3rd Middle East Turbomachinery Symposium (METS III)

15-18 February 2015 | Doha, Qatar | mets.tamu.edu

BABBITTED BEARING HEALTH ASSESSMENT
John K. Whalen, P.E.
Chief Engineer - Bearings
John Crane
Houston, TX, USA

Jim Allen
Senior Rotating Equipment Specialist
Engineering Technologist
Nova Chemicals Corporation
Joffre, Alberta, Canada

Thomas D. Hess, Jr., P.E.

DuPont Engineering Research & Technology
Rotating Machinery Group
Chestnut Run
Wilmington, DE, USA

John K. Whalen is the Chief Engineer Bearings for John Crane Engineered
Bearings. John is a member of STLE,
ASME, and the Vibration Institute and is a
member of the Turbomachinery Symposium
Advisory Committee; John is a registered
Professional Engineer in the State of
Texas.
John received his BSME from the
Rochester Institute of Technology in 1981. He spent 7 years
with Dresser Rand in Wellsville, NY where he worked in Large
Turbine Engineering and Rotordynamics. John joined
Centritech Corp. in 1988 and helped form TCE in 1991.
At John Crane Engineered Bearings, John has Engineering
responsibility for bearing technology across the three bearing
centers in Gttingen, Grafton and Houston. The product range
includes babbitted bearings and seals, labyrinth seals, and
related products and services.
James Jim Allen is a Senior Rotating
Equipment Specialist in the Reliability
Team for Nova Chemicals, in Red Deer,
Alberta.
Jim is responsible for the reliable
operation of the Olefins Plants Equipment
by supporting operations / Maintenance /
Engineering. Prior to joining Nova
Chemicals Jim worked for Dow Chemicals
in Fort Saskatchewan, Alberta and Sherritt Gordon Mines,
Lynn Lake Manitoba.
Jim is as certified Industrial Mechanic and Engineering
Technologist. He is also member of the CMVA, STLE and has
been an active member of the world of rotating equipment for
over 40 years.
Thomas Hess works in the Rotating
Machinery Group for DuPont located in
Wilmington, Delaware. He is responsible
for plant and capital project support and
standards development for DuPont.
Prior to joining DuPont, Mr. Hess
worked for Valero at the Delaware City

Jack Craighton, CLS

Technical Consultant
Schneider Electric
Wilmington, DE, USA

Refinery, ARCO Chemical in Newtown Square, PA, and Bently

Nevada as a field vibration analyst.
Mr. Hess has a BS in Mechanical Engineering from
Villanova University. He is also a member of ASME, the
Vibration Institute, and is a Registered Professional Engineer
in the Commonwealth of Pennsylvania.
John Craighton works for Schneider
Electric as a Technical Consultant
He was a corporate lubrication
consultant responsible for plant level
support and standards development for
DuPont.
Prior to joining DuPont, Mr.
Craighton worked for Invensys Operations
Management
and
Valero
Energy
Corporation as a Lubrication Specialist.
John is a Certified Lubrication Specialist through the
Society of Tribologists and Lubrication Engineers.
ABSTRACT
Within most plants there is often a desire to extend the
time between outages. Several factors determine this but one to
consider is the bearings ability to run for the extended time.
This Tutorial will address common bearing failure
mechanisms, tools available to diagnose these problems and
ways to use that information to make decisions on continued
operation. Evaluation consists of analysis of Condition
Monitoring information coupled with rotor dynamic analysis
and computerized bearing modeling.
A case study will be presented that details the evaluation of
a distressed bearing. Analysis allowed the user to determine
they could slightly reduce running speed (to limit vibration to a
certain level) and continue to run to the scheduled outage date.
There will then be a discussion on bearing robustness and
ways to design bearings for long life. A case study will be
presented where an Ethylene plant desired to extend the time
between outages but the existing bearings prohibited that. The
determination was made that upgrading the bearings to a more
robust design would allow the extended time between outages.

Copyright 2015 by Turbomachinery Laboratory, Texas A&M Engineering Experiment Station

INTRODUCTION
Most process plants and power generation facilities are
required to maximize the use of their assets by running for
longer periods of time (longer time between outages) and
reducing the length of the outage when they are down. Of
course equally important are the need to not come down
unexpectedly (a forced outage) and the requirement to run
efficiently to maximize rates.
Many companies have also been implementing
debottlenecking projects throughout their facility and
sometimes these result in increased demand from other
equipment in the facility. This increased demand could result in
a loss of reliability of that equipment which is being pushed
harder to make rates.
Plant owners and operators realize it is important to
understand which equipment could limit the time between
outages and what component of that equipment is most
vulnerable. It is understood that there is a tremendous amount
of fixed and rotating equipment; any one of which could be the
limiting factor when evaluating the length of time required
between outages. It is also recognized that with rotating
equipment there are several factors which can influence this
scheduling. However this tutorial is only going to discuss oil
lubricated babbitted bearings (journal and thrust) as found in
existing rotating equipment.
To be clear we will be discussing bearings as found in
critical un-spared equipment. Machines that should be fully
instrumented as per API 670 (2010) as a minimum, and with
condition monitoring procedures and processes in place. It is
beyond the scope of this presentation to get deeply into
condition monitoring practices it will just be assumed that
data is available for analysis.
Several failure mechanisms will be introduced discussing
evaluation of the damage mechanism, tools that may help
diagnose the problem while running and methods to avoid the
particular failure mode. The intent is that understanding
common failure modes can lead to design and operation to
avoid them; thus pushing the limits of operation.
Coupled with this is the evaluation of running equipment
and the need to determine; if the machine can run to the next
scheduled outage; if the machine rates (and speed) need to be
reduced to ensure successful operation to the outage; if an
outage needs to be taken before the scheduled outage; or if the
machine needs to be shut down immediately to avoid further
damage.
HYDRODYNAMIC BEARING OPERATION
In order to be able to diagnose bearing problems it is
important to have a basic understanding of bearing design and
operation. The bearings covered in this presentation include
babbitted journal and thrust bearings as found in typical special
purpose API turbomachinery. For simplicity, and in order to
cover the intended machine population effectively, we will not
be discussing bearings lined with materials other than babbitt,
anti friction bearings, magnetic bearings, foil bearings, etc.
The definition of hydrodynamic is: pertaining to forces
in - or motions of - liquids. So we have a fluid in motion (oil)
that generates forces to separate the stationary (bearing) from

the rotating (rotor). This may also be referred to as Fluid Film

Lubrication since in operation a film of oil separates the
bearing from the rotor.
The study of this phenomena falls under the field of
Tribology (Dowson, 1979) a word that was coined in 1966:
the science and technology of interacting surfaces in
relative motion and the practices related thereto. Also often
referred to as the study of friction, wear and lubrication.
The basic requirement for successful operation of
hydrodynamic bearings is the formation of the wedge.
The Hydrodynamic Wedge
Figure 1 is a sketch of a journal in a bearing (with greatly
exaggerated clearances). This figure also contains some terms
that will be used throughout this presentation.

