This international standard was developed in accordance with internationally recognized principles on standardization established in the Decision on Principles

for the
Development of International Standards, Guides and Recommendations issued by the World Trade Organization Technical Barriers to Trade (TBT) Committee.

Designation: E1820 − 23

Standard Test Method for

Measurement of Fracture Toughness1
This standard is issued under the fixed designation E1820; 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 (´) indicates an editorial change since the last revision or reapproval.

1. Scope* mendations issued by the World Trade Organization Technical

1.1 This test method covers procedures and guidelines for Barriers to Trade (TBT) Committee.
the determination of fracture toughness of metallic materials 2. Referenced Documents
using the following parameters: K, J, and CTOD (δ). Tough-
ness can be measured in the R-curve format or as a point value. 2.1 ASTM Standards:2
The fracture toughness determined in accordance with this test E4 Practices for Force Calibration and Verification of Test-
method is for the opening mode (Mode I) of loading. ing Machines
E8/E8M Test Methods for Tension Testing of Metallic Ma-
NOTE 1—Until this version, KIc could be evaluated using this test terials
method as well as by using Test Method E399. To avoid duplication, the
evaluation of KIc has been removed from this test method and the user is E21 Test Methods for Elevated Temperature Tension Tests of
referred to Test Method E399. Metallic Materials
1.2 The recommended specimens are single-edge bend, E23 Test Methods for Notched Bar Impact Testing of Me-
[SE(B)], compact, [C(T)], and disk-shaped compact, [DC(T)]. tallic Materials
All specimens contain notches that are sharpened with fatigue E399 Test Method for Linear-Elastic Plane-Strain Fracture
cracks. Toughness of Metallic Materials
1.2.1 Specimen dimensional (size) requirements vary ac- E1290 Test Method for Crack-Tip Opening Displacement
cording to the fracture toughness analysis applied. The guide- (CTOD) Fracture Toughness Measurement (Withdrawn
lines are established through consideration of material 2013)3
toughness, material flow strength, and the individual qualifi- E1823 Terminology Relating to Fatigue and Fracture Testing
cation requirements of the toughness value per values sought. E1921 Test Method for Determination of Reference
Temperature, T0, for Ferritic Steels in the Transition
NOTE 2—Other standard methods for the determination of fracture Range
toughness using the parameters K, J, and CTOD are contained in Test E1942 Guide for Evaluating Data Acquisition Systems Used
Methods E399, E1290, and E1921. This test method was developed to
provide a common method for determining all applicable toughness in Cyclic Fatigue and Fracture Mechanics Testing
parameters from a single test. E2298 Test Method for Instrumented Impact Testing of
1.3 The values stated in SI units are to be regarded as Metallic Materials
standard. The values given in parentheses after SI units are 2.2 ASTM Data Sets:4
provided for information only and are not considered standard. E1820/1–DS1(2016) Standard data set 1 to evaluate com-
puter algorithms for evaluation of JIc using Annex 9 of
1.4 This standard does not purport to address all of the E1820
safety concerns, if any, associated with its use. It is the E1820/2–DS2(2020) Standard data set 2 to evaluate com-
responsibility of the user of this standard to establish appro- puter algorithms for evaluation of JIc using Annex 9 of
priate safety, health, and environmental practices and deter- E1820
mine the applicability of regulatory limitations prior to use. E1820/3–DS3(2020) Standard data set 3 to evaluate com-
1.5 This international standard was developed in accor- puter algorithms for evaluation of JIc using Annex 9 of
dance with internationally recognized principles on standard- E1820
ization established in the Decision on Principles for the
Development of International Standards, Guides and Recom-
2
For referenced ASTM standards, visit the ASTM website, www.astm.org, or
contact ASTM Customer Service at [email protected]. For Annual Book of ASTM
1
This test method is under the jurisdiction of ASTM Committee E08 on Fatigue Standards volume information, refer to the standard’s Document Summary page on
and Fracture and is the direct responsibility of Subcommittee E08.07 on Fracture the ASTM website.
3
Mechanics. The last approved version of this historical standard is referenced on www.ast-
Current edition approved Feb. 1, 2023. Published March 2023. Originally m.org.
approved in 1996. Last previous edition approved in 2022 as E1820 – 22ɛ1. DOI: 4
These data sets are available for download from ASTM at
10.1520/E1820-23 https://www.astm.org/get-involved/technical-committees/adhoc-e08.html

*A Summary of Changes section appears at the end of this standard

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

1
 E1820 − 23
E1820/4–DS4(2020) Standard data set 4 to evaluate com- available at the front of an ideal crack in an elastic solid during
puter algorithms for evaluation of JIc using Annex 9 of a virtual increment of forward crack extension.
E1820 3.2.5 crack-tip opening displacement (CTOD), δ [L],
E1820/5–DS5(2020) Standard data set 5 to evaluate com- n—crack displacement resulting from the total deformation
puter algorithms for evaluation of JIc using Annex 9 of (elastic plus plastic) at variously defined locations near the
E1820 original (prior to force application) crack tip.
E1820/6–DS6(2020) Standard data set 6 to evaluate com-
puter algorithms for evaluation of JIc using Annex 9 of 3.2.5.1 Discussion—In this test method, CTOD is the dis-
E1820 placement of the crack surfaces normal to the original (un-
E1820/7–DS7(2020) Standard data set 7 to evaluate com- loaded) crack plane at the tip of the fatigue precrack, ao. In this
puter algorithms for evaluation of JIc using Annex 9 of test method, CTOD is calculated at the original crack size, ao,
E1820 from measurements made from the force versus displacement
E1820/8–DS8(2020) Standard data set 8 to evaluate com- record.
puter algorithms for evaluation of JIc using Annex 9 of 3.2.5.2 Discussion—In CTOD testing, δIc [L] is a value of
E1820 CTOD near the onset of slow stable crack extension, here
E1820/9–DS9(2020) Standard data set 9 to evaluate com- defined as occurring at ∆ap = 0.2 mm (0.008 in.) + 0.7δIc.
puter algorithms for evaluation of JIc using Annex 9 of 3.2.5.3 Discussion—In CTOD testing, δc [L] is the value of
E1820 CTOD at the onset of unstable crack extension (see 3.2.36) or
pop-in (see 3.2.22) when ∆ap < 0.2 mm (0.008 in.) + 0.7δc. δc
3. Terminology corresponds to the force Pc and clip-gage displacement vc (see
3.1 Terminology E1823 is applicable to this test method. Fig. 1). It may be size-dependent and a function of test
Only items that are exclusive to Test Method E1820, or that specimen geometry.
have specific discussion items associated, are listed in this 3.2.5.4 Discussion—In CTOD testing, δu [L] is the value of
section. CTOD at the onset of unstable crack extension (see 3.2.36) or
3.2 Definitions of Terms Specific to This Standard: pop-in (see 3.2.22) when the event is preceded by ∆ap >0.2 mm
3.2.1 compliance [LF−1], n—the ratio of displacement in- (0.008 in.) + 0.7δu. The δu corresponds to the force Pu and the
crement to force increment. clip gage displacement vu (see Fig. 1). It may be size-
dependent and a function of test specimen geometry. It can be
3.2.2 crack opening displacement (COD) [L], n—force- useful to define limits on ductile fracture behavior.
induced separation vector between two points at a specific gage
3.2.5.5 Discussion—In CTOD testing, δc* [L] characterizes
length. The direction of the vector is normal to the crack plane.
the CTOD fracture toughness of materials at fracture instability
3.2.2.1 Discussion—In this practice, displacement, v, is the
prior to the onset of significant stable tearing crack extension.
total displacement measured by clip gages or other devices
The value of δc* determined by this test method represents a
spanning the crack faces.
measure of fracture toughness at instability without significant
3.2.3 crack extension, ∆a [L], n—an increase in crack size. stable crack extension that is independent of in-plane dimen-
3.2.4 crack-extension force, G [FL−1 or FLL−2], n—the sions. However, there may be a dependence of toughness on
elastic energy per unit of new separation area that is made thickness (length of crack front).

