Designation: G 39 – 99 (Reapproved 2005)

Standard Practice for

Preparation and Use of Bent-Beam Stress-Corrosion Test
Specimens1
This standard is issued under the fixed designation G 39; the number immediately following the designation indicates the year of original
adoption or, in the case of revision, the year of last revision. A number in parentheses indicates the year of last reapproval. A superscript
epsilon (e) indicates an editorial change since the last revision or reapproval.

1. Scope 1.7 The values stated in SI units are to be regarded as

1.1 This practice covers procedures for designing, prepar- standard. The inch-pound equivalents in parentheses are pro-
ing, and using bent-beam stress-corrosion specimens. vided for information.
1.2 Different specimen configurations are given for use with 1.8 This standard does not purport to address all of the
different product forms, such as sheet or plate. This practice safety concerns, if any, associated with its use. It is the
applicable to specimens of any metal that are stressed to levels responsibility of the user of this standard to establish appro-
less than the elastic limit of the material, and therefore, the priate safety and health practices and determine the applica-
applied stress can be accurately calculated or measured (see bility of regulatory limitations prior to use. (For more specific
Note 1). Stress calculations by this practice are not applicable safety hazard information see Section 7 and 12.1.)
to plastically stressed specimens.
2. Referenced Documents
NOTE 1—It is the nature of these practices that only the applied stress 2.1 ASTM Standards: 3
can be calculated. Since stress-corrosion cracking is a function of the total
D 1141 Practice for the Preparation of Substitute Ocean
stress, for critical applications and proper interpretation of results, the
residual stress (before applying external stress) or the total elastic stress Water
(after applying external stress) should be determined by appropriate G 30 Practice for Making and Using U-Bend Stress-
nondestructive methods, such as X-ray diffraction (1).2 Corrosion Test Specimens
1.3 Test procedures are given for stress-corrosion testing by G 36 Practice for Performing Stress-Corrosion-Cracking
exposure to gaseous and liquid environments. Resistance of MEtals and Alloys in a Boiling Magnesium
1.4 The bent-beam test is best suited for flat product forms, Chloride Solution
such as sheet, strip, and plate. For plate material the bent-beam G 44 Practice for Exposure of Metals and Alloys by Alter-
specimen is more difficult to use because more rugged speci- nate Immersion in 3.5 % Sodium Chloride Solution
men holders must be built to accommodate the specimens. A G 50 Practice for Conducting Atmospheric Corrosion Tests
double-beam modification of a four-point loaded specimen to on Metals
utilize heavier materials is described in 10.5. G 85 Practice for Modified Salt Spray (Fog) Testing
1.5 The exposure of specimens in a corrosive environment 2.2 NACE Documents:4
is treated only briefly since other practices deal with this NACE TM0177-96 Laboratory Testing of Metals for Resis-
aspect, for example, Specification D 1141, and Practices G 30, tance to Specific Forms of Environmental Cracking in H2S
G 36, G 44, G 50, and G 85. The experimenter is referred to Environments
ASTM Special Technical Publication 425 (2). 3. Terminology
1.6 The bent-beam practice generally constitutes a constant
strain (deflection) test. Once cracking has initiated, the state of 3.1 Definitions of Terms Specific to This Standard:
stress at the tip of the crack as well as in uncracked areas has 3.1.1 cracking time—the time elapsed from the inception of
changed, and therefore, the known or calculated stress or strain test until the appearance of cracking.
values discussed in this practice apply only to the state of stress 3.1.1.1 Discussion—The test begins when the stress is
existing before initiation of cracks. applied and the stressed specimen is exposed to the corrosive
environment, whichever occurs later.