Figure 1 Hydrodynamic film generation

There is a journal rotating (CCW) inside a journal bearing

and the rotation of this journal drags oil into a converging
space. Since oil is incompressible, all of the oil cannot be
pulled though this space so a pressure is generated that forces
the oil out the axial ends of the bearing. This pressure is
depicted by the pressure profiles as shown.
For the sleeve bearing being considered we see that the
rotor is displaced down and to the right since the pressure
field starts on the left side of the journal. A line drawn between
the journal center and the bearing center is called the line of
centers and the angle between this line and the applied load is
called the attitude angle.
The amount the journal is displaced from the center of the
bearing is referred to as the eccentricity and the term
eccentricity ratio is the ratio of this eccentricity to the
bearing radial clearance. As such an eccentricity ratio of 1.0
implies the eccentricity equals the clearance and the journal is
touching the bearing surface. Likewise an eccentricity ratio of
zero indicates the journal is centered in the bearing.
The minimum oil film thickness lies on the line of
centers and can be defined by the eccentricity ratio and bearing
radial clearance. Figure 2 is a drawing of the clearance space
unwrapped starting at the point of maximum oil film
thickness (opposite the minimum) and continuing 360 degrees

Copyright 2015 by Turbomachinery Laboratory, Texas A&M Engineering Experiment Station

though the minimum and back to the maximum. Here the

converging and diverging wedges can be more clearly seen.

separate, the film becomes thicker and the shearing of the

thicker film increases the coefficient of friction.
Thrust Bearings

Figure 2 Unwrapped Clearance Space

Terms that are important to retain from this section include

the minimum oil film thickness, the wedge, eccentricity ratio
and attitude angle.

Thrust bearings absorb the axial loads present in

turbomachinery. As journal bearings have to support the rotor,
thrust bearings must support this axial load.
The hydrodynamic principle is the same since the face of
the babbitted bearings must have a converging wedge to
support the load. Figure 4 contains a drawing of a taper land
thrust plate. In this design the wedge (or taper) is machined into
the face of the bearing as noted in the section on the left. The
rotation of the thrust collar pulls oil into this converging space
and hydrodynamic pressure is generated.

The Stribeck Curve

As a machine is brought up from zero speed to full speed it
goes though three hydrodynamic lubrication regimes. The
Stribeck curve contains these regimes and is presented in
Figure 3. The curve plots the coefficient of friction on the
vertical axis and ZN/P on the horizontal. In this plot Z is
viscosity, N is speed and P is the unit load on the bearing. The
best way to first study this curve is to envision a machine
coming up in speed, where the viscosity and load are constant.

Figure 4 Taper Land Thrust Plate

The taper land bearing is limited in load capacity as that

taper can only be machined at a single given angle. In reality
thrust bearings must be able to accommodate varying loads and
speeds and as such a tilting face would accommodate those
variations. For this reason tilting pad thrust bearings are
commonly used. As the name implies the individual pads tilt
depending on the load capacity required.

Figure 3 Stribeck curve

At zero speed and at very low speeds there is boundary

lubrication; here the two surfaces are in contact with a
relatively low coefficient of friction (~0.15) thanks to the
presence of oil and two very good surfaces (babbitt and
typically a ground journal). Once the speed is sufficient to pull
oil into the wedge hydrodynamic pressure begins to develop
and the surfaces begin to separate; hence the drop in friction as
speed increases, this regime is referred to as mixed film
lubrication. Once the surfaces are fully separated we have a
minimum friction (depicted as .001 in the figure) and as speed
increases the friction gradually increases. This increase in
coefficient of friction is due to the fact that as the surfaces

Figure 5 Tilting Pad Thrust Bearings

Figure 5 shows two variations of tilting pad thrust

bearings. On top is the non-equalized bearing with pad pivots in
the circumferential center of the pad and on the bottom is a
fully equalized tilting pad thrust bearing. The equalized bearing
utilizes a set of leveling links or plates under the pads; the links
articulate to allow each pad to take an equal share of the load.

Copyright 2015 by Turbomachinery Laboratory, Texas A&M Engineering Experiment Station

This bottom drawing also shows the use of an offset pivot

which can be utilized to increase load capacity (by allowing the
pad to tilt more thereby providing an optimized converging
wedge).
Babbitt
Babbitt (or white metal) is typically a tin based material
that lines the bore of journal bearings or the faces of thrust
bearings. It was invented in 1839 (by Isaac Babbitt in Boston,
MA) and has been used extensively ever since. The desired
properties for this material are:

Compatibility with the journal (non galling)

Embedability dirt and wear particles
Conformability account for misalignment, etc.
Strength - must be able to support the load
Corrosion resistance
Low Cost
Availability
Low Coefficient of Friction

The main idea is that the bearing is the sacrificial element,

should something happen to cause damage to a turbomachine it
is preferred that only the bearings become damaged and the
rotor and other components survive intact.
There is an ASTM spec for babbitt: ASTM B23-00 (2010)
which specifies compositions and physical properties for
several grades (alloys). The most common grade is Alloy 2
which is nominally composed of 89% Tin, 7 % Antimony and
3 % copper. This grade of babbitt begins to melt at 466 F
and is fully melted at 669 F. Needless to say as temperature
increases above room temperature the babbitt strength drops
off.

All bearings covered here are oil lubricated, so a brief

discussion of oil is in order. Oil in critical turbomachinery
must:

Have adequate viscosity to generate the fluid film

Move the generated heat so it can be removed
Flush the bearings to keep them clean
Coat steel surfaces to reduce rust and oxidation

Viscosity is a measure of a fluids resistance to shear, it

needs to be high enough to support the load with an adequate
film thickness but low enough to limit operating temperatures
and power loss. The viscosity index (VI) is a measure of an oils
viscosity stability (how much it changes with temperature),
most oils today have a VI above 110.
Common oil grades are ISO 32 (most common), followed
by ISO 46 then ISO 68. The heavier oils (46 and 68) are being
used less today in an attempt to standardize on oils and to lower
power losses.
Refined oils are classified by their level of refining in
groups; with Group I oils being supplanted by Group II oils
when used as a turbine oil in the last 15 years or so. The more
highly refined Group II oils have much better resistance to
oxidation but have introduced varnishing potential which will
be discussed later. There are also synthetic turbine oils which
are being used more in recent years. Examples of synthetics are
Polyalphaolefins (PAO) which is a Group IV oil and
Polyalkylene glycols (PAG) in Group V.
Additive packages are also employed with Rust and
Oxidation (R&O), anti wear and anti foam packages being
widely used. Extreme Pressure (EP) additive packages are
being used less in recent years as they have been found to have
the potential to plate out under high heat.
CONDITION MONITORING
While this tutorial is not about condition monitoring it is
helpful to list the common tools used today. Once listed we will
then move onto bearing failures and how these tools may be
used to determine if an operating machine is experiencing a
bearing failure and how to utilize the information to guard
against failure in the future. One important aspect of all of this
is understanding what is happening or what happened in an
attempt to take steps to increase machine reliability going
forward.
Common Condition Monitoring procedures today include:

