the profile
762-764
Profile and placement errors (blades)
classes of error
765-770
766-770
This page has been reformatted by Knovel to provide easier navigation.
543
Index Terms
Profile chord/thickness (blade)
Profile gauging
Links
755
761-764
computer traces
763-764
eye lashing
762-763
guillotine gauge
761
projection methods
762
Profile sectional area/bending modulus
755
Profile setting angle
759
Profile (vane)
blade chord/thickness
20-23
755
blade placement errors
765-770
blade tolerances
749-764
constant-but-reducing
21-22
constant section
20-21
gauging
761-764
sectional area/bending modulus
755
setting angle
759
twisted
25-28
22-23
Program preparation/implementation
(quality assurance)
683-684
Projection methods (gauging)
762
Purchaser assurance (quality)
689-690
Purchasing/requisitioning engineer
692
Q
Quality assurance (QA)
653-710
available quality-assurance program
704-705
definition of quality
658-659
definitions of performance
660-663
design specification
663-665
This page has been reformatted by Knovel to provide easier navigation.
749-770
Index Terms
Links
Quality assurance (QA) (Cont.)
engineering review
680-682
inspection and test plan
688-689
machining of components
705-709
nonconforming situations
699-704
product surveillance
690-699
purchaser assurance of quality
689-690
quality-assurance manual
679-680
quality-assurance program
677-679
683-687
704-705
683-687
694-695
quality-assurance program
responsibility/administration
references
683-687
710
responsibility for quality
657-658
reverse engineering
666-677
Quality-assurance manual
679-680
Quality-assurance program
677-679
704-705
availability
company organization
final quality
inspection department
product quality responsibility
704-705
684
686-687
686
685-686
program implementation
684
program preparation
683
responsibility/administration
683-687
review/inspection
694-695
Quality (definition)
658-659
Quality-program review/inspection
694-695
instrument calibration
records
695
694-695
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
R
Radial alignment (rotating blades)
712-723
blade-tilt effect on tenons
722-723
blade-vane tilt
719-723
center of gravity shift
721-722
incorrect alignment
714-718
Radial rubs
368-369
Radial-entry direction (blade root)
797-801
Radial-seal clearance
589-592
casing distortion
590-592
horizontal half joint
589
Radius of curvature (profile surfaces)
754-755
Rankine cycle
872-873
Ratio O/P
29-30
Raw data
96-106
individual opening-discharge areas
opening (O)
pitch (P)
103-104
106
100
103-104
103-106
97-99
98
106
radial height
100-103
ratio O/P
103-104
Raw-weld deposit
235-238
Reaction-turbine dummy pistons
571-575
Reaction-unit seals
581-585
References
106
199
305-307
544-548
651
710
813
883
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Reforming
Links
371
crack grinding
504-505
root
503-504
rotor rim
503-505
tenons
503-505
371
Refurbishment techniques (coverbands)
396-397
Refurbishment techniques (rotating blades)
201-307
SEE ALSO Rotating blades
(damage and refurbishment)
Refurbishment techniques (rotating components)
309-548
SEE ALSO Rotating components
(damage and refurbishment)
Refurbishment techniques (stationary components)
1-199
SEE ALSO Stationary components
(damage and refurbishment)
Relief diaphragm
168-169
Remachining (rotor)
477-478
Re-measurement of passages
Repair (casing)
Repair (diaphragm)
110
174-198
59-87
inner web
141-148
sidewall
133-141
thermal distortion
124-128
vane
Repair (tie wires)
continuous connection
124-128
59-87
418-426
418-424
deterioration
426
ferrules loss
424
overspeed
426
snubber/integral connection
425
This page has been reformatted by Knovel to provide easier navigation.
133-148
Index Terms
Repair/correction
Links
59-87
124-128
133-148
153-162
174-198
239-241
291-303
418-426
462-463
702-703
axial rubs
462-463
casing
174-198
decision
702-703
erosion
239-241
291-292
59-87
124-128
diaphragm
horizontal joint
153-154
stationary blade
154-162
tie wires
418-426
Responsibility for quality
657-658
product
685-686
quality-assurance program
683-687
Reverse engineering
666-677
advanced planning
670-671
component evaluation/testing
676-677
component installation
677
concept
668-670
dimensional requirements
671-672
material requirements
672-673
material substitution
673-675
special processes
675-676
Reversibility
heat engine
Review (quality-assurance program)
Rework decision
Reworking tenons
683-687
857-858
858
694-695
703
371-372
This page has been reformatted by Knovel to provide easier navigation.
