Hydrogen Induced Stress Cracking (DNV RP F112)
Hydrogen Induced Stress Cracking (DNV RP F112)
Hydrogen Induced Stress Cracking (DNV RP F112)
BY
DATE
JDD
03/05/13
APPROVED
DATE
REV. A01
Page 1 of 25
Introduction
Hydrogen induced stress cracking, or HISC, has gained a lot of attention in the subsea
oil industry over the past 10 years or so because of several very costly failures of
duplex stainless steel (DSS) parts. It is a form of environmentally assisted cracking in
which hydrogen ions from cathodic protection (C/P) are electrostatically charged into a
susceptible metal and embrittle it. Failure may occur at stresses well below the metals
yield strength. Susceptible metals include high strength carbon and low alloy steels,
high strength martensitic stainless steels, and duplex stainless steels. All three of these
alloy families are used extensively subsea under cathodic protection. It is thus very
important that design engineers become familiar with the susceptibility of these
materials to HISC, what their design limitations are, and how the risk of HISC can be
eliminated. Duplex stainless steels are particularly at risk. DNV RP F1122, Design of
Duplex Stainless Steel Subsea Equipment Exposed to Cathodic Protection, was written
to provide guidelines for the safe usage of duplex stainless steels under C/P shortly
after several of the high profile failures had occurred. We will examine this standard in
some detail.
HISC is a form of hydrogen embrittlement (HE) that occurs on subsea parts under
cathodic protection. The exact mechanism of embrittlement has not been clearly
defined yet. It is thought that atomic hydrogen collects at the tip of a microscopic crack
in the metal. Here they can interact with dislocations causing a reduction of bond
strength between the metal atoms. It is also believed that the hydrogen ions from the
electrolysis of seawater are electrostatically charged directly into the cathode without
first forming an adsorbed hydrogen atom. HISC in a susceptible metal depends on the
amount of hydrogen charged into the metal, the mobility of the hydrogen within the
metal, and the location of the hydrogen buildup within the metal.
The amount of hydrogen that is charged into the metal is governed by the cathodic
protection system parameters and the use of protective coatings. The ability of the
hydrogen to move around in the metal is primarily dependent on the microstructure of
the metal. The diffusivity of hydrogen in BCC (ferritic) materials is high, but the total
amount that can be absorbed into a BCC lattice is limited the center of the cubic
lattice already has an atom in it. In contrast the diffusivity of hydrogen in FCC
(austenitic) materials is low, but the total amount that can absorbed into an FCC lattice
is much greater than for BCC (the center of the FCC cube is empty). This makes ferritic
steels much more susceptible to HISC than austenitic steels. Coatings act as barriers to
hydrogen diffusion. If seawater cant come into contact with the metal surface being
cathodically protected, electrolysis cannot occur and no hydrogen ions are formed.
Coatings may mitigate the risk of HISC, but they can never be used as the sole means
for preventing HISC in a susceptible metal.
DATE
JDD
03/05/13
APPROVED
DATE
REV. A01
Page 2 of 25
The risk of HISC occurring in a given metal increases with increasing strength, inclusion
content, stress (applied or residual), and cold work. These provide defects in the
microstructure where hydrogen can accumulate. They also make it easier for the
hydrogen to diffuse into the metal. The entry of hydrogen into a metal part is greatly
enhanced if the part is being strained while under C/P. This is particularly true for DSS.
Strain that occurs under a constant load at ambient temperature and at stresses below
the metals yield strength is called cold creep. Engineers must be careful to limit the
likelihood and extent of cold creep in their designs.
DATE
JDD
03/05/13
APPROVED
DATE
REV. A01
Page 3 of 25
individually hardness tested and met the 35 HRC maximum. The problem was that the
bolting material was heavily banded. Banding describes the appearance of the
microstructure of a rolled product where chemical segregation has occurred (see Figure
1). The chemical compositions of the light and dark colored bands in Figure 1 vary and
consequently their response to heat treatment varies. Properties including hardness will
be different. The spacing between the bands is very small so a Rockwell indentation
spans across many bands and gives an average, bulk hardness value. If we were to
microhardness each band, the results could be very different!