NOTE 1—Construction lines drawn parallel to the elastic loading slope to give vp, the plastic component of total displacement, vg.
NOTE 2—In curves b and d, the behavior after pop-in is a function of machine/specimen compliance, instrument response, and so forth.
FIG. 1 Types of Force versus Clip gage Displacement Records

2
 E1820 − 23
3.2.6 dial energy, KV [FL]—absorbed energy as indicated J-integral is equal to the value obtained from two identical
by the impact machine encoder or dial indicator, as applicable. bodies with infinitesimally differing crack areas each subject to
3.2.7 dynamic stress intensity factor, KJd—The dynamic stress. The parameter J is the difference in work per unit
equivalent of the stress intensity factor KJ, calculated from J difference in crack area at a fixed value of displacement or,
using the equation specified in this test method. where appropriate, at a fixed value of force (1)5.
3.2.11.5 Discussion—The dynamic equivalent of Jc is
3.2.8 effective thickness, Be [L] , n—for side-grooved speci-
Jcd,X, with X = order of magnitude of J-integral rate.
mens Be = B − (B − BN)2/B. This is used for the elastic
unloading compliance measurement of crack size. 3.2.12 Jc [FL−1] —The property Jc determined by this test
method characterizes the fracture toughness of materials at
3.2.9 effective yield strength, σY [FL−2], n—an assumed
fracture instability prior to the onset of significant stable
value of uniaxial yield strength that represents the influence of
tearing crack extension. The value of Jc determined by this test
plastic yielding upon fracture test parameters.
method represents a measure of fracture toughness at instabil-
3.2.9.1 Discussion—It is calculated as the average of the
ity without significant stable crack extension that is indepen-
0.2 % offset yield strength σYS, and the ultimate tensile
dent of in-plane dimensions; however, there may be a depen-
strength, σTS as follows:
dence of toughness on thickness (length of crack front).
σ YS1σ TS
σY 5 (1) 3.2.13 Ju [FL−1]—The quantity Ju determined by this test
2
method measures fracture instability after the onset of signifi-
3.2.9.2 Discussion—In estimating σY, influences of testing cant stable tearing crack extension. It may be size-dependent
conditions, such as loading rate and temperature, should be and a function of test specimen geometry. It can be useful to
considered. define limits on ductile fracture behavior.
3.2.9.3 Discussion—The dynamic effective yield strength, 3.2.13.1 Discussion—The dynamic equivalent of Ju is Jud,X,
σYd, is the dynamic equivalent of the effective yield strength. with X = order of magnitude of J-integral rate.
3.2.10 general yield force, Pgy [F]—in an instrumented
3.2.14 J-integral rate, J̇ @ FL21 T 21 # —derivative of J with
impact test, applied force corresponding to general yielding of
respect to time.
the specimen ligament. It corresponds to Fgy, as used in Test
Method E2298. 3.2.15 machine capacity, MC [FL]—maximum available
energy of the impact testing machine.
3.2.11 J-integral, J [FL−1], n—a mathematical expression, a
line or surface integral that encloses the crack front from one 3.2.16 maximum force, Pmax [F]—in an instrumented im-
crack surface to the other, used to characterize the local pact test, maximum value of applied force. It corresponds to
stress-strain field around the crack front. Fm, as used in Test Method E2298.
3.2.11.1 Discussion—The J-integral expression for a two- 3.2.17 net thickness, BN [L], n—distance between the roots
dimensional crack, in the x-z plane with the crack front parallel of the side grooves in side-grooved specimens.
to the z-axis, is the line integral as follows: 3.2.18 original crack size, ao [L] , n—the physical crack size
* S Wdy 2 T̄· ]] ūx dsD
at the start of testing.
J5 (2) 3.2.18.1 Discussion—In this test method, aoq is used to
Γ

denote original crack size estimated from compliance.

where:
3.2.19 original remaining ligament, bo [L], n—distance
W = loading work per unit volume or, for elastic bodies,
from the original crack front to the back edge of the specimen,
strain energy density,
Γ = path of the integral, that encloses (that is, contains) that is (bo = W − ao).
the crack tip, 3.2.20 physical crack size, ap[L] , n—the distance from a
ds = increment of the contour path, reference plane to the observed crack front. This distance may
T̄ = outward traction vector on ds, represent an average of several measurements along the crack
ū = displacement vector at ds, front. The reference plane depends on the specimen form, and
x, y, z = rectangular coordinates, and it is normally taken to be either the boundary, or a plane
]ū = rate of work input from the stress field into the area containing either the load-line or the centerline of a specimen
T̄· ds
]x enclosed by Γ. or plate. The reference plane is defined prior to specimen
3.2.11.2 Discussion—The value of J obtained from this deformation.
equation is taken to be path-independent in test specimens 3.2.21 plane-strain fracture toughness, JIc [FL−1], KJIc
commonly used, but in service components (and perhaps in test [FL−3/2] , n—the crack-extension resistance under conditions
specimens) caution is needed to adequately consider loading of crack-tip plane-strain.
interior to Γ such as from rapid motion of the crack or the 3.2.21.1 Discussion—For example, in Mode I for slow rates
service component, and from residual or thermal stress. of loading and substantial plastic deformation, plane-strain
3.2.11.3 Discussion—In elastic (linear or nonlinear) solids, fracture toughness is the value of the J-integral designated JIc
the J-integral equals the crack-extension force, G. (See crack [FL−1] as measured using the operational procedure (and
extension force.)
3.2.11.4 Discussion—In elastic (linear and nonlinear) solids 5
The boldface numbers in parentheses refer to the list of references at the end of
for which the mathematical expression is path independent, the this standard.