133-148
Index Terms
Links
Riveting process (tenon)
402-410
hand
405
pneumatic
405
rolled-rivet head
405
409-410
395-396
402-410
Riveting (tenon)
over-riveting
395-396
process
404-410
rolled
Rolled rivets
rivet head
Root
405
409-410
405
409-410
409-410
281-282
503-504
808-811
fastening area
281-282
reforming
503-504
surface finish
wedge angle
802
808-811
Rotating blades (damage and refurbishment)
201-307
blade inlet-edge erosion damage
206-241
blade trailing-edge erosion
255-257
corrosion effects
281-291
erosion-resistant coatings
264-270
erosion-shield cracks
254-255
fretting corrosion
304
massive-particle damage
273-280
moment weighing (refurbished blades)
242-253
references
305-307
rotating-blade refurbishment
291-303
solid-particle erosion by oxide scale
257-264
This page has been reformatted by Knovel to provide easier navigation.
802
Index Terms
Links
Rotating blades (damage and refurbishment) (Cont.)
solid-particle peening
271-273
steam-path cleaning
204-206
water induction
303-304
Rotating-blade corrosion
blade row
281-291
442
corrosion effects
281-291
corrosion fatigue
289
corrosion pitting
287-289
fretting
304
stress-corrosion cracking
287
Rotating-blade erosion
257-258
inlet edge
257
tip-section
258
Rotating-blade refurbishment
erosion damage/repair/control
291-303
291-292
forming new tenons
303
new units
292
older units
292
vane-tip cracks
299-303
vane weld repair (tie-wire hole)
293-299
Rotating components (damage and refurbishment)
304
309-548
blade-root steeples/wheel rim
480-497
components
309-362
coverband damage/repair/refurbishment
362-410
fusion techniques
427-433
induced bends (rotor)
464-479
references
544-548
rotor-damage mechanisms
433-463
rotor-rim damage
498-507
This page has been reformatted by Knovel to provide easier navigation.
442
Index Terms
Links
Rotating components
(damage and refurbishment) (Cont.)
rotor weld repair
508-527
tie-wires damage/repair/refurbishment
410-426
weld-repair process
527-543
Rotor
327-362
433-479
508-527
539
bending
464-479
construction
328-337
corrosion
439-454
damage mechanisms
433-463
discs
342-362
forgings
337-342
functions
327-328
rim damage
498-507
surface
508-514
turbine
342-362
welded
332-336
weld repair
508-527
weld material
539
wheel
508
Rotor and weld material
Rotor bending
539
464-479
bend-damage classification
469-470
causes
464-466
hardness checks
474-475
permanent
467-468
run-out checks
470-474
straightening options
475-479
stress relief
514-522
479
This page has been reformatted by Knovel to provide easier navigation.
498-527
Index Terms
Links
Rotor bending (Cont.)
temporary
466-467
Rotor construction
328-337
built-up
331-332
hybrid
336-337
monobloc
330-331
welded
332-336
Rotor corrosion
blade-tip seal leakage
439-454
444
corrodents
450-454
diaphragm leakage
443-444
disc/spindle interface
445-450
hydrogen sulphide
453-454
oxygen
452-453
rotating-blade rows
442
shaft-end sealing positions
445
sodium hydroxide
451-452
steam throttling
442-443
Rotor-damage mechanisms
433-463
axial rubs
459-463
burns
459-463
corrosion
439-454
low-cycle fatigue
457-458
rough/out-of-phase synchronization
437-438
short circuit of generator
433-437
stage wetness/moisture
454-456
Rotor forgings
basic production
borasonic examination
337-342
337-338
342
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Rotor forgings (Cont.)
central-inspection boreholes
340-341
inspection during manufacture
338-340
Rotor-rim damage
498-507
load-bearing surface skimming
498-503
long-shank blades
505-507
rebuild by welding
498
reforming
503-505
Rotor surface
508-514
Rotor weld repair
508-527
attachment of forgings
522
524-527
cracks
509
522-527
stub attachment crack
508-509
522
surfaces
508-514
wheel
Rotor wheel
508
514-522
508
514-522
blade-root region
516-517
fillet radii
517-519
pressure-balance holes
517
wheel-forging attachment
522
Rubbing (coverband)
385-388
391-392
122
368-370
385-388
391-392
459-463
370
391-392
coverband
368-370
385-388
coverband surface
385-388
axial
368-370
370
coverband surface
385-388
radial
368-369
Rubbing/rubs
axial
This page has been reformatted by Knovel to provide easier navigation.