Figure 6 ASTM B23 Alloy 2 Yield Strength

A plot of babbitt yield strength vs. temperature is presented

in Figure 6. This data comes from ASTM B23 and has been
extrapolated to 300 F. The yield strength is also presented
based upon a 0.125% offset (0.125% deformation of gage
length). Yield data is presented at a 0.2% offset from time to
time and of course that results in higher values for the yield
strength (i.e. 0.125% offset vales are more conservative).
Oil

Proximity probe use Radial and axial

Oil analysis
Bearing Temperature
Grounding brush discharge monitoring

Taking each:
Proximity Probes
Proximity (prox or eddy current) probes sense the rotors
location relative to the surface the probe is mounted on
(typically somewhere on the machine). They can provide

Copyright 2015 by Turbomachinery Laboratory, Texas A&M Engineering Experiment Station

vibration or steady state data. Typically radial probes are

located near journal bearings (2 per bearing location) and are
situated orthogonal (90 degrees apart) so shaft orbits can be
observed. These probes can then be utilized to report
subsynchronous, synchronous or supersynchronous vibration.
Shaft centerline plots
Prox probes can also be used to track rotor movements,
such as when a journal comes up on a film of oil; this is
referred to as shaft centerline data. This use is often overlooked
and when it is desired to analyze shaft centerline data it is
useful to know where the journal is in the bearing at zero speed
(at the bottom of the bearing) as a reference point. Care must be
taken when using data from these probes as thermal affects may
alter the probe length and/or the mounting surface location
relative to the bearing centerline.
Proximity probes can measure change in radial shaft
position with respect to the bearing bore. As such, the data from
the probes can be used to detect bearing wear or other changes
in condition. It is assumed that when the shaft is not rotating,
the shaft is positioned centered in the bottom of the bearing.
There are exceptions to this guideline, such as when the
coupling end of a light weight pinion in a gear drive is held off
the bottom of the bearing by the coupling radial stiffness and
possible alignment offset. Or in some overhung machines
where the coupling end bearing is uploaded.
When the shaft begins to rotate, it is lifted by the oil wedge
and is displaced laterally in the direction of rotation (see Figure
7). Other influences, such as misalignment and gear tangential
and separation forces can also cause movement of the shaft in
the bearing clearance.

Figure 7 Illustration of Proximity Probe Layout and Shaft

Movement for a Counterclockwise Rotating Shaft (Eisenmann and
Eisenmann, 1997)

Proximity probes are essentially electronic dial

indicators. The change in the DC voltage output from the
probe system is proportional to the change in distance from the
probe to the shaft. The distance the shaft has moved toward or
away from the probe is determined by measuring the change in
the DC voltage values and applying the probe sensitivity or
scale factor (usually 200mV/mil). With two probes mounted
orthogonally, a vector summation of the movement measured

on the two probes determines the actual change in shaft

position relative to the probes. The vector summation describes
the magnitude and direction of the change in shaft position
from the initial state.
High speed data acquisition systems can record the probe
DC components and show changes in shaft position in some
detail. Figure 8 is a plot of continuous sampling of DC gap
voltages and resultant shaft movement during a start up for a
machine with clockwise rotation. The initial point is with the
shaft resting on the bottom of the bearing and all the changes in
position are relative to that shut down condition. This machine
was equipped with tilt pad bearings, and this direction and
magnitude change in shaft position is considered normal for
this type and size bearing. (Eisenmann and Eisenmann 1997)

Figure 8 Shaft Position Change during Start Up for a Machine

with Clockwise Rotation (Eisenmann and Eisenmann, 1997)

Shaft centerline data can be used to evaluate changes in

radial shaft position in relation to other variables as well as
speed. Position change with respect to time while the machine
is at a steady operating state can indicate whether radial bearing
wear, or some other phenomenon that causes a change in radial
shaft position, is occurring. Even without a high speed data
acquisition system, logging the DC probe gap voltages with the
machine shut down and/or at full speed and load can show
changes in bearing condition. If the gap voltages with the
machine in shut down condition show an increase in DC value,
perhaps there is wear occurring in the bottom of the bearing. If
the gap voltages at normal operating conditions indicate a
change, perhaps bearing wear due to shaft alignment or some
other force is causing wear of the bearing, or at least a change
in shaft position that should be evaluated.
One particular example of hand logged gap data used for
diagnosis of bearing wear was on a turbine generator used in
peaking service. The hand logged DC voltage data showed a
somewhat rapid decrease in probe gap voltages (shaft moving
toward the top of the bearing) on the number one (front
standard) turbine bearing. The data was logged daily by the
machinery engineer at the plant because, although the radial
shaft vibration was low at the bearing, the casing vibration was
slightly elevated and just didnt feel right. The engineer had
just read an article on use of shaft centerline data to detect
bearing wear and decided to give it a try. All indications from
the gap voltage data were that the wear was occurring in the top
of the bearing. Plant management did not believe the data or the

Copyright 2015 by Turbomachinery Laboratory, Texas A&M Engineering Experiment Station