3
 E1820 − 23
satisfying all of the qualification requirements) specified in this lim
K 3 5 r→0 @ τ yz~ 2πr ! 1/2 # (5)
test method, that provides for the measurement of crack-
where r = distance directly forward from the crack tip
extension resistance near the onset of stable crack extension.
to a location where the significant stress is calculated.
3.2.21.2 Discussion—For example, in Mode I for slow rates 3.2.32.2 Discussion—In this test method, Mode 1 or Mode
of loading, plane-strain fracture toughness is the value of the I is assumed. See Terminology E1823 for definition of mode.
stress intensity designated KJIc calculated from JIc using the
equation (and satisfying all of the qualification requirements) 3.2.33 stress-intensity factor rate, K̇ [FL-3/2T-1]—derivative
specified in this test method, that provides for the measurement of K with respect to time.
of crack-extension resistance near the onset of stable crack 3.2.34 stretch-zone width, SZW [L], n—the length of crack
extension under dominant elastic conditions (2). extension that occurs during crack-tip blunting, for example,
3.2.21.3 Discussion—The dynamic equivalent of JIc is JIcd,X prior to the onset of unstable brittle crack extension, pop-in, or
, with X = order of magnitude of J-integral rate. slow stable crack extension. The SZW is in the same plane as
the original (unloaded) fatigue precrack and refers to an
3.2.22 pop-in, n—a discontinuity in the force versus clip
extension beyond the original crack size.
gage displacement record. The record of a pop-in shows a
sudden increase in displacement and, generally a decrease in 3.2.35 time to fracture, tf [T]—time corresponding to speci-
force. Subsequently, the displacement and force increase to men fracture.
above their respective values at pop-in. 3.2.36 unstable crack extension [L], n—an abrupt crack
3.2.23 R-curve or J-R curve, n—a plot of crack extension extension that occurs with or without prior stable crack
resistance as a function of stable crack extension, ∆ap or ∆ae. extension in a standard test specimen under crosshead or clip
3.2.23.1 Discussion—In this test method, the J-R curve is a gage displacement control.
plot of the far-field J-integral versus the physical crack 3.3 Symbols:
extension, ∆ap. It is recognized that the far-field value of J may 3.3.1 ti [T]—time corresponding to the onset of crack
not represent the stress-strain field local to a growing crack. propagation.
3.2.24 remaining ligament, b [L], n—distance from the 3.3.2 v0 [LT-1]—in an instrumented impact test, striker
physical crack front to the back edge of the specimen, that is velocity at impact.
(b = W − ap). 3.3.3 Wm [FL]—in an instrumented impact test, absorbed
3.2.25 specimen center of pin hole distance, H* [L], n—the energy at maximum force.
distance between the center of the pin holes on a pin-loaded 3.3.4 Wt [FL]—in an instrumented impact test, total ab-
specimen. sorbed energy calculated from the complete force/displacement
3.2.26 specimen gage length, d [L], n—the distance be- test record.
tween the points of displacement measure (for example, clip 3.3.5 W0 [FL]—in an instrumented impact test, available
gage, gage length). impact energy.
3.2.27 specimen span, S [L], n—the distance between speci-
4. Summary of Test Method
men supports.
4.1 The objective of this test method is to load a fatigue
3.2.28 specimen thickness, B [L], n—the side-to-side di-
precracked test specimen to induce either or both of the
mension of the specimen being tested.
following responses (1) unstable crack extension, including
3.2.29 specimen width, W [L], n—a physical dimension on significant pop-in, referred to as “fracture instability” in this
a test specimen measured from a reference position such as the test method; (2) stable crack extension, referred to as “stable
front edge in a bend specimen or the load-line in the compact tearing” in this test method. Fracture instability results in a
specimen to the back edge of the specimen. single point-value of fracture toughness determined at the point
3.2.30 stable crack extension [L], n—a displacement- of instability. Stable tearing results in a continuous fracture
controlled crack extension beyond the stretch-zone width (see toughness versus crack-extension relationship (R-curve) from
3.2.34). The extension stops when the applied displacement is which significant point-values may be determined. Stable
held constant. tearing interrupted by fracture instability results in an R-curve
up to the point of instability.
3.2.31 strain rate, ε̇—derivative of strain ε with respect to
time. 4.2 This test method requires continuous measurement of
force versus load-line displacement or crack mouth opening
3.2.32 stress-intensity factor, K, K1, K2, K3, KI, KII, KIII
displacement, or both. If any stable tearing response occurs,
[FL−3/2], n—the magnitude of the ideal-crack-tip stress field
then an R-curve is developed and the amount of slow-stable
(stress-field singularity) for a particular mode in a
crack extension shall be measured.
homogeneous, linear-elastic body.
3.2.32.1 Discussion—Values of K for the Modes 1, 2, and 3 4.3 Two alternative procedures for measuring crack exten-
are given by the following equations: sion are presented, the basic procedure and the resistance curve
procedure. The basic procedure involves physical marking of
lim
K 1 5 r→0 @ σ yy~ 2πr ! 1/2 # (3)
the crack advance and multiple specimens used to develop a
K2 5 lim
r→0 @ τ xy~ 2πr ! #
1/2
(4) plot from which a single point initiation toughness value can be