Figure 1: Banding
There are various causes of banding, but all stem from chemical segregation that
occurs during solidification. The elongated bands of course form during the rolling (hot
working) of the bar stock used to make the fastener. Carbon often segregates out
particularly if boron is present. Boron is sometimes added to steels to increase
hardenability. A band with a high carbon content and boron present may have a
hardness equivalent to 45-55 HRC or higher while an adjacent low carbon band may be
20-30 HRC. Rockwell testing a finished fastener such as a stud will not detect this
condition, but will produce an average hardness that may very well meet 35 HRC
maximum. The high hardness bands are very susceptible to HISC where cracks can
initiate under C/P. API Spec 20E addresses banding and requires that fastener
microstructures be checked for excessive banding.
DATE
JDD
03/05/13
APPROVED
DATE
REV. A01
Page 4 of 25
DATE
JDD
03/05/13
APPROVED
DATE
REV. A01
Page 5 of 25
with islands (isolated grains) of austenite (see Figure 2). As might be expected, there
is some alloy partitioning between the ferrite and austenite phases so their
compositions are not the same. Nickel and nitrogen, both strong austenite formers, will
tend to concentrate in the austenite. Strong ferrite formers such as chromium and
molybdenum will tend to concentrate in the ferrite. Both phases will have approximately
the same pitting resistance equivalency number (PREN = Cr% = 3.3Mo% + 16N%).
Ferrite will be the first solid phase to form as a molten DSS alloy begins to solidify..
Austenite will begin to nucleate and grow from the ferrite as the temperature decreases.
Ferrite (Dark)
Austenite (Light)
Ideally the microstructure of DSS at room temperature will contain only austenite and
ferrite, but unfortunately there are a number of undesirable secondary phases that may
be present. DSS are used in the solution annealed and quenched condition.
Undesirable secondary phases such as chi, rho, sigma, carbides, nitrides, or other
intermetallics may precipitate out within certain temperature ranges during cooling after
solution annealing if held for a sufficient length of time within these ranges. The
presence of these miscreants can totally screw up the corrosion resistance and
toughness of our DSS. If we cool the material as rapidly as possible from the solution
annealing temperature, there wont be enough time in the offending temperature bands
for these troublesome secondary phases to precipitate out. We can get too much of a
good thing, however, because too fast a cooling rate may result in less than the desired
amount of austenite in the final microstructure. Cooling from the annealing temperature
must be rapid enough to prevent the precipitation of undesirable secondary phases, but
slow enough to allow sufficient austenite to form. Duplex stainless steels cannot be
hardened or strengthened through heat treatment.
DATE
JDD
03/05/13
APPROVED
DATE
REV. A01
Page 6 of 25
As you would expect, the ferrite and austenite phases have different properties. The
ferrite phase has superior chloride stress corrosion cracking resistance compared to the
austenitic grades hence the ferrite matrix. This permits DSS to be used at higher
temperatures than the austenitic stainless steels in chloride environments. The ferrite is
also stronger than the austenite phase. The austenite phase imparts good general
corrosion resistance and toughness to the alloy. Together they form a natural
composite material.
The permeability of hydrogen in ferrite with its BCC structure is high, but the solubility is
low. In austenite with its FCC structure the permeability is low, but the solubility is high.
Hydrogen that is charged into the microstructure from C/P will tend to accumulate
around defects in the ferrite because of the low solubility in the BCC lattice.
Stresses will accumulate along the ferrite-austenite grain boundaries as a DSS part is
loaded. The ferrite matrix will initially carry the load. The low strength, ductile austenite
will plastically deform first. The austenite will strain harden as it deforms and produce a
complex stress state along the grain boundaries. The susceptibility of ferrite to HISC is
much greater than the susceptibility of the austenite. The ductile austenite will act as an
obstacle to crack propagation. In most HISC failures, brittle crack propagation will be in
the ferrite phase with the austenite phase failing by ductile overload. Understanding
this failure mechanism is the key to the successful use of DSS under C/P.