conclusions, even though the data indicated almost 40 mils of

wear. The shaft vibration eventually did increase to high levels
(as the casing vibration decreased), and the turbine generator
was shut down for inspection. There was indeed in excess of 60
mils babbitt lost from the top of the bearing caused by
misalignment of the bearing low relative to the turbine during
maintenance at the previous shut down.
Caution should be used when evaluating probe gap voltage
data, as thermal effects (heating of the probe or bearing area)
can cause thermal growth of the probe mount relative to the
shaft. This thermal growth can mimic shaft movement in the
bearing. It is advisable to examine the probe and bearing areas
for possible thermal influences on the probe mounts before
evaluating the data. Proximity probe holders (stingers) are
also subject to some change in length from thermal effects. A
316 stainless steel, 12 inch (0.3 m) long probe holder will
increase in length 0.005 inch (5 mils 0.13 mm) as the bearing
area increases in temperature from a nominal ambient of 70oF
to 120oF (21-49oC). The bearing housing also experiences
thermal growth, so determining the actual effects of thermal
growth on the shaft centerline plots can be quite complex. The
best solution is to take a lot of data at different operating
conditions so that normal and abnormal conditions are
easily discerned.
One plant that recorded shaft vibration, bearing
temperatures, and probe gap data in the plant data historian
were examining the historical data to determine if a bearing
inspection was warranted at the next maintenance shutdown.
When the data for the last two years was evaluated, a distinct
sawtooth oscillation of about 0.20.4 VDC (1-2 mils) was
seen on some of the probe gap trends, even though the mean
value of the trend was constant. This effect was very
pronounced in the summer months, and the period of the
sawtooth oscillation was approximately 24 hours. The cause of
the oscillation was that the probes in question were on the sun
side of the compressor train, and the change in gap voltages
was due to solar thermal heating of the probe holders.
Axial position
Proximity probes are used to measure rotor axial position
in the machine casing. This can yield valuable information on
the health of the thrust bearings. Usually since only DC data is
desired (not vibration only gap voltages which yield rotor
axial location) the surface the probe observes is not considered
critical. However when diagnosing a machine problem every
piece of information is important; so if the target area is treated
like a radial probe surface it is possible to get axial vibration
data from these probes. Granted axial vibration problems are
rare but if you have a good surface for the axial probes to
observe you can at least rule that out.
Since there is only one axis of motion, the calculation of
rotor position change relative to the bearing is less complicated
than with two orthogonal radial probes. Thrust position
monitors in a permanently installed vibration rack display the
change in rotor position relative to some zero point established
when the machine was shut down.
In order to know how much the rotor position has changed
(and possible thrust bearing wear) it is necessary to know
where the rotor was positioned when the thrust monitor was

zeroed. Some users set the thrust monitor zero point with the
rotor centered in the thrust clearance. Some users set the thrust
monitor zero point with the rotor positioned against active
thrust bearing surface. Whatever convention is used, it must be
known when evaluating thrust position changes. If the zero
point was set with the rotor in the midpoint of a 12 mil (0.012
inch - 0.3 mm) thrust clearance and the thrust position monitor
indicates 6 mils in the normal direction of thrust, the thrust
bearing is probably not worn. If given the same monitor set up
and clearance, the thrust position monitor indicates 25 mils in
the normal direction of thrust, some bearing wear has likely
occurred. It is always good to compare thrust bearing
temperatures to thrust position values to verify that the
measurement is valid.
As with radial position measurement, there are factors that
can affect the accuracy of thrust position measurement.
Thermal growth of the thrust bearing housing is one possible
effect. If the rotor was not positioned correctly in the center of
the clearance, or the bearing assembly was not properly
positioned with the rotor against the active thrust face, the
thrust position reading will be affected. Some movement of the
entire thrust bearing assembly in the housing due to improper
shimming or assembly has also occurred and will shift the
operating thrust position. The best way to incorporate all of
these variables into the thrust position evaluation is to record
the operating thrust position after a thrust bearing replacement
when the machine reaches normal operating speed, load, and
temperature. The initial startup is usually a true zero point, as
the thrust bearing has most likely not experienced any
significant wear.
Orbit Plots
The timebase signals from orthogonal proximity probes
used to measure radial vibration can be combined with a onceper-turn phase reference to produce an orbit plot, which
provides a graphical representation of the shaft motion in the
bearing (see Figure 9). Alarm and shutdown limits for radial
shaft vibration are sometimes established as a percentage of
bearing clearance in order to avoid bearing damage from
excessive shaft radial movement. Orbit shape is often elliptical
rather than round, so it is advisable to set the alarm and
shutdown using the amplitude of the major axis of the ellipse
(Smax), since that is the highest amplitude motion that could
cause bearing damage. The Smax calculation is available as a
standard function on many modern monitoring systems.

Figure 9 1X Filtered Orbit (Eisenmann and Eisenmann, 1997)

Bode, Polar, and Cascade Plots

Copyright 2015 by Turbomachinery Laboratory, Texas A&M Engineering Experiment Station

Vibration data acquired during machine start up and shut

down (transient data) can also be an indicator of radial
bearing condition. The Bode plot (see Figure 10) is a
representation of synchronous rotational speed (1X) vibration
amplitude and phase angle plotted as a function of machine
speed. The polar plot (see Figure 11) is the same data presented
in a Nyquist format. The Bode plot is helpful in detecting
changes with respect to speed, the polar plot shows the change
in phase angle and physical orientation in a more easily
interpreted format. Cascade plots (see Figure 12) are a data
format where spectrum plots acquired at speed intervals during
start up or shut down are plotted on the same graph.
The vibration response through the startup and shut down
operating speed ranges for machines with hydrodynamic radial
bearings is affected by the design and condition of those
bearings. If radial bearing damage has occurred, it will likely be
reflected in changes to the vibration characteristics during start
up and shut down, especially as the machine traverses a
resonance or critical speed. Changes to measurements such as
amplitude, amplification factor, relative phase, and the
appearance of subsynchronous frequency components are all
indicators that something has affected the rotor/bearing system.
If the bearing is worn and the wear has affected the stiffness or
damping characteristics, it is usually detectable by comparison
of the startup or shut down data in the Bode and polar plots
with previously acquired data. Cascade plots are useful in
detecting faults such as oil whirl or whip, which are often
related to bearing design or damage. Combining the transient
data with shaft position and bearing temperature data gives a
broad range of information with which to analyze potential
bearing problems.

Figure 11 Polar Plot (Eisenmann and Eisenmann, 1997)

Figure 12 Cascade Plot (Eisenmann and Eisenmann, 1997)

Oil analysis

Figure 10 Bode Plot (Eisenmann and Eisenmann, 1997)

Oil analysis can yield information that will provide the

early detection of bearing failures and the ability to proactively
prevent potential failure mechanisms. This is all accomplished
by attacking oil analysis from three different directions.
First, lets look at the health of the oil. This is
accomplished by examining certain lubricant properties and
comparing them to base line data. It involves testing the oils
viscosity, additive elements and acid concentration levels. This
can be accomplished through a series of tests.
Viscosity: ASTM-D445 Viscosity is, of course, the
most important property of oil. It will determine whether the
oil film is viscous enough to separate the bearing surface from
the shaft under the current operating temperatures. Typical
allowances are +/- 10% of the baseline value of the oil.
Changes of +/-5% should be closely monitored for
contamination or chemical changes.
Additive Elements: ASTM D5185 Certain elements
are categorized as additives and should be trended to monitor
their depletion as the additives are consumed. Some of these
additives help protect the surfaces of the bearing and shaft from
wear and corrosion.