1
This practice is under the jurisdiction of ASTM Committee G01 on Corrosion
3
of Metals, and is the direct responsibility of Subcommittee G01.06 on Environmen- For referenced ASTM standards, visit the ASTM website, www.astm.org, or
tally Assisted Cracking. contact ASTM Customer Service at [email protected]. For Annual Book of ASTM
Current edition approved May 1, 2005. Published May 2005. Originally Standards volume information, refer to the standard’s Document Summary page on
approved in 1973. Last previous edition approved in 1999 as G 39 – 99. the ASTM website.
2 4
The boldface numbers in parentheses refer to the list of references appended to Available from National Association of Corrosion Engineers (NACE), 1440
this practice. South Creek Dr., Houston, TX 77084-4906.

Copyright © ASTM International, 100 Barr Harbor Drive, PO Box C700, West Conshohocken, PA 19428-2959, United States.

1
 G 39 – 99 (2005)
3.1.1.2 Discussion—The specimen is considered to have
failed when cracks are detected. Presence of cracks can be
determined with or without optical, mechanical, or electronic
aids. However, for meaningful interpretation, comparisons
should be made only among tests employing crack detection
methods of equivalent sensitivity.
3.1.2 stress-corrosion cracking—a cracking process requir-
ing the simultaneous action of a corrodent and sustained tensile
stress. This excludes corrosion-reduced sections that fail by
fast fracture. It also excludes intercrystalline or transcrystalline
corrosion which can disintegrate an alloy without either
applied or residual stress.

4. Summary of Practice
4.1 This practice involves the quantitative stressing of a
beam specimen by application of a bending stress. The applied
stress is determined from the size of the specimen and the
bending deflection. The stressed specimens then are exposed to
the test environment and the time required for cracks to
develop is determined. This cracking time is used as a measure
of the stress-corrosion resistance of the material in the test
environment at the stress level utilized.

5. Significance and Use

5.1 The bent-beam specimen is designed for determining
the stress-corrosion behavior of alloy sheets and plates in a
variety of environments. The bent-beam specimens are de-
signed for testing at stress levels below the elastic limit of the
alloy. For testing in the plastic range, U-bend specimens should
be employed (see Practice G 30). Although it is possible to
stress bent-beam specimens into the plastic range, the stress FIG. 1 Schematic Specimen and Holder Configurations
level cannot be calculated for plastically-stressed three- and
four-point loaded specimens as well as the double-beam
accomplished by (1) making the holder of the same material as
specimens. Therefore, the use of bent-beam specimens in the
the individual specimens, (2) inserting electrically insulating
plastic range is not recommended for general use.
materials between specimen and holder at all points of contact
6. Apparatus (see Note 4), (3) making the entire holder out of a nonmetallic
material (see Note 4), or (4) coating the holder with an
6.1 Specimen Holders—Bent-beam specimens require a electrically nonconducting coating that effectively prevents
specimen holder for each specimen, designed to retain the contact between holder and electrolyte.
applied stress on the specimen. Typical specimen holder 6.1.3 Crevice corrosion may occur in an electrolyte at
configurations are shown schematically in Fig. 1. contact points between specimen and holder (or spacer). In
NOTE 2—The double-beam specimen, more fully described in 10.5, is these instances the critical areas should be packed with a
self-contained and does not require a holder. hydrophobic filler (such as grease or wax).
NOTE 3—Specimen holders can be modified from the constant defor-
mation type shown in Fig. 1 to give a constant-load type of stressing. For NOTE 5—In atmospheres (gas) galvanic action between specimen and
instance, the loading bolt can be supplanted by a spring or dead-weight holder either does not exist or is confined to a very small area as
arrangement to change the mode of loading. experienced in outdoor exposure tests.