We can reduce the risk of our DSS part failing by HISC if we minimize the distance
between the austenite grains (called the austenite spacing). This minimizes the
distance that a HISC crack can propagate through the ferrite phase before it is blocked
by a tough, crack resistant austenite grain. Austenite spacing is primarily a function of
composition, hot work, and heat treatment. The greater the hot work, the finer the
austenitic spacing and the greater the HISC resistance. Wrought tubing for instance
has significantly more hot work than a large DSS forging and consequently will be more
HISC resistant.
The HISC resistance of a DSS part is dependent on the orientation of the grain flow in
relation to the tensile stresses it sees in service. We can minimize the risk of HISC by
properly orientating the grain flow within the part. The greatest driving force for crack
growth is transverse to the tensile load direction. By having the grain flow of the part
orientated parallel to the direction of greatest tensile stress, crack growth will occur
perpendicular to the grain flow. This orientation presents the greatest number of
austenite grains (consequently the smallest austenite spacing) in the path of a
propagating crack (see Figure 3). Note in Figure 3A that the grain flow is parallel to the
longitudinal axis of the flange. A crack growing in the radial direction in the highly
stressed weld neck will have to cross the maximum number of austenite grains (the
light colored phases). Figure 3B, on the other hand, has the grain flow perpendicular to
DATE
JDD
03/05/13
APPROVED
DATE
REV. A01
Page 7 of 25
the axis of the flange. A radial crack can grow through the wall entirely within the ferrite
matrix (the dark
color
phase)!
DATE
JDD
03/05/13
APPROVED
DATE
REV. A01
Page 8 of 25
the stronger 25CR DSS grades. Cold working a DSS, however, will increase its HISC
susceptibility.
Maintaining the proper ferrite/austenite mix in weldments can be challenging! DSS are
easily welded and do not require a stress relief. The problem is to insure a HISC
resistant microstructure in the weldment. Heat input is of paramount importance. If the
heat input is too low, the weld metal may cool too quickly for the desired amount of
austenite to precipitate out resulting in excessive ferrite and poor HISC resistance. Too
much heat input may retard the cooling rate and result in unwanted, secondary phases.
The filler metal for DSS generally has a higher nickel content than the DSS parts being
welded. Nitrogen, a very strong austenite former, is not transferred across the weld arc
with the molten filler metal. The increased nickel in the filler metal, also a strong
austenite former, will help restore the proper ferrite/austenite balance. Dont specify
fillet welds on DSS parts! They tend to cool very rapidly resulting in excess ferrite in the
microstructure that is susceptible to HISC. Great care should be taken to insure that no
fillet welds are made on a DSS part or assembly later on in the form of attachment
welds or seal welds for fittings, sensors, anodes, etc. Always follow the design
guidelines in DNV-RP-F112 for weldment design. Some of the earliest DSS failures by
HISC were the result of fillet welds that had cooled too rapidly and ended up having 8090% ferrite.
The risk of HISC is greatest when a DSS part is under CP and is being strained at the
same time. A crack will often initiate at a ferrite/austenite grain boundary and then
propagate transgranularly through the ferrite. The austenite, although less susceptible
to HISC than the ferrite, may undergo HISC given the high plastic strains (with the
resulting cold working) and the high hydrogen charging. A DSS part may undergo
dynamic strain at stresses below its yield strength. This is known as cold creep. Cold
creep is plastic deformation at ambient temperatures that occurs under a constant load
at stresses below the yield strength of the material. HISC susceptibility increase as the
total stresses approach the yield strength of the material.