Copyright 2015 by Turbomachinery Laboratory, Texas A&M Engineering Experiment Station

 Acid Concentration: ASTM D664 Total Acid

Number or the acid concentration is one method of determining
the onset of oil oxidation. This method of determining oil
oxidation is usually an indication that the oil is approaching the
end of its useful life and that potential coking or varnishing
may have already taken a toll on the bearing and shaft surfaces.
Additional testing methods, such as the monitoring of
antioxidants, should be considered as well to fend off the
potential for oxidation.
Secondly, the oil should be tested for wear materials and
wear rates. It is important to consider wear rates and not just
limitations on the quantity of wear elements. Wear rates
provide a means to which the analysts can determine how
severe a particular bearing problem is. This type of trending
analysis is greatly improved by the sampling frequency (e.g.
monthly sampling based on a 24 hour a day operation). The
recommended testing for detecting wear is:
Wear Elements: ASTM D5185 The known wear
elements of ASTM B23-00 Alloy 2 Babbitt material are known
to be Tin, Antimony and Copper with an emphasis on Tin. The
ICP spectrometer can be used to monitor the wear materials
produced by the bearing in the range of 5um and below. It will
reveal traces of these elements based on the percentage of each
element in the Babbitt. In order to properly analyze the amount
of wear that the bearing produces, it is important to track
system top-offs and oil changes to accurately analyze wear
patterns. Keep in mind wear particles greater than 5um are not
detectible. If wear is suspected but not revealed through ICP,
consider using large particle spectroscopy or analytical
ferrography.
Finally, the oil should be tested routinely for
contamination. Particulate, water, use of the wrong lubricant,
process contamination and varnish can have detrimental effects
on the health of the oil as well as the bearing. Most of these
types of contaminants can be detected with the following tests:
Particle Counts: ISO 4406 Performing particle counts
are an extremely important part of proactively eliminating the
potential wear mechanisms that lead to early bearing failure
and/or shortening the health of the lubricant. Establishing ISO
cleanliness levels and maintaining them with the proper
filtration devices will improve the reliability of the bearing.
Particle counts can also be used as good indicators of
accelerated wear and possible filter bypass conditions.
Karl Fischer: ASTM D6304 Water concentration
levels in oil are a primary cause of sludge buildup and
increased oxidation rates. Maintaining water levels well below
the oils saturation rate is important to reduce varnish and
sludge buildup on bearings. Online water detection is preferred
over KF when an oil system size and criticality warrants it.
Elemental Analysis: ASTM D5185 Elemental
analysis has been mentioned twice for monitoring elemental
additives and wear elements. ICP can also be a useful tool in
detecting sources of contamination. It can serve as a good
indicator of lubricant cross contamination and identify sources
of internal and external contamination.
Varnish detection and mitigation has become an industry
upon itself with the introduction of Group II base oils. Many of
the routine tests used to monitor oil quality cannot detect
varnish problems until it is too late. It has been a source of
many bearing failures similar to the one mentioned later in this

tutorial. This section only deals with ways to discover the

presence of varnish and not the mitigation of the problem. If
varnish is discovered there are ways of treating the problem
through special types of filtration. Ultimately, the source of the
heat generation should be located to eliminate the problem.
The following tests have shown promise in the early detection
and the onset of varnishing.
The potential for varnish build-up often correlates with
a decrease in an oils antioxidant levels. Monitoring these levels
have shown to provide a source of early detection. The
monitoring of oxidation inhibitors (Amines & Phenols) can be
accomplished with the use of FTIR and/or RULER. To
perform these tests properly, it is important to provide a base
line sample before monitoring depletion rates.
Typical
depletion limits are 25 to 50% of the baseline before a partial
change out or a complete oil change out is required. The need
for frequent change outs could be a key indicator there is an
internal heat source causing the accelerated depletion of the
antioxidants.
QSA If the presence of varnish already exists within
the system, a well known test developed by Analysts Inc. can
be used to detect the severity of the varnish or sludge that
exists. Analysts Inc developed a color patch test (QSA) that
correlates the contrast of a sample patch to a numerical severity
of the varnish or sludge that plagues system. It also provides a
patch weight that can be monitored and trended as well.
Before applying these practices, it is imperative that the
physical act of obtaining a representative sample by a trained
individual is accomplished. If this is ignored, much of the data
gathered will be misleading in performing a final diagnosis. To
determine the best location for sampling a bearing, the method
of oil application must be understood. (e.g. In force fed
systems it is best practice to set up a primary and secondary
sample point. The primary sample point should be located on
the common oil return line before it empties into the oil
reservoir. The secondary sample point should be located on the
dedicated return line of each bearing to further narrow down
wear and contamination related problems).
Much of the analysis data stated above is geared around
building a primary testing slate for routine testing. Secondary
test methods may be required to further troubleshoot negative
sample results.
Bearing temperature
This is probably the most critical indicator of bearing
health. As pointed out in the babbitt section above babbitt loses
strength at elevated temperatures so it is best to keep bearings
running cool. As will be presented later one can take a known
temperature and calculate an estimated maximum oil film
pressure and then compare that pressure to the strength of the
babbitt at that temperature. The location of the temperature
sensors is also important; it is not necessary to know the point
in the bearing of maximum temperature but it is important to
know the temperature at the most distressed location. The most
distressed location is determined by analysis of the temperature
coupled with the applied hydrodynamic pressure. Because of
this understanding API 670 specifies very good locations for
temperature sensor placement.

Copyright 2015 by Turbomachinery Laboratory, Texas A&M Engineering Experiment Station

Note that these are very good standard locations for sensor
placement. They closely represent the most distressed location
in the bearing and the standardization allows one to effectively
analyze the condition of the bearing. For instance if there was a
sensor placed at the center of the pad there would need to be
some of correlation required to estimate the temperature at the
API location.
Grounding brush discharge monitoring
Figure 13 API 670 Figure for TPJ Instrumentation

Proper thermocouple or RTD placement in a 5 pad tilting

pad journal bearing oriented in a load between pad (LBP)
configuration id denoted in Figure 13. Note the instrument is
located very close to the babbitt bond line and 75% in
circumferentially from the pad leading edge. The 75% location
may not represent the hottest portion of the pad surface but it
does closely represent the most distressed location taking into
consideration temperature and film pressure. Also note the
instrument should not compromise the bond line as dimpling of
the babbitt surface is possible.

In recognition of the possibility of static charge build up in

condensing steam turbines API 612 (2005) specifies that
grounding brushes be installed. While it is beyond the scope of
this presentation it is important to know that the electrical flow
to ground through these brushes can be monitored and useful
information can be extracted.
Grounding brushes take current from the rotor to ground so
that a charge does not build up on the rotor to the point where it
discharges to ground though the best path possible which is
usually the closest point between the rotor and stator which is
usually (hopefully) the point of minimum film thickness in a
bearing. Typically this point of minimum film thickness is
found in the active thrust bearing (as will be shown later).
Shaft grounding brushes serve two purposes. The brushes
are able to transmit modest amounts of stray current to ground
for prevention of arc damage through parts of the machine
(especially the bearings). The brushes also permit measurement
of the shaft voltage and current, which allows assessment of the
electromagnetic condition of the machine.