6.1.1 The holder shall be made of a material that would 6.2 Stressing Jigs—Three-point and four-point loaded
withstand the influence of the environment without deteriora- specimen holders, Fig. 1 ( b and c), contain a stressing feature
tion or change in shape. in the form of a loading screw. To stress two-point loaded
specimens (Fig. 1(a)), a separate stressing jig shall be used. A
NOTE 4—It should be recognized that many plastics tend to creep when convenient stressing jig is shown in Fig. 2.
subjected to sustained loads. If specimen holders or insulators are made of
such materials, the applied stress on the specimen may change appreciably NOTE 6—The double-beam specimen, described in 10.5, requires a
with time. By proper choice of holder and insulator materials, however, mechanical or hydraulic stressing frame (a universal tension testing
many plastics can be used, especially in short-time tests. machine can also be used) as well as welding equipment.
6.1.2 When the stress-corrosion test is conducted by immer- 6.3 Deflection Gages—Deflection of specimens is deter-
sion in an electrolyte, galvanic action between specimen and mined by separate gages or by gages incorporated in a loading
holder (or spacer) shall be prevented (see Note 5). This is apparatus as shown in Fig. 3. In designing a deflection gage to

2
 G 39 – 99 (2005)
anticipated loading direction in service with respect to process-
ing conditions, for example, rolling direction.
8.2 Paragraphs 7.4 and 7.5 deal specifically with specimen
selection as related to the original material surface.
9. Test Specimen
9.1 The bent-beam, stress-corrosion specimens shall be flat
strips of metal of uniform, rectangular cross section, and
uniform thickness.
9.2 The identification of individual specimens should be
permanently inscribed at each end of the specimen because this
is the area of lowest stress and cracking is not expected to be
initiated by the identification markings. If stenciling is used for
identification, this shall be done only on softened material
before any hardening heat treatments to prevent cracking in the
stenciled area. Care must be taken to prevent the identification
from being obliterated by corrosion.
9.3 Mechanical properties should be determined on the
same heat-treatment lot from which stress-corrosion specimens
are obtained.
FIG. 2 Stressing Jig and Two-Point Loaded Specimen with Holder 9.4 The specimens can be cut from sheet or plate in such a
(approximately 1⁄4 actual size) fashion that the original material surface is retained. This
procedure is recommended when it is desired to include the
effect of surface condition in the test.
9.5 If, however, it is desired that surface conditions should
not influence the test results of several materials with different
surface conditions, the surfaces of all specimens must be
prepared in the same way. It is recommended that grinding or
machining to a surface finish of at least 0.7 µm (30 µin.) and to
a depth of at least 0.25 mm (0.01 in.) be utilized for surface
preparation. It is desirable to remove the required amount of
metal in several steps by alternately grinding opposite surfaces.
This practice minimizes warpage due to residual stresses
caused by machining. All edges should be similarly ground or
machined to remove cold-worked material from previous
shearing. Chemical or electrochemical treatments that produce
hydrogen on the specimen surface must not be used on
materials that may be subject to embrittlement by hydrogen or
FIG. 3 Specimen Loading Apparatus for Three-Point Loaded that react with hydrogen to form a hydride.
Beam Specimens with Integral Deflection Gage 9.6 Immediately before stressing, the specimens should be
degreased and cleaned to remove contamination that occurred
during specimen preparation. Only chemicals appropriate for
suit individual circumstances care must be taken to reference the given metal or alloy should be used. Care must be exercised
the deflection to the proper support distance as defined in not to contaminate cleaned specimens. Also, it is suggested that
10.2-10.5. specimens be examined for cracks before exposure to the test
environment.
7. Hazards
7.1 Bent-beam specimens made from high-strength materi- 10. Stress Calculations
als may exhibit high rates of crack propagation and a specimen 10.1 The equations given in this section are valid only for
may splinter into several pieces. Due to high stresses in a stresses below the elastic limit of the material. At stresses
specimen, these pieces may leave the specimen at high velocity above the elastic limit, but below the engineering yield strength
and can be dangerous. Personnel installing and examining (0.2 % offset) only a small error results from use of the
specimens should be cognizant of this possibility and be equations (see Note 1). The equations must not be used above
protected against injury. the yield strength of the material. The following paragraphs
give relationships used to calculate the maximum longitudinal
8. Sampling stress in the outer fibers of the specimen convex surface.
8.1 Test specimens shall be selected so that they represent Calculations for transverse stress or edge-to-edge variation of
the material to be tested. In simulating a service condition, the longitudinal stress are not given; the specimen dimensions are
direction of load application in the specimen shall represent the chosen to minimize these stresses consistent with convenient