Preventing or minimizing cold creep in is one of the engineers chief responsibilities
when designing DSS parts to avoid failure by HISC. The engineer must take into
account strains resulting from applied stresses, residual stresses (resulting from
welding, fabrication, etc.), and stress concentration factors. DNV-RP-F112, Design of
Duplex Stainless Steel Subsea Equipment Exposed to Cathodic Protection, was written
to assist designers on how to perform the appropriate cold creep analysis and minimize
the risk of HISC. Most customers invoke this spec for any subsea equipment with DSS
components.
.
Cold work will increase the HISC susceptibility of DSS. Cold working generates
numerous defects in the microstructure of the material that acts as traps where
hydrogen can accumulate. DSS should always be used in the solution annealed and
DATE
JDD
03/05/13
APPROVED
DATE
REV. A01
Page 9 of 25
quenched condition whenever possible. Heavy cold work should be followed by resolution annealing and water quenching. DSS tubing that is cold bent should have a
bend radius as large as possible, but not less than 3.3D where D is the diameter of the
tubing.
DSS parts should always be coated whenever possible. This is true even for super
duplexes that are seawater corrosion resistant. The coating acts as a protective barrier
that prevents the seawater from contacting the part surface. This prevents hydrolysis
from occurring so hydrogen will not be generated. Coatings cannot be used as the only
means for preventing HISC because any holiday in the coating will permit hydrogen to
form. In general sacrificial anodes should not be directly mounted on DSS. This avoids
having to make attachment welds and avoids damaging the protective coating. It is not
practical to coat small diameter DSS tubing (control lines, injection lines, etc.), but
fortunately its not necessary. Small diameter tubing is very resistant to HISC because it
has a very fine grain size due to the high amount hot working it undergoes.
DATE
JDD
03/05/13
APPROVED
DATE
REV. A01
Page 10 of 25
for sour service. Sulfide stress cracking (SSC) is an accelerated form of hydrogen
embrittlement and alloys with good SSC resistance will have generally have good HISC
resistance. When in doubt about an alloy, laboratory testing should be performed to
verify its compatibility with cathodic protection.
DATE
JDD
03/05/13
APPROVED
DATE
REV. A01
Page 11 of 25
COATINGS
Coatings are an important weapon in our arsenal for preventing HISC in our equipment.
As previously mentioned, coatings can never be used as the sole means of preventing
HISC and certainly cannot be relied upon to protect a susceptible alloy from HISC.
Coatings get mechanically damaged. Over a period of time all organic coatings will
degrade. Sooner or later an area of the substrate will become exposed to seawater,
electrolysis will begin, and hydrogen will be charged into the metal. But coatings can
form an effective dielectric barrier over the substrate for many years and significantly
reduce the risk of HISC.
Coatings themselves are not immune to degradation from cathodic protection. Coatings
typically consist of long, interconnected chains of organic molecules. They are not
impervious to small hydrogen ions, gases, and even the highly polarized water
DATE
JDD
03/05/13
APPROVED
DATE
REV. A01
Page 12 of 25
molecule. Over a period of time these can diffuse through the coating as a result of
osmotic pressure or electrostatic attraction and collect at the coating/metal interface
particularly at areas with poor adhesion or minute defects. Gases, ions, and water
molecules that collect at the coating/metal interface may create a localized environment
that chemically and mechanically degrade the coating and reduces adhesion. As gases
and water molecules continue to accumulate in a given area, a blister may form and
grow. As the blister grows it will lift off adjacent areas of the coating from the substrate.
When the blister finally ruptures, the metal substrate will be directly exposed to
seawater and hydrogen charging will significantly increase. This loss of coating
adhesion due to the C/P system is called cathodic disbondment. If we are going to rely
on our coating to reduce the risk of HISC, it obviously must stay on the substrate. We
need our coating to have good cathodic disbondment resistance. Proper formulation is
the key here. The coating material manufacturer must use the right coating ingredients
to provide ease of application, good strength and adhesion, and good resistance to
environmental attack. Pigments need to be selected that impede the diffusion of
gasses, ions, and water. The manufacturer must be able to provide documented field
experience and the results of cathodic disbondment tests.