Figure 14 API 670 figure for LOP instrumentation

Figure 14 is for Load on Pad (LOP) configurations. There

is also a similar figure for sleeve bearings. Figure 15, also from
API 670, denotes the proper location of temperature
instrumentation
in
tilting
pad
thrust
bearings.

Figure 16- Typical Grounding Brush Installation (Sohre, 2010)

Figure 15 Temperature sensor placement in thrust bearings

There are two types of electrical currents in rotating

machines: electrostatic and electromagnetic.
Electrostatic currents are primarily generated by impinging
particles or droplet atomization (such as in wet stages of steam
turbines) and can generally be handled with one grounding
brush per shaft. Electrostatic currents usually cause minor
damage, such as frosting of bearing babbitt, and the progress of
damage is relatively slow.
Electromagnetic currents are usually due to residual
magnetism and/or stray currents created by electrical machines,
such as generators, motors, and exciters. Events such as
machinery rubs, improper welding, or lightning strikes can
induce residual magnetism in the machine parts and cause the

Copyright 2015 by Turbomachinery Laboratory, Texas A&M Engineering Experiment Station

rotating machine to become a generator. Electromagnetic

currents can be extremely destructive. Severe damage can occur
rapidly because of high current density. Damage such as burned
areas and welding of components are possible.
Monitoring to determine shaft current level and changes is
performed using a shaft voltage current monitor (VCM). The
VCM measures voltage and current through the grounding
brushes. Ideally, the output of the VCM should be connected to
a recording device or plant data historian so that trends of the
voltage and current can be used to detect changes in shaft
electrical properties. If readings increase slowly, it could be an
indication of deterioration of the electrical machinery (motor,
generator, or exciter) or possible progressive self-magnetization
of the machine. Sudden increases in current may indicate that
self-magnetization has occurred, possibly due to the reasons
previously mentioned.

Varnish build up
Electrostatic discharge damage (frosting)
Electromagnetic discharge damage (Spark tracks)
Oil burn or additive plating due ot high temperatures
Loss of bond between babbitt and base metal
Chemical attack
Pivot wear in tilting pad bearings
Unloaded pad flutter
Cavitation damage

This is a relatively short list but it does contain the vast

majority of failures observed in actual operation. Some of the
more common of these are presented below:
Low speed wiping and fatigue
Figure 18 is a photograph of a bearing that exhibits slight
low speed wiping coupled with at-speed fatigue damage,
brought on by bearing to journal misalignment.

Figure 17 Shaft Voltage Current Monitor Installation on a

Turbine Generator (Nippes and Galano)

If readings from the VCM decrease to zero, the grounding

brush has most likely lost contact due to wear, or some other
part of the brush circuit has opened.
More detailed analysis of the shaft currents to determine
cause of excessive shaft currents can be performed using an
oscilloscope. In-depth discussion of those techniques is beyond
the scope of this tutorial.
Other tools
There are also other condition monitoring tools available
such as:

Infrared Thermography
Acoustic Emission (AE) analysis
Motor Current Signature Analysis (MCSA)
Machine performance monitoring

The more information analyzed the more accurate the

diagnosis and prognosis.
BEARING FAILURES
This section will discuss bearing failure mechanisms and
introduce possible monitoring activities that could detect the
failure. First of all there are literally dozens of ways bearings
can fail. Some of the more common include:

Babbitt fatigue
Babbitt wiping due to rotor to stator contact
Babbitt flow due to high operating temperatures
Foreign particle damage

Figure 18 Wiping and fatigue/misalignment

Here we see that the bottom portion of this bearing shows

an even wipe (or polish) pattern from low speed operation
(operating in the boundary lubrication regime). We also see
fatigue damage; however this is presenting itself only on the
right side of the bearing indicating that the bearing became
misaligned once up to speed.
Recall that when a sleeve bearing is operating in the full
film lubrication regime a film of oil separates the journal from
the bearing and the hydrodynamics pressure profile puts the
operating position of the journal off at an attitude angle. That is
why the babbitt damage is downstream from bottom dead
center.
The user theorized that as the machine became loaded and
thermally soaked that the bearing pedestal moved axially,
actually tilting, thereby edge loading the bearing. This edge
loading resulted in basically half the bearing taking the load
causing high babbitt temperature and high peak film pressure.
Babbitt fatigue is caused by dynamic loads on the babbitt
surface. Typically in bearings of this type the dynamic loads are
caused by vibration and result in peak film pressure
fluctuations. Cracks initiate on the babbitt surface and
propagate radially towards the bond line. As the cracks get
closer to the bond line the strength of the backing material
reinforces the babbitt and causes the cracks to turn and spread

Copyright 2015 by Turbomachinery Laboratory, Texas A&M Engineering Experiment Station

circumferentially, meeting with other cracks and dislodging

pieces of babbitt. This is conceptually illustrated in Figure 19.

Loss of bond between babbitt and base metal

This is most common when babbitt is applied to copper
alloy. Copper is used extensively to reduce bearing operating
temperatures by allowing a significantly increased heat transfer
coefficient. Most bearing companies use a copper alloy that
has good stiffness and strength while still having good heat
transfer properties. Figure 23 is a photograph of a sheet of
babbitt that came off of a 16 journal pad.

Figure 19 Fatigue crack propagation

A close up of the babbitt surface (Figure 20) shows

portions of babbitt missing and a section showing only surface
cracks.

Figure 23 Loss of bond with copper pads

Figure 20 Close up of fatigue damage

Babbitt fatigue strength can be increased by utilizing

thinner babbitt (to take advantage of the backing material
strength) and by keeping the babbitt temperature low.

Figure 21 Babbitt Fatigue

Strength vs. Temperature

Figure 22 Babbitt Fatigue

Strength vs. Thickness

Figures 21 and 22 demonstrate the effect of these

parameters on fatigue strength. These are based upon dynamic
tests with 2000 psi mean load. Obviously for a running machine
changing babbitt thickness is not an option, but taking steps to
reduce temperature and vibration often times are available.

It has been discovered that copper has a strong affinity for

tin and this is magnified at elevated temperatures. When babbitt
is applied to bearings the process includes a tinning operation
so the tin adheres to the base metal and then the babbitt bonds
to the tin. Should the tin diffuse into the copper the bond will
become weaker and weaker, and brittle. Oddly enough this is
often times not discovered during normal NDT inspection (such
as visual, PT or UT) since the bond is intact, it is just
compromised and measures of bond strength are destructive.
It has been discovered that applying a barrier layer prior to
babbitting can eliminate this issue. A material must be selected
that has good bond strength to copper and the tin will adhere to,
while providing a barrier to tin migration into the copper.
This failure mode is not possible to predict with
conventional condition monitoring tools. As such the best way
to address this issue is to verify the copper backed bearings,
new or repaired, have a proven barrier layer. If there is a
question as to whether the barrier layer exists it is
recommended that during the next outage this is addressed.
Also while the primary authors company has seen dozens of
bearings that have had this problem none of them have failed in
service; it is theorized that the brittle bond is not challenged
during normal operation.
Electrostatic discharge
As mentioned earlier it is known that rotors can build up a
static charge and this charge will jump to ground through the
easiest path available. Preferably grounding brushes are utilized
and these brushes take the charge off the rotor to ground in a
controlled way. If there are no grounding brushes or they are
not working properly then this charge can go to ground through
the location where the rotor is closest to a grounded stationary

Copyright 2015 by Turbomachinery Laboratory, Texas A&M Engineering Experiment Station

This damage can be avoided with a lower stress pivot

design and/or control of vibration levels.

part.