391-392
Index Terms
Links
Rubbing/rubs (Cont.)
radial
368-369
rotor
459-463
Run-out checks (rotor bending)
470-474
S
Sand-cast materials
642-643
Saturated-steam line
863
Saturated-water line
862-863
Scale particle
263
hardness
263
size
263
Scrap and replace decision
700-702
Screw attachment (coverband)
382-383
Seal knife-edge discharge forms
604-618
knife-edge form
605-607
multi-strip seal configurations
612-618
seal-location geometry
608-612
Seal-location geometry
608-612
Seal-maintenance economics
595-604
examples
597-603
kW loss prediction
599-604
Seals, glands, and sealing systems
549-651
labyrinth leakage
559-594
maintenance economics
595-604
steam-path seals
551-557
steam-sealing system functions
557-559
steam-seal leakage
559-567
Seal-strip forms
627-630
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Seal-strip insertion/securing
630-641
gland and seal-strip assembly
Seal-strip materials
alloy steel
sand cast
634-641
641-643
643
642-643
Seal-strip/gland assembly
634-641
caulked strips
636-637
inserted segments
637-638
staked strips
635-636
Setback (inlet/trailing edge)
31-32
Setting angle (vane)
30-31
Set-up (welding)
66-68
Shaft end
glands
packing head
sealing
15
445
Shield cleaning
231-232
Shield damage
208-210
433-437
Shrink fit (disc)
350-352
design
351-352
machining
350-351
dressing
discharge diameters
washing erosion
15
699
Short circuit (generator)
Sidewall
568-471
15
13
Side-grip damage (blade root)
445
568-571
Shells (casing)
Shipment of products
641
488-492
32
52-54
78-81
32
52-54
This page has been reformatted by Knovel to provide easier navigation.
78-81
Index Terms
Links
Snubber/integral connection (tie wires)
414-418
repair
Snubber repair (tie wires)
425
425
Sodium hydroxide
451-452
Solid-particle erosion (oxide scale)
257-264
preventive/corrective actions
259-264
rotating-blade inlet edge
257
rotating-blade tip-section suction face
258
tenons attaching coverband
Solid-particle erosion (SPE)
oxide scale
Solid-particle impact (SPI)
425-426
258-259
46-50
153
273-280
365-366
257-264
40-46
heavy damage
44-46
light damage
41-42
medium damage
43-44
Solid-particle peening
271-273
273-280
363-364
697
Spatial relationships
724
Special processes
665
675-676
725
774
reverse engineering
675-676
specification
665
vane
774
Specification (design)
257-264
663-665
assembly/alignment
665
material
664
nondestructive tests
665
physical dimensions
664
special processes
665
surface finish
665
724
724
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Specific heats (steam)
839-841
ratio of specific heats
Specific volume (water and steam)
Spring loading (gland ring)
Stage-discharge area/angle (determination)
841
816-822
621
88-96
adjustment methods
93-96
material removal
95-96
vane bending
95-96
weld buildup
95-96
Stage-hardware requirements
812
Stage wetness/moisture damage
454-456
Staggered seal configuration
615-617
Staked strips (gland/seal strip)
635-636
State change (heat addition)
Stationary-blade damage
repair
Stationary-blade repair
Stationary-blade row components
casings
834
154-162
154-162
154-162
1-15
2
definitions
5-15
diaphragms
2-12
Stationary-blade row geometry
15-35
dimensional requirements
28-35
three-dimensional
20-23
two-dimensional
16-20
vane profile
25-28
vane tilt
23-25
Stationary blades
damage
definition
1-35
4-5
154-162
154-162
5-6
This page has been reformatted by Knovel to provide easier navigation.
13-15
Index Terms
Links
Stationary blades (Cont.)