Correct surface preparation is essential for successful coating performance. Poor
surface preparation is the number one cause of coating failures. Areas improperly
cleaned or prepared will often result in poor coating adhesion. This in turn allows
premature coating failure as gases, water molecules, etc., then collect at these points.
Multilayer coatings are a more effective barrier than single, thick coatings: there is little
chance of a holiday (hole, pore, or localized area not coated) extending through a
multilayer coating. Coating application must adhere strictly to the coating material
manufacturers recommendations to minimize defects.
Coatings must be compatible with thermal insulation if used. In general the thicker the
coating, the more effective it is as a diffusion barrier. There is of course a limit that a
given coating system such as the standard epoxy polyamide can be applied before the
coating system itself becomes degraded. Sometimes a more robust coating system
such as multi-layer FBE/PP or chlorinated rubber may be required..
DATE
JDD
03/05/13
APPROVED
DATE
REV. A01
Page 13 of 25
DNV RP F1122
Design of Duplex Stainless Steel Subsea Equipment Exposed to
Cathodic Protection
INTRODUCTION
DATE
JDD
03/05/13
APPROVED
DATE
REV. A01
Page 14 of 25
Duplex stainless steels are used extensively for subsea equipment because of their
unique combination of strength, corrosion resistance, and the fact that they do not
require stress relieving after welding. They began to be used extensively subsea about
20 years or so ago. It wasnt long after their introduction that a number of isolated
failures occurred. Among the first to receive attention was a small pressure sensor case
made out of duplex in the North Sea. The failure occurred in a weld that had a
predominately ferrite structure rather than the desired ferrite/austenite balance. The
failure of a large duplex hub in a manifold on the Foinaven Project was another clarion
call to the industry that duplex stainless steels have some unique issues associated
with their use. The Foinaven hub failed its weld neck area, but the failure was not
associated with the weld. Extensive investigation of these and other failures attributed
the failures to hydrogen embrittlement from the cathodic protection system, known as
hydrogen induced stress cracking or HISC. Much effort was expended trying to
ascertain why some duplex parts failed and others gave trouble free service for years. It
quickly became apparent that susceptibility to HISC was a function of not only the
microstructure of the duplex part, but how the part is loaded in service. Standard design
codes then in use were not sufficient in preventing conditions that may lead to HISC in
duplex stainless steels (DSS).
DNV-RP-F112 was written as a recommended best practice that takes the lessons
learned from past duplex failures and provides guidelines for the safe use of duplex
stainless steel parts under cathodic protection (C/P). It addresses the inherent
properties of the DSS itself that influence its HISC susceptibility and how these can be
controlled through manufacturing and fabrication. It addresses loads and conditions that
must be considered when designing with DSS under C/P including stress/strain design
criteria. And finally it addresses other factors that may influence HISC susceptibility
including C/P potential, environmental factors, etc. Although this standard is called a
recommended practice, ARGUS Subsea engineers must strictly adhere to its
provisions when designing DSS components for use under C/P. All major end users
make it mandatory for their projects if DSS will be used. The review in this Training
Module is based upon the October 2008 revision of DNV-RP-F112. As more knowledge
about DSS and HISC is gained, there will undoubtedly be further revisions. Always work
to the latest!
DNV-RP-F112 is not an all-inclusive design code to be used in designing a part from
scratch. It is meant to be used in conjunction with a standard design code such ANSI
B31.3, ASME Section VIII, API 6A, etc. Where there is a conflict between DNV-RPF112 and the standard design code for a part, the more stringent requirement shall
apply.
DATE
JDD
03/05/13
APPROVED
DATE
REV. A01
Page 15 of 25
External loads that may be imparted to the DSS part from the connecting
system during installation and operation
Incidental or shock loads such as those from dropped objects, fishing gear
impact, anchorage, earthquakes, etc. If the duration of an incidental load will
persist longer than a few minutes, the effect on HISC must be considered.