Oil coking
From time to time oil analysis may indicate that the oil is
oxidizing and/or an additive package content is diminishing.
This could also be accompanied by slight rotor position moves
and is usually present when there are elevated temperatures.

Figure 24 Thrust pad with electrostatic discharge damage

A pad exhibiting signs of electrostatic discharge is shown

in Figure 24. Note that with a bearing like this the point of
minimum film thickness is typically on the upper trailing edge
as shown, Note that the babbitt has been spark eroded away
starting in the minimum film thickness corner and working its
way down the pad, as the pad shape changes.
This damage can be avoided by ensuring there is adequate
grounding.
Pivot wear
Since pads in tilting pad journal and thrust bearings
actually tilt they can exhibit damage at the pivot interface(s)
that can affect the performance of the bearing. Tilting Pad
Journal (TPJ) bearing outer shell and a pad are shown in Figure
17. Note the pad has been flipped over so we can see the
mating surfaces. Also note the extreme wear on both surfaces.
Damage can be cased by dynamic load being transmitted across
this interface (typically brinnelling damage) and by relative
motion of the pad to the shell (typically fretting). Either or both
of these wear mechanisms can act to wear the pivot interface
which can manifest itself as an opening up of the bearing
clearance, possible resulting in increased vibration.
One way to detect the possibility of this happening is by
monitoring the shaft centerline position for journal bearings and
axial position for thrust. If moves are detected and there is no
tin found in the oil then one may conclude there is a possibility
of pivot wear. For journal bearings this could also be
accompanied with increased vibration (due to loss of stiffness
and damping with the increased clearance).

Figure 26 Oil coking

Figure 26 illustrates this phenomena. Note that the

presence of this coke is typically located in the area of the
bearing exposed to high film pressure coupled with high
operating temperatures; also note that the 75-75 location does
seem to represent the most distressed portion of the pad.
This damage mechanism can be avoided with careful
selection of additive packages, monitoring of oil analysis and
control of operating temperature.
Varnish
As mentioned earlier the formation of varnish on lubricated
surfaces has increased over the last 15 years or so as the more
highly refined Group II oils have been utilized. At first glance it
may appear as though the damage is similar to coking but
further analysis will show that varnish deposits can be found on
any surface in contact with the oil, high temperatures and
pressures are not required.

Figure 27 Varnish

Figure 25 Pivot wear

Figure 28 Varnish

A severe case of varnish formation is depicted in Figures

27 and 28. Note how the varnish is not isolated to the hot spot
of the thrust pads but can be seen on the retainer, links and even
the back side of the thrust pads.
This damage can be avoided by being aware of varnish
potential as predicted by oil analysis and taking steps to
mitigate the phenomena. Signs that you may be experiencing
varnish deposits include rotor position changes (due to the
deposits getting thicker) and temperature changes due to the

Copyright 2015 by Turbomachinery Laboratory, Texas A&M Engineering Experiment Station

insulating behavior of the deposits and the thickness of the

deposits resulting in reduced clearances.
Note that even worse side effects could be found where the
oil is used for other purposes in the machine (such as servo
valves, etc). In these cases the tacky varnish coated surfaces
can result in sticky operation of the components in question.

same. However the actual machine had 5 mils vibration on one

bearing and 1 mil on the other. It was determined that an
unbalance between bearings could not duplicate the vibration
spread observed.

ROBUST BEARING DESIGN

Case study 1
This case study involves a large centrifugal compressor in
charge gas service in an ethylene plant. The unit was running
fine when a power outage shut the plant down. The train came
down fine but when brought back up it was noted that one
bearing exhibited high levels of vibration.
This is the low pressure A case compressor located
between the steam turbine driver and a gearbox which drives
the medium pressure B case compressor. Alarm and trip
levels were set to 3 and 5 mils respectively, and the compressor
normally runs around 4750 rpm. The bearing in question is a 6
5 pad LBP bearing with an L/D ratio of 0.5 loaded to 225 psi.
The low pressure compressor outboard bearing vibration
had increased from under 1 mil of vibration pre-outage to over
4 mils vibration after coming back up; this vibration was
substantially synchronous, had a forward precession, a round
orbit and was speed dependant, there was also a slight phase
shift. All of this (and other data, history and condition analysis)
pointed to a balance condition change on that end of the
machine.
There was a lot of speculation on how the rotor became
unbalanced however rotor dynamic analysis indicated only a
coupling unbalance (or other similar synchronous forcing
function) could cause the vibration levels observed.

Figure 30 Rotor deflection shaft with coupling unbalance

As shown in Figure 30 when applying an unbalance to the

coupling we find the roughly 5 to 1 variation in vibration as
observed from one end of the rotor to the other. It was
determined that somehow the coupling became unbalanced
even though the vibration on the other side of the coupling
remained low. This analysis also replicated the speed
dependency found in the actual machine.
Now the focus shifted to determine if continued operation
would lead to an unplanned shutdown due to bearing failure.

Figure 29 Rotor deflection due to imbalance

Figure 31 Bearing force due to coupling unbalance

The rotor model the rotor deflection shape from an

unbalance at the first wheel are presented in Figure 29. Note
that this is a drive though machine so there is a coupling on
each end. Also note that the analysis predicts that for this
unbalance the vibration at each bearing should be about the

The dynamic force imparted on the bearing due to a

coupling unbalance sufficient to raise the vibration to 5 mils is
depicted in Figure 31. Note the significant dynamic load (which
is zero to peak). The bearing had (roughly) a static load (journal

Copyright 2015 by Turbomachinery Laboratory, Texas A&M Engineering Experiment Station

load) of 4000 pounds and a dynamic load of +/- 20,000 lbs.