repair
row components
row geometry
Stationary components (damage and refurbishment)
adjustment computation
154-162
1-15
15-35
1-199
96-121
casing components
162-173
casing operating problems/repair methods
174-198
components
1-15
diaphragm inner web correction
141-148
diaphragm sidewall repair methods
133-141
diaphragm thermal distortion
121-133
diaphragm-vane repair
horizontal-joint weld repair
operating phenomena
outer-ring damage
references
59-87
153-154
36-59
148-153
199
stage-discharge area/angle
88-96
stationary-blade row geometry
15-35
stationary-blade damage
Stationary/rotating seal configuration
154-162
614
Steam discharge angle
30
Steam-discharge angle
108-109
adjustment
108-109
Steam expansion
842-852
adiabatic
869
847-850
constant pressure
845
constant volume
845
general case
846
throttling
757-758
850-851
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Steam-flow distribution
Links
260
Steam forces (steam seal)
624-627
Steam formation
834-835
Steam heat
833-852
Steam-inlet angle
Steam leakage
757
559-594
enthalpy-entropy (Mollier) diagram)
865-870
intermediate-seal location
880-882
rotating-blade row
878-880
seal
559-594
stationary-blade row
877-878
Steam-path cleaning
204-206
blast cleaning
205
hand cleaning
206
water washing
205-206
Steam-path expansion/mixing
842-852
expansion
842-852
moisture removal
881-882
steam leakage (intermediate-seal location)
880-882
steam leakage (rotating-blade row)
878-880
steam leakage (stationary-blade row)
877-878
Steam-path seals
877-882
149-153
face damage
149-153
general requirements
554-556
impulse unit
578-581
leakage
559-594
nuclear units
557
reaction unit
581-585
sealing system
557-559
869
551-594
This page has been reformatted by Knovel to provide easier navigation.
877-882
Index Terms
Links
Steam properties
859-873
Carnot cycle
870-872
Rankine cycle
872-873
steam tables
859-860
temperature-entropy diagram
861-865
Steam-seal face damage
metallic inserts
149-153
152-153
Steam seal (gland ring)
623-627
Steam-seal leakage
559-594
calculation
565-567
Martin method
567-594
Steam-sealing system
557-559
Steam tables
859-860
Steam throttling (corrosion)
442-443
Steam-turbine blades
(manufacture/inspection requirements)
711-813
blade-manufacturing processes
726-749
blade-pitch errors (factors)
803-811
blade-root tolerances
775-802
manufacturing techniques
723-726
passage swallowing capacity
770-774
profile and cascade tolerances
749-764
profile and placement errors
765-770
radial alignment (rotating blades)
712-723
references
813
special processes (vane)
774
stage-hardware requirements
812
Steam-turbine components (quality assurance)
653-710
available quality-assurance program
704-705
definition of quality
658-659
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Steam-turbine components (quality assurance) (Cont.)
definitions of performance
660-663
design specification
663-665
engineering review
680-682
inspection and test plan
688-689
machining of components
705-709
nonconforming situations
699-704
product surveillance
690-699
purchaser assurance of quality
689-690
quality-assurance manual
679-680
quality-assurance program
677-679
683-687
quality-assurance program
responsibility/administration
references
683-687
710
responsibility for quality
657-658
reverse engineering
666-677
Steam-turbine technology
xii
Steam turning angle
758-759
Stitching repair
181-184
Straightening (rotor bend)
475-479
bending
476
peening
477
remachining
477-478
stress relief
476
thermal
476
Straight-through seal configuration
613-614
Stress-concentration centers
481-482
Stress-corrosion cracking
287
This page has been reformatted by Knovel to provide easier navigation.
704-705
Index Terms
Stress relief
Links
82-83
540-541
479
heat-affected zone
541
rotor straightening
476
tempering
Stress (tie wires)
82-83
411-413
batch discontinuity
413
batch twist
413
centrifugal
411
compressive
411
misalignment
411-412
thermal
412
vibration
411
Stub-attachment crack
Stub shaft
attachment
Suction-face finishing
Superheated steam
ice conversion
508-509
522
530
537-538
530
86-87
836-837
864-865
836-837
Supersaturation (steam)
851-852
Supplier facility inspection (preparation)
691-693
component operating modes (changes)
693
design-review meeting minutes
693
experience with supplier
693
inspection and test plan
692
operating experience
692
purchasing/requisitioning engineer
692
technical-purchase specification
692
Supplier-purchaser control
479
698-699
This page has been reformatted by Knovel to provide easier navigation.