Even brief incidental loads may produce permanent deformation and induce
residual stresses that remain after the load is removed or dissipated.
Installation loads while the equipment is subsea and under C/P must be
considered except if present for less than a few minutes
Design for HISC prevention shall take into account the full design life of the
equipment. Seabed subsidence over the design life must be considered.
DATE
JDD
03/05/13
APPROVED
DATE
REV. A01
Page 16 of 25
DATE
JDD
03/05/13
APPROVED
DATE
REV. A01
Page 17 of 25
DATE
JDD
03/05/13
APPROVED
DATE
REV. A01
Page 18 of 25
means residual stresses associated with a circumferential girth weld do not need to be
evaluated for a hoop direction, and a longitudinal weld need not be considered for the
axial direction. If a weldment is solution annealed after welding, residual stresses and
strains are considered negligible.
Lres may be estimated by Lres = 2.5(Rt) where R is the nominal pipe radius and t is the
wall thickness. The residual strain associated with a girth weld may be measured or
may be estimated from the following Table C1 in the RP.
Table C1 Estimates of Residual Longitudinal Strains for Girth Welds
Location
res
At weld toe
0.15%
0.25%
These values may be reduced if an FEA, which includes an initial strain distribution,
shows a beneficial effect of pressure testing, but the acceptable strain within Lres shall
not exceed that outside Lres. (Note the toe of a weld is where the weld face meets the
base material.)
The criteria for linear elastic stress is given in Section 4, Sub-section D. Residual
stresses related to welds are taken into account in the stress limits so do not need to be
included in the linear stress analysis. Misalignment between the two mating
components at a weld connection is a source of bending stress and must be analyzed.
As an alternative, misalignment can be addressed as a stress concentration factor
(SCF) at the weld. SCF can be calculated using DNV-RPC203. Stress analysis
methodology shall be in accordance with the project design code. Stress must be
linearized over the wall thickness in the principal directions into two elements:
membrane stress and bending stress.
Stress limits are given as follows (both equations must be met.):
1) m < m HISC SMYS
2) m+b < m+b HISC SMYS
DATE
JDD
03/05/13
APPROVED
DATE
REV. A01
Page 19 of 25
DATE
JDD
03/05/13
APPROVED
DATE
REV. A01
Page 20 of 25
This figure illustrates the criteria given in Table D1in the RP (given below).
DATE
JDD
03/05/13
APPROVED
DATE
REV. A01
Page 21 of 25
It is obvious from the above that the geometry of weld neck flanges is extremely
important in preventing HISC. A poorly designed weld neck hub flange was a major
contributing factor in the Foinaven failure.
As weve previously discussed, a microstructure with fine austenite spacing is more
HISC resistant than one with coarse spacing. The RP makes an allowance for this in
the stress criteria by allowing a higher HISC material factor, HISC, for fine austenite
spacing (100%) than for coarse spacing (85%) in Table D2. Austenite spacing
characterization will be addressed more fully in Section 5 of the RP, Materials, and in
Section 7, Procedure for Assessment of Austenite Spacing.
If the stress fails to meet the prescribed limits, then the RP gives the following options.
1. A more detailed assessment of the component can be made
2. The non-linear strain criteria may be invoked
3. The component may have to be redesigned to lower stresses.
Stress components parallel to a weld can be considered outside of L res, this does not
apply to the equivalent stress.
Sub-section E of Section 4 prescribes the non-linear strain criteria that may be used as
an alternative to the linear elastic stress criteria we just covered. It is based upon
limiting loads so that significant cold creep and HISC do not occur in the material. Cold
creep
facilitates the entry of hydrogen into the material and greatly increases the risk of HISC.
This approach requires that a non-linear finite element analysis (FEA) be run using a
non-linear material hardening description. Creep strain should not be included in the
non-linear strain evaluation. The calculations must consider misalignment, geometric
transitions, and welds. The FEA must have sufficient through-thickness mesh
refinement to capture the relevant strain gradients. The material hardening curve used
in the calculations shall have the SMYS and strain hardening curve adjusted for the
highest design temperature.