3
 G 39 – 99 (2005)
use of the specimens. The specimen dimensions given here can e 5 s/E m
be modified to suit specific needs. However, if this is done, the
approximate specimen proportions should be preserved to give 10.2.2.2 From Eq 1 determine the value of k corresponding
a similar stress distribution (for instance, if the length is to the required value of e.
doubled the width should be doubled also). 10.2.2.3 By using appropriate values of k, evaluate Eq 2 for
10.1.1 When specimens are tested at elevated temperatures, L. To facilitate calculations, a computer can be used to generate
the possibility of stress relaxation should be investigated. a table for a range of strain e and H/t with resultant values of
Relaxation can be estimated from known creep data for the (L − H)/H.
specimen, holder, and insulating materials. Differences in 10.2.3 Calculate the deflection of the specimen as follows:
thermal expansion also should be considered. y/H 5 k/~2E 2 K! (3)
10.1.2 The applied stress is determined by specimen dimen-
where:
sions and the amount of bending deflection. Thus, the errors in y = maximum deflection.
the applied stress are related to those inherent in the use of The other quantities are given in 10.2.1.
measuring instruments (micrometers, deflection gages, strain This relationship can be used as a simple check to ensure that
gages, and so forth). For the two-point loaded specimens, most the maximum stress does not exceed the proportional limit. If
measured values lie within 5 % of the values calculated in it should exceed the proportional limit, the measured deflection
accordance with the procedures given in 10.2.1-10.2.3, as will be greater than that calculated from Eq 3.
reported by Haaijer and Loginow (4). The calculated stress 10.2.4 As an alternative method the following approximate
applies only to the state of stress before initiation of cracks. relationship can be used for calculating specimen length:
Once cracking is initiated, the stress at the tip of the crack, as
well as in uncracked areas, has changed. L 5 ~ktE/s! sin21 ~Hs/ktE! (4)

10.2 Two-Point Loaded Specimens—This specimen can be where:

used for materials that do not deform plastically when bent to L = specimen length,
(L − H)/H = 0.01 (see section 10.2.5). The specimens shall be s = maximum stress,
approximately 25 by 254-mm (1- by 10-in.) flat strips cut to E = modulus of elasticity,
appropriate lengths to produce the desired stress after bending H = holder span,
as shown in Fig. 1(a). t = thickness of specimen, and
k = 1.280, an empirical constant.
10.2.1 Calculate the elastic stress in the outer fiber at
This equation can be solved by computer, by trial and error,
midlength of the two-point loaded specimens from relation-
or by using a series expansion of the sine function. Eq 4 shall
ships derived from a theoretically exact large-deflection analy-
be used only when the quantity (Hs/ktE) is less than 1.0.
sis (4), as follows: 10.2.5 Choose specimen thickness and length and holder
F
k 2E 2 K t
e 5 4~2E 2 K! 2 2 12 H S DG t
H (1)
span to obtain a value for (L − H)/H of between 0.01 and 0.50,
thus keeping the error of stress within acceptable limits. A
and specimen thickness of about 0.8 to 1.8 mm (0.03 to 0.07 in.)
and a holder span of 177.8 to 215.9 mm (7.00 to 8.50 in.) has
~L 2 H!/H 5 [K/2E 2 K!# 2 1 (2)
been very convenient when working with very high strength
where: steels and aluminum alloys with applied stresses ranging from
L = length of specimen, about 205 MPa (30 ksi) for aluminum to 1380 MPa (200 ksi)
H = distance between supports (holder span), for steel. The specimen dimensions given here can be modified
t = thickness of specimen, to suit specific needs. However, if this is done, approximate
e = maximum tensile strain, dimensional proportions shall be preserved.
K = *0p/2 (1 − k2 sin2z) −1/2 dz (complete elliptic integral of 10.2.6 In two-point loaded specimens the maximum stress
the first kind), occurs at midlength of the specimen and decreases to zero at
E = *0p/2 (1 − k2 sin2z)1/2 dz (complete elliptic integral of the specimen ends.
second kind), 10.2.7 The two-point loaded specimen is preferred to three-
k = sin u/2, point loaded specimens because in many instances crevice
u = maximum slope of the specimen, that is, at the end of corrosion of the specimen occurs at the central support of the
the specimen, and three-point loaded specimen. Since this corrosion site is very
z = integration parameter (4). close to the point of highest tension stress, it may cathodically
10.2.2 The mathematical analysis establishes that Eq 1 and protect the specimen and prevent possible crack formation or
Eq 2 define the relationship between the strain e and (L − H)/H cause hydrogen embrittlement. Furthermore, the pressure of
in parameter form. The common parameter in these equations the central support at the point of highest load introduces
is the modulus k of the elliptic integrals. Thus, the following biaxial stresses at the area of contact and could introduce
procedure can be used to determine the specimen length L that tension stresses where normally compression stresses are
is required to produce a given maximum stress s: present.
10.2.2.1 Divide the stress s by the modulus of elasticity Em NOTE 7—Occasionally two-point loaded specimens having a nonuni-
to determine the strain e. form cross section are used for special purposes. A description of such a