476
Index Terms
Surface curvature
Surface finish
coverband
curvature
integrity
Links
28
665
724
802
28
385-388
508-514
529
531
665
707-709
724
802
28
665
707-709
724
802
385-388
707-708
machining
542
707-709
rebuild
529
531
requirements
724
rotor
specification
Surface finish (machining)
cutting fluids
integrity
machining rates
Surface integrity
built-up edge retention
508-514
665
542
707-709
709
707-708
709
707-708
707-708
surface burning
707
tears and gouges
707
tool chatter
707
Surface rebuild
707-709
529
Surface rubs (coverband)
385-388
Surface skimming
498-503
axial-entry roots
502-503
blade replacement
501-502
blade reuse
499-501
Synchronization (rotor)
437-438
531
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
T
Tangential-entry direction (blade root)
788-796
Tangential position determination (blade)
247-248
blade listing
247
blade placement
248
blade selection
Technical-purchase specification
247-248
692
Temperature-changing cycle (entropy)
853-854
Temperature-entropy diagram
861-865
critical point
latent heat
lines of constant superheat
863-864
863
864-865
moisture content
864
saturated-steam line
863
saturated-water line
862-863
Temperature (water and steam)
822-824
absolute temperatures
821-822
Tempering/stress relief
82-83
Tenon and hole requirements (coverband)
400-401
Tenon corrosion
283-287
398-399
Tenon erosion
241
258-259
Tenon failures
395
Tenon forming
303
new
303
reforming
371
Tenon-hole form
Tenons
371
393-394
400-401
241
258-259
283-287
303
371-381
393-396
398-410
722-723
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Tenons (Cont.)
corrosion
283-287
398-399
erosion
241
258-259
failures
395
forming
303
hole form/requirements
393-394
over-riveting
395-396
reforming
371
reworking
371-372
riveting
395-396
weld deposit
372-373
weld rebuild
373-381
Thermally-hardened inlet edges
Thermal straightening (rotor)
stress relief
Thermal stress
Thermal transients
Thermodynamics of water and steam
476
412
58-59
813-883
entropy of steam
852-857
gas equations
831-833
heating and expansion of steam
833-852
physical properties
816-830
883
reversibility
857-858
steam properties
859-873
increased
reduced
479
479
873-882
Thickness (coverband)
402-410
212-213
basic power cycles
references
400-401
366-368
368
366-368
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Three-dimensional considerations (stationary-blade row) 20-23
vanes (constant-but-reducing profile)
21-22
vanes (constant section)
20-21
vanes (twisted profile)
22-23
Throat opening
Throttle-controlled units
Throttling (steam expansion)
non-expansive expansion
756-757
34-35
850-851
868-869
Thumbnail crack (vane-discharge tail)
58
Tie-wire damage/repair/refurbishment
410-426
forming/connection methods
414-418
repair considerations
418-426
stresses
411-413
Tie-wire forming/connection
414-418
continuous connection
414-416
snubber/integral connection
414-418
Tie-wire hole
hole plugging/re-drilling
Tie wires
868-869
282-283
293-299
296-299
282-283
293-299
317-326
410-426
batching with coverbands
323-326
cross sections
320-323
damage/repair/refurbishment
410-426
functions
317-318
forming/connection
414-418
forms
318-320
tie-wire hole
282-283
293-299
Tilt (vane)
23-25
31
angle
31
axial
23-25
This page has been reformatted by Knovel to provide easier navigation.
719-723
Index Terms
Links
Tilt (vane) (Cont.)
compound radial
tangential
Turbine-rotor discs
24
23-24
342-362
disc removal
356-359
disc rim
348-349
disc-spindle assembly
352-356
disc-to-hub fillet radii
350
functions
343
forms
343-350
interstage seals
346-348
keyways and securing
358-362
machining for shrink fits
350-351
pressure-balance holes
349-350
shrink-fit design
351-352
Twisted profiles
22-23
Two-dimensional considerations (stationary-blade row)
16-20
742-743
U
Unpredictable phenomena (casing)
176
V
Vane adjustment
62-63
Vane bending
95-96
Vane cross-section irregularities
32-35
extended vanes
33-34
nozzle box
32-33
throttle-controlled units
34-35
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Vane-discharge tail cracking
thumbnail crack
Links
56-58
58
Vane dressing
78-81
Vane extension
33-34
Vane extrusion
729-731
Vane inclination
806-807
Vane material
Vane mismatch errors
262
115-120
small discharge edge
118-120
small inlet portion
117-118
Vane pitch (P)
Vane placement (root platform)
Vane profile
29-30
805-806
20-23
constant-but-reducing
21-22
constant section
20-21
discharge tail
27
inlet nose
26
surface curvature/finish
28
twisted
Vane projection across half joint
25-28
22-23
120-121
Vane repair methods
64-87
metallic inserts
84-86
suction-face finishing
86-87
weld buildup
64-84
Vanes
747
293-299
7-8
20-35
56-58
62-87
95-96
115-121
262
293-303
719-723
729-731
747
805-808
adjustment
62-63
bending
95-96
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Vanes (Cont.)