The material hardening curve may have the following characteristics:
DATE
JDD
03/05/13
APPROVED
DATE
REV. A01
Page 22 of 25
Kt2Se =
Where Kt = elastic stress concentrator
DATE
JDD
03/05/13
APPROVED
DATE
REV. A01
Page 23 of 25
SECTION 5 MATERIALS
This section of the RP gives guidance on the use of duplex stainless steels: limitations
and general requirements, quality control to insure sound material, and processing
suggestions that will improve their HISC resistance. The RP applies to both 22Cr
(standard) and 25Cr (super) duplex stainless steels. It references the following UNS
numbers:
22Cr S31803, S32205
25Cr S32550, S32750, S32760
There are new grades of duplex stainless steels being introduced all the time. If you
want to use a new grade that is outside of the composition limits of these UNS
numbers, the RP states that qualification (HISC resistance) testing may need to be
carried out.
The provisions of the RP are based primarily on field data from equipment that was
designed to a SMYS. It therefore limits the room temperature SMYS of 22CR DSS to
450MPa (65.3ksi) maximum and 550MPa (80.0ksi) maximum for 25Cr. Using higher
strengths may require qualification (HISC resistance) testing. The SMYS of welds is
assumed to be equal to that of the base material. The RP assumes the material is in
DATE
JDD
03/05/13
APPROVED
DATE
REV. A01
Page 24 of 25
the solution annealed condition. DSS intentionally cold worked to enhance mechanical
properties is thus outside the scope of the RP.
The material quality factor, HISC, used in the linear elastic stress evaluation approach
and given in Table D2 is dependent on whether or not the material has a fine or a
coarse austenite spacing. The RP states that the following materials can be assumed to
have a fine austenitic spacing:
All other material is assumed to have coarse spacing unless you make actual
measurements of the austenite spacing. The procedure for making these
measurements is given in Section 7 of the RP. The spacing should always be
measured in a plane in which crack propagation is most likely to occur (generally the
through-thickness direction). The RP suggests that 30 micrometers be used as the
break off point between fine and coarse spacing, but this is subject to agreement.
The orientation of the grain flow is very important as previously discussed (see Figure
3). The grain flow within the part should always be oriented parallel to the principal
tensile stress whenever possible. Crack growth will be transverse to this direction and
will have the shortest ferrite path before encountering the tougher austenite.
The RP references DNV-OS-F101, NORSOK M-601, and NORSOK M-630 for the
quality control of duplex stainless steels. It notes that there is no agreed upon, standard
test for HISC susceptibility of duplex stainless steels.
DATE
JDD
03/05/13
APPROVED
DATE
REV. A01
Page 25 of 25
Duplex stainless steels have some unique NDT issues associated with them. They are
approximately half ferritic and half austenitic. The ferrite phase is ferromagnetic while
the austenitic is not. While DSS may be sufficiently ferromagnetic to magnetize to the
proper field strength, magnetic particle inspection is impractical as a surface inspection
technique because the non-ferromagnetic austenite phase generates a tremendous
amount of background noise (like a fine spider web laid over the surface) masking
relevant indications. Dye penetrant (also called liquid penetrant inspection) is the
preferred surface NDT method. The RP recommends a dwell time of at least 60
minutes before removing the penetrant and applying the developer.
Ultrasonic testing (UT) is more complex on DSS than on carbon and low alloy steels
because of the duplex microstructure, grain size, and grain flow. The RP recommends
following the NDT recommendations in DNV-OS-F101 or EN 10228 for DSS. Both UT
and radiographic testing (RT) have their strengths and weaknesses as far as their
ability to detect certain types of defects as far as the location, size, and orientation of
the defect is concerned. Many design standards such as API 6A that require volumetric
NDT of components allows the use either method. The RP suggests that in many cases
both techniques should be used as they often complement each other. This is
particularly true of NDT of butter welds.