4
 G 39 – 99 (2005)
specimen is given by Wilson and Spier (5). H = distance between outer supports, and
10.3 Three-Point Loaded Specimens—The specimen shall A = distance between inner and outer supports.
be a flat strip typically 25 to 51-mm (1 to 2-in.) wide and 127 The dimensions are often chosen so that A = H/4.
to 254-mm (5 to 10-in.) long. The thickness of the specimen is 10.4.2 An alternative method of calculating the elastic stress
usually dictated by the mechanical properties of the material between the inner supports is as follows:
and the product form available. Support the specimen at the s 5 4Ety8/h 2 (7)
ends and bend the specimen by forcing a screw (equipped with
a ball or knife-edge tip) against it at the point halfway between
the end supports in a fashion shown in Fig. 1(b). The specimen where:
dimensions given here can be modified to suit specific needs. h = distance between inner supports, and
However, if this is done, approximate dimensional proportions y8 = deflection between inner supports.
shall be preserved. (This equation is a special case of 10.4.1 when A = 0.)
10.4.3 The above relationships are based on small deflec-
10.3.1 Calculate the elastic stress at midspan in the outer
tions (y/H less than 0.1). In sheet-gage bent-beam specimens
fibers of three-point loaded specimens from the relationship:
the deflections are usually large, and thus, the relationships are
s 5 6Ety/H 2 (5) only approximate. To obtain more accurate stress values, use
for calibration a prototype specimen equipped with strain
where: gages. This prototype specimen should have the same dimen-
s = maximum tensile stress, sions as the test specimens and should be stressed in the same
E = modulus of elasticity, way.
t = thickness of specimen, 10.4.4 In four-point loaded specimens the maximum stress
y = maximum deflection, and occurs between the contact points with the inner supports; in
H = distance between outer supports. this area the stress is constant. From the inner supports the
10.3.2 The above relationship is based on small deflections stress decreases linearly toward zero at the outer supports.
(y/H less than 0.1). In sheet-gage, bent-beam specimens the 10.5 Double-Beam Specimen—The specimen shall consist
deflections are usually large, and thus, the relationship is only of two flat strips 25 to 51-mm (1 to 2-in.) wide and 127 to
approximate. To obtain more accurate stress values, use a 254-mm (5 to 10-in.) long. Bend the strips against each other
prototype specimen, equipped with strain gages, for calibra- over a centrally located spacer until both ends of the specimens
tion. This prototype should have the same dimensions as the touch. Hold them in this position by welding the ends together
test specimens and should be stressed in the same way. as shown in Fig. 1(d) (see Note 8). An equivalent procedure for
10.3.3 In three-point loaded specimens the maximum stress bolted specimens is described on pp. 319–321 of Ref (2).
occurs at midlength of the specimen and decreases linearly to NOTE 8—If the test is to be conducted in an electrolyte, the spacer shall
zero at the outer supports. be made of the same material as the specimen (or of an electrically
10.3.4 For limitation in the use of three-point loaded speci- nonconducting material such as glass, ceramic, and so forth) to prevent
mens see 10.2.7. galvanic action between specimen and spacer. See also 6.1.2 and Note 4
and Note 5.
10.4 Four-Point Loaded Specimens—The specimen shall be
a flat strip typically 25 to 51-mm (1 to 2-in.) wide and 127 to 10.5.1 Calculate the elastic stress for the midportion of the
254-mm (5 to 10-in.) long. The thickness of the specimen is specimen (between contact points of the spacer) in the outer
usually dictated by the mechanical properties of the material fibers of the doublebeam specimens from the following rela-
and the product form available. Support the specimen at the tionship:
ends and bend the specimen by forcing two inner supports 3Ets
against it in a fashion shown in Fig. 1(c). The two inner s5 (8)
H 2@1 2 ~h/H!#@1 1 ~2h/H!#
supports shall be located symmetrically around the midpoint
between the outer supports. The specimen dimensions given where:
here can be modified to suit specific needs. However, if this is s = maximum tensile stress,
done, approximate dimensional proportions shall be preserved. E = modulus of elasticity,
10.4.1 Calculate the elastic stress for the midportion of the t = thickness of specimen strip,
specimen (between contact points of the inner support) in the s = thickness of spacer,
H = see Fig. 1(d), and
outer fibers of four-point loaded specimens from the following
h = length of spacer.
relationship:
10.5.2 When the length of the spacer h is chosen so that
s 5 12Ety/~3H 2 2 4A 2! (6) H = 2h the equation in 10.5.1 is simplified to:
s 5 3Ets/H2
where:
s = maximum tensile stress, 10.5.3 The above relationships are based on small deflec-
E = modulus of elasticity, tions (s/H being less than 0.2). In sheet-gage bent-beam
t = thickness of specimen, specimens the deflections are usually large, and thus, the
y = maximum deflection (between outer supports), relationships are only approximate. To obtain more accurate