4
 E1820 − 23
evaluated. The resistance curve procedure is an elastic- 5.2.3 The values of δc, δu, Jc, and Ju may be affected by
compliance method where multiple points are determined from specimen dimensions.
a single specimen. In the latter case, high precision of signal
resolution is required. These data can also be used to develop 6. Apparatus
an R-curve. Other procedures for measuring crack extension 6.1 Apparatus is required for measurement of applied force,
are allowed. load-line displacement, and crack-mouth opening displace-
4.4 The commonality of instrumentation and recommended ment. Force versus load-line displacement and force versus
testing procedure contained herein permits the application of crack-mouth opening displacement may be recorded digitally
data to more than one method of evaluating fracture toughness. for processing by computer or autographically with an x-y
Annex A4 and Annex A6 – Annex A11 define the various data plotter. Test fixtures for each specimen type are described in the
treatment options that are available, and these should be applicable Annex.
reviewed to optimize data transferability. 6.2 Displacement Gages:
4.5 Data that are generated following the procedures and 6.2.1 Displacement measurements are needed for the fol-
guidelines contained in this test method are labeled qualified lowing purposes: to evaluate J from the area under the force
data. Data that meet the size criteria in Annex A4 and Annex versus load-line displacement record, CTOD from the force
A6 – Annex A11 are insensitive to in-plane dimensions. versus crack-mouth opening displacement record and, for the
elastic compliance method, to infer crack extension, ∆ap, from
4.6 Supplementary information about the background of elastic compliance calculations.
this test method and rationale for many of the technical 6.2.2 The recommended displacement gage has a working
requirements of this test method are contained in (3). The range of not more than twice the displacement expected during
formulas presented in this test method are applicable over the the test. When the expected displacement is less than 3.75 mm
range of crack size and specimen sizes within the scope of this (0.15 in.), the gage recommended in Fig. 2 may be used. When
test method. a greater working range is needed, an enlarged gage such as the
one shown in Fig. 3 is recommended. Accuracy shall be within
5. Significance and Use 61 % of the full working range. In calibration, the maximum
5.1 Assuming the presence of a preexisting, sharp, fatigue deviation of the individual data points from a fit (linear or
crack, the material fracture toughness values identified by this curve) to the data shall be less than 60.2 % of the working
test method characterize its resistance to: (1) fracture of a range of the gage when using the elastic compliance method
stationary crack, (2) fracture after some stable tearing, (3) and 61 % otherwise. Knife edges are required for seating the
stable tearing onset, and (4) sustained stable tearing. This test gage. Parallel alignment of the knife edges shall be maintained
method is particularly useful when the material response to within 1°. Direct methods for measuring load-line displace-
cannot be anticipated before the test. Application of procedures ment are described in Refs (3-6).
in Test Method E1921 is recommended for testing ferritic 6.2.2.1 Gage Attachment Methods—The specimen shall be
steels that undergo cleavage fracture in the ductile-to-brittle provided with a pair of accurately machined knife edges that
transition. support the gage arms and serve as the displacement reference
points. These knife edges can be machined integral with the
5.1.1 These fracture toughness values may serve as a basis
specimen or they may be attached separately. Experience has
for material comparison, selection, and quality assurance.
shown that razor blades serve as effective attachable knife
Fracture toughness can be used to rank materials within a
edges. The knife edges shall be positively attached to the
similar yield strength range.
specimen to prevent shifting of the knife edges during the test
5.1.2 These fracture toughness values may serve as a basis method. Experience has shown that machine screws or spot
for structural flaw tolerance assessment. Awareness of differ- welds are satisfactory attachment methods.
ences that may exist between laboratory test and field condi- 6.2.3 For the elastic compliance method, the recommended
tions is required to make proper flaw tolerance assessment. signal resolution for displacement should be at least 1 part in
5.2 The following cautionary statements are based on some 32 000 of the transducer signal range, and signal stability
observations. should be 64 parts in 32 000 of the transducer signal range
5.2.1 Particular care must be exercised in applying to measured over a 10-min period. Signal noise should be less
structural flaw tolerance assessment the fracture toughness than 62 parts in 32 000 of the transducer signal range.
value associated with fracture after some stable tearing has 6.2.4 Gages other than those recommended in 6.2.2 are
occurred. This response is characteristic of ferritic steel in the permissible if the required accuracy and precision can be met
transition regime. This response is especially sensitive to or exceeded.
material inhomogeneity and to constraint variations that may 6.3 Force Transducers:
be induced by planar geometry, thickness differences, mode of 6.3.1 Testing is performed in a testing machine conforming
loading, and structural details. to the requirements of Practices E4. Applied force may be
5.2.2 The J-R curve from bend-type specimens recom- measured by any force transducer capable of being recorded
mended by this test method (SE(B), C(T), and DC(T)) has been continuously. Accuracy of force measurements shall be within
observed to be conservative with respect to results from tensile 61 % of the working range. In calibration, the maximum
loading configurations. deviation of individual data points from a fit to the data shall be

5
 E1820 − 23

FIG. 2 Double-Cantilever Clip-In Displacement gage Mounted by Means of Integral Knife Edges

less than 60.2 % of the calibrated range of the transducer when

using elastic compliance, and 61 % otherwise.
6.3.2 For the elastic compliance method, the signal resolu-
tion on force should be at least 1 part in 4000 of the transducer
signal range and signal stability should be 64 parts in 4000 of
the transducer signal range measured over a 10-min period.
Recommended maximum signal noise should be less than 62
parts in 4000 of the transducer signal range.
6.4 System Verification—It is recommended that the perfor-
mance of the force and displacement measuring systems be
verified before beginning a series of continuous tests. Calibra-
tion accuracy of displacement transducers shall be verified with
due consideration for the temperature and environment of the
test. Force calibrations shall be conducted periodically and
documented in accordance with the latest revision of Practices
E4.
6.5 Fixtures:
6.5.1 Bend-Test Fixture—The general principles of the
bend-test fixture are illustrated in Fig. 4. This fixture is
designed to minimize frictional effects by allowing the support
rollers to rotate and move apart slightly as the specimen is
loaded, thus permitting rolling contact. Thus, the support
rollers are allowed limited motion along plane surfaces parallel
to the notched side of the specimen, but are initially positively
positioned against stops that set the span length and are held in
place by low-tension springs (such as rubber bands). Fixtures
and rolls shall be made of high hardness (greater than 40 HRC)
NOTE 1—All dimensions are in millimeters. steels.
FIG. 3 Clip Gage Design for 8.0 mm (0.3 in.)
and More Working Range
6.5.2 Tension Testing Clevis:
6.5.2.1 A loading clevis suitable for testing compact speci-
mens is shown in Fig. 5. Both ends of the specimen are held in
such a clevis and loaded through pins, in order to allow rotation