Analysis of the steady state condition of the machine
predicted a maximum film pressure of 750 psi so one could
estimate that the dynamic pressure was 3750 psi. Analyzing
this as a compressive stress of 750 psi and an alternating stress
of 3750 psi we can then look at the babbitt compressive
strength at 180 F of 4000 psi (see Figure 6) and realize we are
very near failure.
It was decided to reduce speed to maintain vibration levels
at or below 4.5 mils and continue to run. This of course resulted
in reduced plant output put did allow the machine to continue to
run. Financial analysis dictated that they run for another 3 years
under this constraint. The bearing health was monitored by
observing shaft centerline data, bearing temperature readings
and performing frequent oil analyses.
An alignment consultant was brought in and determined
that there was an offset misalignment that would have
presented itself as a synchronous force on the rotor. When the
machine came down they found that the coupling had spun and
friction-welded itself to the rotor this caused the
misalignment and change in balance condition. It was theorized
that there may have been a significant liquid ingestion when
coming back up after the power outage.
What did this bearing have going for it that allowed
continued operation for 3 years under this extreme dynamic
loading condition? The OEM bearing was replaced a long time
ago with a bronze and babbitt bearing with ball & socket (B&S)
pivots. The B&S pivots deteriorated very little over this run
thanks to the significantly lower stresses compared to non ball
& socket designs. The user is convinced they would not have
been able to run this long under these conditions with the
originally supplied bearings. The fact that the bearing was
running at only 180 F was also a major factor as this meant the
babbitt strength was still relatively high.
The bearings were removed and inspected and it was found
that the lower two pads had opened up about 2.0 mils due

Figure 32

Figure 33

1 mil.
Case History 2
Nova Chemicals company operates 3 large ethylene plants
at its Joffre facility in Edmonton, Canada.
It was decided that the length of their runs needed to be
extended so they undertook a program to address items that
could affect that goal. One area they knew that had an issue was
with the bearings in the main turbine compressor trains
(Ethylene, propylene and charge gas). The existing bearings
had substantial pivot wear, resulting in increased clearances and
ultimately increased vibration. The decision was made to
upgrade the bearings to Ball & Socket design with thin babbitt
(for fatigue strength) on bronze backing. The B&S afforded
significantly reduced pivot stresses and the thin babbitt on
bronze yielded an increase in babbitt fatigue strength (see
Figure 22).
In total 8 turbines and compressors were retrofitted with
more robust TPJs and one journal bearing in the gas turbine
was upgraded from a sleeve bearing to a TPJ. Most of these
bearings have gone through 2 outages now; one at 4 years, one
at 5 and the plan is to now run 6 years. Of the 17 journal
bearings upgraded the worse of them had an increase in
clearance of 0.6 mils and all the rest opened up less than 0.5
mils. So after 9 years of operation they cleaned up, checked and
reinstalled all 17 bearings. The figures below represent what a
few of these bearings looked like after cleaning.

Figure 37 201 compressor after 9 years operation

Figure 34

Figure 28 601 bearing after 9 years in operation

Figure 35

Figure 36

to some pivot wear and babbitt loss. Figure 32 is a photo of the

bearing assembly; Figure 33 shows the bottom two pads; Figure
34 the top pads; Figure 35 the lower pad pivots and Figure 36 is
a close up of the pivot wear. The bearing on the other end had
the same bore as when it shipped 11 years earlier.
The rotor, bearings and coupling were replaced and the
machine came back up with vibration levels on both ends under

Figure 39 201A Bearing after 9 years in service

Copyright 2015 by Turbomachinery Laboratory, Texas A&M Engineering Experiment Station

& Sons.

CONCLUSIONS
Babbitted bearings theoretically can have an infinite life
when utilized in ideal situations; unfortunately the real world is
not ideal. As such the rotating equipment engineer must take
whatever data he has available to determine the health of his
bearings. Knowledge of possible failure mechanisms, past
history of the machine, oil analysis, real time vibration, shaft
position and bearing temperature monitoring and other tools
can provide significant information to aid in this determination.
Steps taken to extend the time between outages can and
should include evaluation of the robustness of the bearings
installed and may include bearing upgrade projects to increase
bearing reliability.

REFERENCES

API 612, 2005, Petroleum Petrochemical and Natural Gas

IndustriesSteam TurbinesSpecial-Purpose
Applications (6th Edition), Washington D.C, API
Publishing Services.
API 617, 2002, Axial and Centrifugal Compressors and
Expander-Compressors for Petroleum, Chemical and Gas
Industry Services (7th Edition), Washington D.C, API
Publishing Services.
API 670, 2010, Machinery Protection systems (4th Edition),
Washington D.C, API Publishing Services.
ASTM B23-00, 2010, Standard Specification for White Metal
Bearing Alloys (Known Commercially as "Babbitt
Metal).

Vance, J. M., Zeidan, F. and Murphy, B, 2010, Machinery

Vibration and Rotordynamics, John Wiley & Sons.
Wilcock, D. F. and Booser, E. R., 1957, Bearing Design
and Application, 1st Edition, McGraw-Hill Book
Company, Inc.

ACKNOWLEDGEMENTS
Being a tutorial the authors thought it may be best to list
some reference materials that they have found invaluable
through the years. Everything cited here is listed in either the
reference section or the bibliography.
A great overall book on rotating machine health is the book
by the Eisenmanns (father and son). This book contains a
wealth of information and was the main source of material for
the section on proximity probes under the Condition
Monitoring section.
A couple great books on bearings are the classic Wilcock
and Booser book from 1957 and the newer Applied Tribology:
Bearing Design and Lubrication second edition (2008) by
Khonsari and Booser.
For Rotordynamics there are a few good books but the
most useful to engineers practicing in the field are Dara Childs
Turbomachinery Rotordynamics (1993) and Machinery
Vibration and Rotordynamics (2010) by Vance, Zeidan and
Murphy.
Lastly there is a growing source of information on oil
analysis but good resources include material written by Greg
Livingstone (Fluitec International) and Dr. David Wooton with
Wooton Consulting.

Dowson, D., 1979, History of Tribology, London: Longman

Group Limited.
Eisenmann, R.C., Sr., Eisenmann, R.C., Jr., 1997, Machinery
Malfunction Diagnosis and Correction, Upper Saddle
River, New Jersey, Prentice Hall PTR.
Nippes, P.I., and Galano, E.S., Understanding Shaft Voltage
and Grounding Currents of Turbine Generators.
Sohre, J.S., and Nippes, P.I., 1978, Electromagnetic Shaft
Currents and Demagnetization on Rotors of Turbines and
Compressors, Proceedings of the Seventh
Turbomachinery Symposium, Turbomachinery Laboratory,
Texas A&M University, College Station, Texas, pp. 13-33.
Sohre, T. K., 2010, Shaft Riding Brushes, Type 8.5A
Operating Instructions, Sohre Turbomachinery, Inc.

BIBLIOGRAPHY
Khonsari, M. M. and Booser, E. R., 2001, Applied Tribology:
Bearing Design and Lubrication, 2nd Edition, John Wiley
Copyright 2015 by Turbomachinery Laboratory, Texas A&M Engineering Experiment Station