5
 G 39 – 99 (2005)
stress values, use a prototype specimen, equipped with strain severe enough to produce varying degrees of cracking in the
gages, for calibration. The prototype specimen should have the alloys of interest, in mechanical or thermal treatments used, or
same dimensions as the test specimens and should be stressed in sensitivity to specific environments investigated. By testing
in the same way. a set of specimens at a series of stress levels, the stress
10.5.4 In double-beam specimens the maximum stress oc- dependence of alloys can be assessed. The bent-beam speci-
curs between the contact points with the spacer; in this area the men is very well suited for establishing the relative merits of
stress is constant. From the contact with the spacer the stress several alloys for the relative severity of several environments.
decreases linearly toward zero at the ends of specimens. 11.3 Ideally, the environmental test conditions should be the
same that would prevail in the intended use of the alloys. In
11. Choice of Test Conditions choosing a set of test conditions, it is important that they
11.1 The purpose of stress-corrosion testing is to simulate (environment and stress) be well defined and reproducible. A
on a small scale the conditions (materials, stress, and environ- detailed discussion is given by Loginow (6).
ment) that exist in an engineering application. The stresses in 11.4 The presence of a machined notch in the middle of the
an engineering structure can be varied between operational tension side of a bent beam will induce a severe triaxial stress
(design) stresses and residual stresses (from heat treatment or state at the root of the notch. The actual bending stress there
fabrication). Residual stresses are frequently the more impor- will be greater by a concentration factor dependent on the
tant, primarily because current design practices and close notch geometry, than the minimal test stress, and generally,
control of processes have kept operational stresses well below may be expected to be in the range of plastic stain. Advantages
the yield strength of the metal in use. On the other hand, of such a notched specimen include the probable localized
magnitude and direction of residual stresses frequently are cracking in the notch and an acceleration of failure. However,
difficult to predict and also difficult to measure. Depending on unless directly related to practical conditions of usage, mis-
the degree of restraint, residual stresses may even exceed the leading failures may ensue.
initial yield strength of the material.
11.4.1 Another type of stress concentration at the site of two
11.2 Generally stress-corrosion testing falls into two broad
drilled holes located half way between the end supports of a
categories: (1) evaluation of materials for a specific applica-
three-point loaded bent beam has been used in the evaluation of
tion, and (2) comparison of the relative behavior of several
metals for oilfield equipment. Details on the preparation and
materials or environments.
use of this specimen are described in NACE TM0177-96.
11.2.1 To evaluate materials for specific applications, the
Laboratory test data for carbon and low-alloy steels have been
testing conditions should be representative of the most severe
found to correlate with field data (7).
conditions to which the materials would be subjected in
service. Testing at nominal or design conditions could be
misleading. An engineering structure, because of residual 12. Specimen Exposure
stresses, is expected to be stressed to its yield strength at some 12.1 Expose the stressed specimens to the environment
points even if the design stress for that structure is appreciably (gaseous or liquid) of interest. This can be accomplished by
below yield strength. Thus, the use of the elastically stressed mounting the specimen holders on appropriate racks and
bent-beam specimens for materials evaluation is of limited exposing the entire rack to the environment. A typical atmo-
value. spheric exposure rack is shown in Fig. 4. As noted in 7.1,
11.2.2 To compare materials or environments for relative bent-beam specimens may break violently and thus cause
stress-corrosion behavior, the test conditions may be only injury. To protect personnel and to prevent specimen loss, drill