6
 E1820 − 23
satisfactory fracture toughness test result. The most effective
artifice for this purpose is a narrow notch from which extends
a comparatively short fatigue crack, called the precrack. (A
fatigue precrack is produced by cyclically loading the notched
specimen for a number of cycles usually between about 104
and 106 depending on specimen size, notch preparation, and
stress intensity level.) The dimensions of the notch and the
precrack, and the sharpness of the precrack shall meet certain
conditions that can be readily met with most engineering
materials since the fatigue cracking process can be closely
controlled when careful attention is given to the known
contributory factors. However, there are some materials that
are too brittle to be fatigue-cracked since they fracture as soon
as the fatigue crack initiates; these are outside the scope of the
present test method.
7.4.1 Fatigue Crack Starter Notch—Three forms of fatigue
FIG. 4 Bend Test Fixture Design
crack starter notches are shown in Fig. 6. To facilitate fatigue
cracking at low stress intensity factor levels, the root radius for
a straight-through slot terminating in a V-notch should be 0.08
of the specimen during testing. In order to provide rolling mm (0.003 in.) or less. If a chevron form of notch is used, the
contact between the loading pins and the clevis holes, these root radius may be 0.25 mm (0.010 in.) or less. In the case of
holes are provided with small flats on the loading surfaces. a slot tipped with a hole it will be necessary to provide a sharp
Other clevis designs may be used if it can be demonstrated that stress raiser at the end of the hole. The combination of starter
they will accomplish the same result as the design shown. notch and fatigue precrack shall conform to the requirements of
Clevises and pins should be fabricated from steels of sufficient Fig. 7.
strength (greater than 40 HRC) to elastically resist indentation
of the clevises or pins. 7.4.2 Fatigue Crack Size—The crack size (total average
6.5.2.2 The critical tolerances and suggested proportions of length of the crack starter configuration plus the fatigue crack)
the clevis and pins are given in Fig. 5. These proportions are shall be between 0.45 and 0.70 W for J and δ determination.
based on specimens having W/B = 2 for B > 12.7 mm (0.5 in.) 7.4.3 Equipment—The equipment for fatigue cracking
and W/B = 4 for B ≤ 12.7 mm. If a 1930-MPa (280 000-psi) should be such that the stress distribution is uniform through
yield strength maraging steel is used for the clevis and pins, the specimen thickness; otherwise the crack will not grow
adequate strength will be obtained. If lower-strength grip uniformly. The stress distribution should also be symmetrical
material is used, or if substantially larger specimens are about the plane of the prospective crack; otherwise the crack
required at a given σYS/E ratio, then heavier grips will be may deviate from that plane and the test result can be
required. As indicated in Fig. 5 the clevis corners may be cut significantly affected. The K calibration for the specimen, if it
off sufficiently to accommodate seating of the clip gage in is different from the one given in this test method, shall be
specimens less than 9.5 mm (0.375 in.) thick. known with an uncertainty of less than 5 %. Fixtures used for
6.5.2.3 Careful attention should be given to achieving good precracking should be machined with the same tolerances as
alignment through careful machining of all auxiliary gripping those used for testing.
fixtures. 7.4.4 Fatigue Loading Requirements—Allowable fatigue
force values are limited to keep the maximum stress intensity
7. Specimen Size, Configuration, and Preparation applied during precracking, KMAX, well below the material
7.1 Specimen Configurations—The configurations of the fracture toughness measured during the subsequent test. The
standard specimens are shown in Annex A1 – Annex A3. fatigue precracking shall be conducted with the specimen fully
heat-treated to the condition in which it is to be tested. No
7.2 Crack Plane Orientation—The crack plane orientation
intermediate treatments between precracking and testing are
shall be considered in preparing the test specimen. This is
allowed. There are several ways of promoting early crack
discussed in Terminology E1823.
initiation: (1) by providing a very sharp notch tip, (2) by using
7.3 Alternative Specimens—In certain cases, it may be a chevron notch (Fig. 6), (3) by statically preloading the
desirable to use specimens having W/B ratios other than two. specimen in such a way that the notch tip is compressed in a
Suggested alternative proportions for the single-edge bend direction normal to the intended crack plane (to a force not to
specimen are 1 ≤ W/B ≤ 4 and for the compact (and disk shaped exceed Pm as defined in Annex A1 – Annex A3), and (4) by
compact) specimen are 2 ≤ W/B ≤ 4. However, any thickness using a negative fatigue force ratio; for a given maximum
can be used as long as the qualification requirements are met. fatigue force, the more negative the force ratio, the earlier
7.4 Specimen Precracking—All specimens shall be pre- crack initiation is likely to occur. The peak compressive force
cracked in fatigue. Experience has shown that it is impractical shall not exceed Pm as defined in Annex A1 – Annex A3.
to obtain a reproducibly sharp, narrow machined notch that 7.4.5 Fatigue Precracking Procedure—Fatigue precracking
will simulate a natural crack well enough to provide a can be conducted under either force control or displacement

7
 E1820 − 23

NOTE 1—Corners may be removed as necessary to accommodate the clip gage.

FIG. 5 Tension Testing Clevis Design

displacement cycle is maintained constant, the reverse will

happen. The initial value of the maximum fatigue force should
be less than Pm. The specimen shall be accurately located in the
loading fixture. Fatigue cycling is then begun, usually with a
sinusoidal waveform and near to the highest practical fre-
quency. There is no known marked frequency effect on fatigue
precrack formation up to at least 100 Hz in the absence of
adverse environments. The specimen should be carefully
monitored until crack initiation is observed on one side. If
crack initiation is not observed on the other side before
appreciable growth is observed on the first, then fatigue cycling
should be stopped to try to determine the cause and find a
remedy for the unsymmetrical behavior. Sometimes, simply
turning the specimen around in relation to the fixture will solve
the problem.
7.4.5.1 The fatigue precrack extension from the machined
FIG. 6 Fatigue Crack Starter Notch Configurations notch at the nine measurement points along the crack front (see
8.5.3) shall not be less than 0.5h where h is the notch height,
control. If the force cycle is maintained constant, the maximum or 0.25 mm, whichever is larger, and the combination of
K and the K range will increase with crack size; if the precrack size and sharpened notch length shall not be less than

8
 E1820 − 23

NOTE 1—The crack-starter notch shall be centered between the top and bottom specimen edges within 0.005 W.
FIG. 7 Envelope of Fatigue Crack and Crack Starter Notches

2.0h. Precracking shall be accomplished in at least two steps. 7.4.5.3 To transition between steps, intermediate levels of
For the first step the maximum stress intensity factor applied to force shedding can be used if desired.
the specimen shall be limited by: 7.5 Side Grooves—Side grooves are highly recommended
K MAX 5S D~
σ YS
f

σ YS
T
f
0.063σ YS MPa=m ! (6)
when the compliance method of crack size prediction is used.
The specimen may also need side grooves to ensure a straight
crack front as specified in Annex A4 – Annex A11. The total
or thickness reduction shall not exceed 0.25B. A total reduction of

S D~
σ YS
f 0.20B has been found to work well for many materials. Any
K MAX 5
σ YS
T
f
0.4σ YS ksi=in. ! included angle of side groove less than 90° is allowed. Root
radius shall be 0.5 6 0.2 mm (0.02 6 0.01 in.). In order to
where: produce nearly straight fatigue precrack fronts, the precracking
σYSf and σYST = the material yield stresses at the fatigue should be performed prior to the side-grooving operation. BN is
precrack and test temperatures respectively. the minimum thickness measured at the roots of the side
grooves. The root of the side groove should be located along
7.4.5.2 It is generally most effective to use R = PMIN/PMAX
the specimen centerline.
= 0.1. The accuracy of the maximum force values shall be
known within 65 %. Precracking should be conducted at as
8. Procedure
low a KMAX as practical. For some aluminum alloys and high
strength steels the above KMAX relationship can give very high 8.1 Objective and Overview:
precracking forces. This is especially true if precracking and 8.1.1 The overall objective of the test method is to develop
testing are conducted at the same temperature. It is suggested a force-displacement record that can be used to evaluate K, J,
that the user start with approximately 0.7 KMAX given by the or CTOD. Two procedures can be used: (1) a basic procedure
above relationship, and if the precrack does not grow after 105 directed toward evaluation of a single K, J, or CTOD value
cycles the loading can be incrementally increased until the without the use of crack extension measurement equipment, or
crack begins to extend. For the second precracking step, which (2) a procedure directed toward evaluation of a complete
shall include at least the final 50 % of the fatigue precrack, the fracture toughness resistance curve using crack extension
maximum stress intensity factor that may be applied to the measurement equipment. This also includes the evaluation of
specimen shall be given by: single-point toughness values.
8.1.2 The basic procedure utilizes a force versus displace-
σ YS
f
K MAX 5 0.6 T KF (7) ment plot and is directed toward obtaining a single fracture
σ YS
toughness value such as Jc, KJIc, or δc. Optical crack measure-
where: ments are utilized to obtain both the initial and final physical
KF = KQ, KJQ, KJQc or KJQu depending on the result of the crack sizes in this procedure. Multiple specimens can be used
test, and KF is calculated from the corresponding JF to evaluate J at the initiation of ductile cracking, JIc or δIc.
using the relationship that: 8.1.3 The resistance curve procedure utilizes an elastic