cross-section irregularities
32-35
discharge-tail cracking
56-58
extension
33-34
extrusion
729-731
inclination
806-807
material
mismatch errors
pitch
placement
profile
projection across half joint
262
115-120
29-30
805-806
20-23
64-87
setting angle
30-31
sidewall dressing
78-81
tilt
23-25
twist/shape variations
weld repair
31
719-723
31
719-723
808
293-299
30-31
Vane tilt
23-25
angle
31
axial
23-25
24
tangential
23-24
Vane-tip cracks
299-303
Vane twist/shape variations
293-299
299-303
Vane setting angle
compound radial
25-28
120-121
repair methods
tip cracks
747
808
Vane weld repair (tie-wire hole)
293-299
blade removal/access
295-296
hole plugging/re-drilling
296-299
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Vane weld repair (tie-wire hole) (Cont.)
welding in situ
295
welding integral snubbers
299
Vapor deposition
266
Vibration stress
411
Viscosity (water and steam)
829-830
Vortex profiles
742-743
W
Washing erosion
454-456
Water and steam thermodynamics
813-883
basic power cycles
873-882
entropy of steam
852-857
gas equations
831-833
heating and expansion of steam
833-852
physical properties
816-830
references
883
reversibility
857-858
steam properties
859-873
Water-impact damage
364-365
Water-induced damage
50-56
205-206
454-456
497
centrifuged moisture
54-56
sidewall washing
52-54
washing erosion
454-456
water-impact damage
364-365
water washing
205-206
worming/wire drawing
Water washing
erosion
497
364-365
497
454-456
497
205-206
454-456
497
454-456
497
51-52
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Weld attachment
209-210
213-230
522
524-527
inlet edge/base material buildup
224-230
shield
209-210
forgings to rotor
solid-bar inlet nose
Weld-attaching inlet nose
522
213-224
213-216
final moment weighing
224
inlet-edge profile finishing
222-223
post-weld heat treatment
220-222
weld inspection
223
welding process
220-221
welding process preparation
217-219
64-84
copper plugs/backing strips
70-73
dimensional requirements before adjustment
83-84
diaphragm
66-68
initial weld deposit
74-77
layout/measurements
95-96
65
initial set-up
inlet edge
524-527
213-224
blade vane preparation
Weld buildup
431-433
224-230
65-66
material removal
69
required measurements
69
stress relief/tempering
82-83
vane/sidewall dressing
78-81
weld metal/filler materials
77-78
weld preheat
73-74
69
This page has been reformatted by Knovel to provide easier navigation.
224-230
Index Terms
Weld deposit
Links
74-77
235-238
372-373
540
tenons
372-373
Welding process
217-221
preparation
217-219
rotating components
527-543
Welding repair
527-543
180-181
508-543
attachment of forgings
522
524-527
deep cracks
509
522-527
options
529-530
process
527-543
rotor surfaces
508-514
stub attachment crack
508-509
522
508
514-522
64-84
95-96
135-137
139-141
180-181
209-210
213-230
233
235-238
293-299
372-381
431-433
498
508-543
209-210
213-230
522
524-527
64-84
95-96
224-230
235-238
372-373
wheel
Welding
attachment
buildup
cast iron
139-141
construction
60-61
deposit
74-77
540
in situ
295
inspection
233
integral snubbers
299
metal/filler materials
431-433
77-78
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Welding (Cont.)
preheat
73-74
process
217-221
527-543
rebuild
135-137
498
repair
180-181
508-543
repair options
529-530
rotating components
527-543
set-up
66-68
tenons
373-381
Weld inspection
233
Weld metal/filler materials
77-78
Weld preheat
73-74
Weld rebuild
135-137
diaphragm sidewall
rotor rim
tenons
Weld-repair options
crack filling
135-137
498
373-381
529-530
529-530
rebuilding surface
529
rebuilding wheel
529
stub-shaft attachment
530
Weld-repair process (rotating components)
527-543
damage examination
531-538
finish machining
541-543
material removal
539
preheating
540
procedure evaluation
530
repair options
rotor and weld material
373-381
529-530
539
This page has been reformatted by Knovel to provide easier navigation.
498
Index Terms
Links
Weld-repair process (rotating components) (Cont.)
stress relief
weld deposit
Wheel gouging
Wheel rebuild
540-541
540
494-496
529
Work done (steam expansion)
845-851
adiabatic expansion
847-850
expansion at constant pressure
845
expansion at constant volume
845
general case
846
throttling
Worming/wire drawing
531-535
850-851
51-52
This page has been reformatted by Knovel to provide easier navigation.