FIG. 4 Bent-Beam Specimens on Atmospheric Exposure Rack

6
 G 39 – 99 (2005)
holes in specimen ends and holders and secure the specimens specimen contains only one or a few cracks, the shape of the
by wires to their holders. bend can be considerably changed, predominantly by kinking;
12.2 Determination of cracking time is a subjective proce- this feature helps in identifying cracked specimens. However,
dure involving visual examination that under some conditions if many cracks are present, a change in shape may not be
can be very difficult, as noted in Section 13, and depends on the apparent. It should also be noted that presence of voluminous
skill and experience of the inspector. corrosion products may obscure cracks, thus making a careful
12.3 Laboratory Exposure of Bent Beams—In both alternate examination mandatory. In these instances metallographic
and sustained immersion of bent beams, avoid galvanic corro- sectioning of the specimen may be necessary to detect cracks.
sion between fixtures and specimens as discussed in 6.1.2 and
Note 4 and Note 5. It should be recognized that, at points of 14. Report
contact between specimen and fixture, crevice corrosion may 14.1 Results of stress-corrosion tests with bent-beam speci-
occur on some materials, which in turn may result in galvanic mens are expressed as the time to produce failure by cracking
protection of the stressed area. If this condition occurs, either or as the fraction of specimens that have cracked in a fixed
eliminate the crevice or consider a different kind of specimen. time. In addition to the cracking time the following data shall
In alternate immersion, expose the specimen to allow complete be reported:
drainage and drying of the surface. In immersion tests, arrange 14.1.1 Specimen identification,
the specimens so as to prevent contact with each other. In both 14.1.2 Material name or specification code,
sustained and alternate immersion, the solution volume should 14.1.3 Chemical composition,
be large enough to prevent depletion of corrosive agents. In 14.1.4 Heat treatment,
elevated-temperature tests, make arrangements to reflux the 14.1.5 Mechanical properties,
solution to maintain a constant concentration. 14.1.6 Type and orientation of specimen used and surface
12.4 Atmospheric Exposure of Bent Beams—Expose the condition (hot rolled, cold rolled, machined, surface ground,
specimens in an area that is representative of the atmospheric and so forth),
conditions of interest. 14.1.7 Applied stress (and residual stress, if known),
14.1.8 Details of specimen preparation if different from
13. Inspection of Specimens those specified here (or if not specified),
13.1 As continuous observation of specimens is usually 14.1.9 Detailed description of test environment, and
impractical, inspect specimens for appearance of cracks at 14.1.10 Remarks concerning the size and appearance of
predetermined time intervals. These intervals are usually in- cracks may be included.
creased as the test progresses because the logarithms of
observed cracking times are often normally distributed as 15. Keywords
described by Loginow (6) and by Booth et al (8). 15.1 bent-beam; constant deformation; constant load; elastic
13.2 Determine presence of cracks by visual observation, strain; quantitative stress; stress-corrosion cracking; stress-
usually with the aid of a 5 to 10 power magnifying glass. If the corrosion test specimen