Œ
unloading procedure or equivalent procedure to obtain a J- or
EJF CTOD-based resistance curve from a single specimen. Crack
KF 5 (8)
~ 1 2 ν 2! size is measured from compliance in this procedure and

9
 E1820 − 23
verified by post-test optical crack size measurements. An specimen with respect to the clevis opening within 0.76 mm
alternative procedure using the normalization method is pre- (0.03 in.). Seat the displacement gage in the knife edges firmly
sented in Annex A15: Normalization Data Reduction Tech- by wiggling the gage lightly.
nique.
8.4 Basic Procedure—Load all specimens under displace-
8.1.4 Three or more determinations of the fracture tough-
ness parameter are suggested to ascertain the effects of material ment gage or machine crosshead or actuator displacement
and test system variability. If fracture occurs by cleavage of control. If a loading rate that exceeds that specified here is
ferritic steel, the testing and analysis procedures of Test desired, please refer to Annex A14 (“Special Requirements for
Method E1921 are recommended. Rapid-Load J-Integral Fracture Toughness Testing”).
8.4.1 The basic procedure involves loading a specimen to a
8.2 System and Specimen Preparation:
selected displacement level and determining the amount of
8.2.1 Specimen Measurement—Measure the dimensions,
crack extension that occurred during loading.
BN, B, W, H*, and d to the nearest 0.050 mm (0.002 in.) or
0.5 %, whichever is larger. 8.4.2 Load specimens at a constant rate such that the time
8.2.2 Specimen Temperature: taken to reach the force Pm, as defined in Annex A1 – Annex
8.2.2.1 The temperature of the specimen shall be stable and A3, lies between 0.1 and 3 min.
uniform during the test. Hold the specimen at test temperature 8.4.3 If the test ends by fracture instability, measure the
63°C for 1⁄2 h/25 mm of specimen thickness. initial crack size and any ductile crack extension by the
8.2.2.2 Measure the temperature of the specimen during the procedure in 9. Ductile crack extension may be difficult to
test to an accuracy of 63°C, where the temperature is distinguish but should be defined on one side by the fatigue
measured on the specimen surface within W/4 from the crack precrack and on the other by the brittle region. Proceed to
tip. (See Test Methods E21 for suggestions on temperature Section 9 to evaluate fracture toughness in terms of K, J, or
measurement.) CTOD.
8.2.2.3 For the duration of the test, the difference between 8.4.4 If stable tearing occurs, test additional specimens to
the indicated temperature and the nominal test temperature evaluate an initiation value of the toughness. Use the procedure
shall not exceed 63°C. in 8.5 to evaluate the amount of stable tearing that has occurred
8.2.2.4 The term “indicated temperature” means the tem- and thus determine the displacement levels needed in the
perature that is indicated by the temperature measuring device additional tests. Five or more points favorably positioned are
using good-quality pyrometric practice. required to generate an R curve for evaluating an initiation
NOTE 3—It is recognized that specimen temperature may vary more point. See Annex A9 and Annex A11 to see how points shall be
than the indicated temperature. The permissible indicated temperature positioned for evaluating an initiation toughness value.
variations in 8.2.2.3 are not to be construed as minimizing the importance
of good pyrometric practice and precise temperature control. All labora- 8.5 Optical Crack Size Measurement:
tories should keep both indicated and specimen temperature variations as 8.5.1 After unloading the specimen, mark the crack accord-
small as practicable. It is well recognized, in view of the dependency of ing to one of the following methods. For steels and titanium
fracture toughness of materials on temperature, that close temperature
control is necessary. The limits prescribed represent ranges that are
alloys, heat tinting at about 300°C (570°F) for 30 min works
common practice. well. For other materials, fatigue cycling can be used. The use
of liquid penetrants is not recommended. For both recom-
8.3 Alignment:
mended methods, the beginning of stable crack extension is
8.3.1 Bend Testing—Set up the bend test fixture so that the
marked by the end of the flat fatigue precracked area. The end
line of action of the applied force passes midway between the
of crack extension is marked by the end of heat tint or the
support roll centers within 61 % of the distance between the
beginning of the second flat fatigue area.
centers. Measure the span to within 60.5 % of the nominal
length. Locate the specimen so that the crack tip is midway 8.5.2 Break the specimen to expose the crack, with care
between the rolls to within 1 % of the span and square to roll taken to minimize additional deformation. Cooling ferritic steel
axes within 62°. specimens to ensure brittle behavior may be helpful. Cooling
8.3.1.1 When the load-line displacement is referenced from nonferritic materials may help to minimize deformation during
the loading jig, there is potential for introduction of error from final fracture.
two sources. They are the elastic compression of the fixture as 8.5.3 Along the front of the fatigue crack and the front of the
the force increases and indentation of the specimen at the marked region of stable crack extension, measure the size of
loading points. Direct methods for load-line displacement the original crack and the final physical crack size at nine
measurement are described in Refs (4-7). If a remote trans- equally spaced points centered about the specimen centerline
ducer is used for load-line displacement measurement, take and extending to 0.005 W from the root of the side groove or
care to exclude the elastic displacement of the load-train surface of plane-sided specimens. Calculate the original crack
measurement and brinelling displacements at the load points size, ao, and the final physical crack size, ap, as follows:
(8). average the two near-surface measurements, combine the result
8.3.2 Compact Testing—Loading pin friction and eccentric- with the remaining seven crack size measurements and deter-
ity of loading can lead to errors in fracture toughness determi- mine the average. Calculate the physical crack extension, ∆ap
nation. The centerline of the upper and lower loading rods = ap − ao. The measuring instrument shall have an accuracy of
should be coincident within 0.25 mm (0.01 in.). Center the 0.025 mm (0.001 in.).

10
 E1820 − 23
8.5.4 None of the nine measurements of original crack size 8.6.8 After completing the final unloading cycle, return the
and final physical crack size may differ by more than force to zero without additional crosshead displacement be-
0.1(boBN)1/2 from the average physical crack size defined in yond the then current maximum displacement.
8.5.3. 8.6.9 After unloading the specimen, use the procedure in 8.5
8.6 Resistance Curve Procedure: to optically measure the crack sizes.
8.6.1 The resistance curve procedure involves using an 8.7 Alternative Methods:
elastic compliance technique or other technique to obtain the J 8.7.1 Alternative methods of measuring crack extension,
or CTOD resistance curve from a single specimen test. The such as the direct current electric potential difference method
elastic compliance technique is described here, while the described in Annex A18, are allowed. Methods shall meet the
normalization technique is described in Annex A15. qualification criteria given in 9.1.5.2 or in the case of the
8.6.2 Load the specimens under displacement gage or ma- potential difference method, in A18.16.2. If an alternative
chine crosshead or actuator displacement control. Load the method is used to obtain JIc, at least one additional, confirma-
specimens at a rate such that the time taken to reach the force tory specimen shall be tested at the same test rate and under the
P m, as defined in Annex A1 – Annex A3, lies between 0.1 and same test conditions. From the alternative method the load-line
3.0 min, not including the time required to preform unload/ displacement corresponding to a ductile crack extension of 0.5
reload cycles to estimate compliance. The time to perform an mm shall be estimated. The additional specimen shall then be
unload/reload sequence should be as needed to accurately loaded to this load-line displacement level, marked, broken
estimate crack size, but not more than 10 min. If a higher open and the ductile crack growth measured. The measured
loading rate is desired, please refer to Annex A14 (“Special crack extension shall be 0.5 6 0.25 mm in order for these
Requirements for Rapid-Load J-Integral Fracture Toughness results, and hence the JIc value, to be qualified according to this
Testing”). method.
8.6.3 Take each specimen individually through the follow- 8.7.2 If displacement measurements are made in a plane
ing steps: other than that containing the load-line, the ability to infer
8.6.3.1 Measure compliance to estimate the original crack load-line displacement shall be demonstrated using the test
size, ao, using unloading/reloading sequences over a force material under similar test temperatures and conditions. In-
range of 0.5 to 1.0 times the final maximum precracking force. ferred load-line displacement values shall be accurate to within
Estimate a provisional initial crack size, aoq, from at least three 61 %.
unloading/reloading sequences. No individual value shall differ
from the mean by more than 60.002 W. 9. Analysis of Results
8.6.3.2 Proceed with the test using unload/reload sequences
9.1 Qualification of Data—The data shall meet the follow-
that produce crack extension measurements at intervals pre-
ing requirements to be qualified according to this test method.
scribed by the applicable data analysis section of Annex A8 or
If the data do not pass these requirements, no fracture tough-
Annex A10. Note that at least eight (∆a, J) data points are
ness measures can be determined in accordance with this test
required before the specimen achieves maximum force. If
method.
crack size values change negatively by more than 0.005 ao
(backup), stop the test and check the alignment of the loading NOTE 4—This section contains the requirements for qualification that
train. Crack size values determined at forces lower than the are common for all tests. Additional qualification requirements are given
maximum precracking force should be ignored. with each type of test in the Annexes as well as requirements for
determining whether the fracture toughness parameter developed is
8.6.4 For many materials, stress relaxation may occur prior insensitive to in-plane dimensions.
to conducting compliance measurements, causing a time-
dependent nonlinearity in the unloading slope. One method 9.1.1 All requirements on the test equipment in Section 6
that may be used to remedy this effect is to hold the specimen shall be met.
for a period of time until the force becomes stable at a constant 9.1.2 All requirements on machining tolerance and pre-
displacement prior to initiating the unloading. cracking in Section 7 shall be met.
8.6.5 The maximum recommended range of unload/reload 9.1.3 All requirements on fixture alignment, test rate, and
for crack extension measurement should not exceed either temperature stability and accuracy in Section 8 shall be met.
50 % of Pm, as defined in Annex A1 – Annex A3, or 50 % of 9.1.4 The following crack size requirements shall be met in
the current force, whichever is smaller. all stable tests. Unstable tests need only meet the original crack
8.6.6 Experience has shown that satisfactory results may be size requirement.
obtained with unloads of 10-20% of Pm. A consistent force 9.1.4.1 Original Crack Size—None of the nine physical
range should be used for all unloadings in a test. measurements of initial crack size defined in 8.5.3 shall differ
8.6.7 A minimum of twenty (crack opening displacement, by more than 0.1(boBN)1/2 from the average ao.
force) data points, uniformly spaced over the unload interval, 9.1.4.2 Final Crack Size—None of the nine physical mea-
are required to estimate the specimen compliance. The uncer- surements of final physical crack size, ap, defined in 8.5.3 shall
tainty of the compliance estimates can be improved by increas- differ by more than 0.1(boBN)1/2 from the average ap. In
ing the number of data points used in the regression analysis. subsequent tests, the side-groove configuration may be modi-
It is recommended that forty or more data points be used in the fied within the requirements of 7.5 to facilitate meeting this
regression analysis of each unload (reload). requirement.

11
 E1820 − 23
9.1.5 The following crack size requirements shall be met in Basic Test Information
all stable tests using the resistance curve procedure of 8.6. Loading Rate, time to Pm = [min]
Test temperature = [°C]
9.1.5.1 Crack Extension—None of the nine physical mea-
surements of crack extension shall be less than 50 % of the Crack Size Information
average crack extension. Initial measured crack size, ao = [mm]
Initial predicted crack size, aoq = [mm]
9.1.5.2 Crack Extension Prediction—The crack extension, Final measured crack size, af = [mm]
∆apredicted, predicted from elastic compliance (or other Final ∆ap = [mm]
method), at the last unloading shall be compared with the Final ∆apredicted = [mm]

measured physical crack extension, ∆ap. The difference be- Analysis of Results
tween these shall not exceed 0.15 ∆ ap for crack extensions less Fracture type = (Fracture instability or stable tearing)
than 0.2 bo, and the difference shall not exceed 0.03 bo
K Based Fracture
thereafter. KJIc = [MPa-m1/2]
9.2 Fracture Instability—When the test terminates with J Based Fracture
fracture instability, evaluate whether the fracture occurred Jc = [kJ/m2]
before stable tearing or after stable tearing. The beginning of JIc = [kJ/m2]
Ju = [kJ/m2]
stable tearing is defined in A6.3 and A7.3. For fracture
instability occurring before stable tearing proceed to Annex δ Based Results
A6, and Annex A7 to evaluate the toughness values in terms of δc* = [mm]
δIc = [mm]
K, J, or CTOD. For fracture instability occurring after stable δc = [mm]
tearing, proceed to Annex A6, and Annex A7 to evaluate δu = [mm]
toughness values and then go to 9.3 to evaluate stable tearing.
Final ∆a/b =
9.3 Stable Tearing: Final Jmax/σYS = [mm]
9.3.1 Basic Procedure—When the basic procedure is used, Specimen Information
only an initiation toughness can be evaluated. Proceed to Type =
Annex A9 and Annex A11 to evaluate initiation toughness Identification =
Orientation =
values.
9.3.2 Resistance Curve Procedure—When the resistance Basic dimensions
curve procedure is used, refer to Annex A8 and Annex A10 to B = [mm]
BN = [mm]
develop the R-curves. Proceed to Annex A9 and Annex A11 to W = [mm]
develop initiation values of toughness. h (Notch Height) = [mm]
aN(Notch Length) = [mm]