REFERENCES

(1) Measurement for Stress by X-Ray, SAE Information Report TR-182, (6) Loginow, A. W., “Stress-Corrosion Testing of Alloys,” Materials
Society of Automotive Engineers, New York, NY. Protection, Vol 5, No. 5, May 1966, pp. 33–39.
(2) Symposium on Stress-Corrosion Testing, ASTM STP 425, ASTM, (7) Fraser, J. P., Eldredge, G. G., and Treseder, R. S., “Laboratory and
1967. Field Methods for Quantitative Study of Sulfide Corrosion Cracking,”
(3) Snape, E., “Roles of Composition and Microstructure in Sulfide H25 Corrosion in Oil and Gas Production—A Compilation of Classic
Cracking of Steel,” Corrosion, Vol 24, 1968, pp. 261–282.
Papers, Tuttle, R. N. and Kane, R. D., eds. (Houston, TX: NACE,
(4) Haaijer, G., and Loginow, A. W., “Stress Analysis of Bent-Beam
1981) p. 283. Original publication: Corrosion 14, 11 (1958) p. 517t.
Stress-Corrosion Specimen,” Corrosion, Vol 21, 1965, pp. 105–112.
(5) Wilson, P. E., and Spier, E. E., “Nonlinear Bending of a Stress- (8) Booth, F. F., Tucker, G. E. G., and Goddard, H. P., “Statistical
Corrosion Specimen,” Journal of Engineering for Industry, Vol 88, Distribution of Stress-Corrosion Endurance,” Corrosion, Vol 19, 1963,
1966, pp. 31–36. pp. 390t–395t.

7
 G 39 – 99 (2005)

ASTM International takes no position respecting the validity of any patent rights asserted in connection with any item mentioned
in this standard. Users of this standard are expressly advised that determination of the validity of any such patent rights, and the risk
of infringement of such rights, are entirely their own responsibility.

This standard is subject to revision at any time by the responsible technical committee and must be reviewed every five years and
if not revised, either reapproved or withdrawn. Your comments are invited either for revision of this standard or for additional standards
and should be addressed to ASTM International Headquarters. Your comments will receive careful consideration at a meeting of the
responsible technical committee, which you may attend. If you feel that your comments have not received a fair hearing you should
make your views known to the ASTM Committee on Standards, at the address shown below.

This standard is copyrighted by ASTM International, 100 Barr Harbor Drive, PO Box C700, West Conshohocken, PA 19428-2959,
United States. Individual reprints (single or multiple copies) of this standard may be obtained by contacting ASTM at the above
address or at 610-832-9585 (phone), 610-832-9555 (fax), or [email protected] (e-mail); or through the ASTM website
(www.astm.org).