10. Report Particular dimensions

C(T) H = [mm]
10.1 Recommended tables for reporting results are given in SE(B) S = [mm]
Figs. 8 and 9. DC(T) D = [mm]

10.2 Report the following information for each fracture Material

toughness determination: Material designation =
10.2.1 Type of test specimen and orientation of test speci- Form =

men according to Terminology E1823 identification codes, Tensile Properties

10.2.2 Material designation (ASTM, AISI, SAE, and so E (Young’s modulus) = [MPa]
ν (Poisson’s ratio) =
forth), material product form (plate, forging, casting, and so σYS (Yield Strength) = [MPa]
forth), and material yield and tensile strength (at test σTS (Ultimate Strength) = [MPa]
temperature),
Precracking Information
10.2.3 Specimen dimensions (8.2.1), thickness B and BN, Final Pmax = [N]
and width W, Final Pmin = [N]
10.2.4 Test temperature (8.2.2), loading rate (8.4.2 and Pm = [N]
Final ∆K/E = [MPa-m1/2]
8.6.2), and type of loading control, Fatigue temperature = [°C]
10.2.5 Fatigue precracking conditions (7.4), Kmax, ∆K Fatigue crack growth information
range, and fatigue precrack size (average),
FIG. 8 Suggested Data Reporting Format
10.2.6 Load-displacement record and associated calcula-
tions (Section 9),
10.2.7 If the loading rate is other than quasi-static, report the
applied dK/dt,
10.2.8 Original measured crack size, ao (8.5), original 10.2.9 Qualification of fracture toughness measurement
predicted crack size, aoq , final measured crack size, ap, final (Annex A4 and Annex A6 – Annex A11), based on size
predicted crack extension, ∆apredicted, physical crack extension requirements, and based on crack extension, and
during test, ∆ap, crack front appearance—straightness and 10.2.10 Qualified values of fracture toughness, including
planarity, and fracture appearance, R-curve values.

12
 E1820 − 23

FIG. 9 Suggested Data Reporting Format

11. Precision and Bias research report.6 In addition, the overall analysis procedures of
11.1 Bias—There is no accepted “standard” value for any of this test method were evaluated in an interlaboratory test
the fracture toughness criteria employed in this test method. In program. Note that for the evaluation of JIc, if the slope of the
the absence of such a true value no meaningful statement can power law regression line, dJ/da, evaluated at the abscissa
be made concerning bias of data. value ∆aQ is greater than σY, the uncertainty of the JIc
measurement is likely to be much greater than that obtained
11.2 Precision—The precision of any of the various fracture during the interlaboratory test programs cited. Likewise note
toughness determinations cited in this test method is a function that for the evaluation of δIc , if the slope of the power law
of the precision and bias of the various measurements of linear regression line, dδ/da, evaluated at the abscissa value ∆aQ, is
dimensions of the specimen and testing fixtures, the precision greater than 1, the uncertainty of the δIc measurement is likely
of the displacement measurement, the bias of the force mea- to be much greater than that obtained during the interlaboratory
surement as well as the bias of the recording devices used to test programs cited.
produce the force-displacement record, and the precision of the
constructions made on this record. It is not possible to make 12. Keywords
meaningful statements concerning precision and bias for all
12.1 crack initiation; crack-tip opening displacement;
these measurements. However, it is possible to derive useful
CTOD; ductile fracture; elastic-plastic fracture toughness;
information concerning the precision of fracture toughness
fracture instability; J-integral; JIc; KJic; Jc; δc; plane-strain
measurements in a global sense from interlaboratory test
fracture toughness; resistance curve; stable crack growth
programs. Most of the measures of fracture toughness that can
be determined by this procedure have been evaluated by an
interlaboratory test program. The JIc was evaluated in (9), the 6
Supporting data have been filed at ASTM International Headquarters and may
J-R curve was evaluated in (10), and δc was evaluated in a be obtained by requesting Research Report RR:E24-1013.

13
 E1820 − 23
ANNEXES

(Mandatory Information)

A1. SPECIAL REQUIREMENTS FOR TESTING SINGLE EDGE BEND SPECIMENS

NOTE A1.1—Annex A1 – Annex A3 cover specimen information. A1.4.2 Calculation of J:

For the single edge bend specimen, calculate J as follows:
A1.1 Specimen
J 5 J el1J pl (A1.4)
A1.1.1 The standard bend specimen is a single edge-
notched and fatigue-cracked beam loaded in three-point bend- where:
ing with a support span, S, equal to four times the width, W. Jel = elastic component of J, and
The general proportions of the specimen configuration are Jpl = plastic component of J.
shown in Fig. A1.1.
A1.4.2.1 J Calculations for the Basic Test Method—At a
A1.1.2 Alternative specimens may have 1 ≤ W/B ≤ 4. These point corresponding to v and P on the specimen force versus
specimens shall also have a nominal support span equal to 4W. displacement record, calculate the J integral as follows:
A1.2 Apparatus K 2 ~ 1 2 ν 2!
J5 1J pl (A1.5)
E
A1.2.1 For generally applicable specifications concerning
the bend-test fixture and displacement gage see 6.2 and 6.5.1. where K is from A1.4.1 with a = ao, and
η plA pl
A1.3 Specimen Preparation: J pl 5 (A1.6)
B Nb o
A1.3.1 For generally applicable specifications concerning
specimen configuration and preparation see Section 7. where:
A1.3.2 All specimens shall be precracked in three-point Apl = area under force versus displacement record as shown
bending fatigue based upon the force Pm, as follows: in Fig. A1.2,
ηpl = 1.9 if the load-line displacement is used for Apl,
0.5Bb2o σ Y = 3.667 − 2.199(ao /W) + 0.437(ao /W )2 if the crack
Pm 5 (A1.1)
S mouth opening displacement record is used for Apl,
See 7.4.5 for fatigue precracking requirements. BN = net specimen thickness (BN = B if no side grooves are
present), and
A1.4 Calculation b o = W − ao .
A1.4.1 Calculation of K—For the bend specimen at a force All basic test method J integral values shall be corrected for
P(i), calculate K as follows: crack growth using the following relationship (11):

K ~i! 5 F~ P iS
BBN ! 1/2 W 3/2 G~f a i /W ! (A1.2) J 5 J e⁄0 1
11
J p⁄0

S
α 2 0.5 ∆a
D
(A1.7)
α10.5 b o
where: with α = 1 for SE(B) specimen.

S D f
ai
W
5 (A1.3) A1.4.2.2 J Calculations for the Resistance Curve Test
Method—At a point corresponding to a(i), v(i), and P(i) on the

S D F S DS DS S D S D DG
ai 1/2
ai ai ai ai 2 specimen force versus displacement record, calculate the J
3
W
1.99 2
W W
12 2.15 2 3.93
W
12.7
W
integral as follows:

S DS
2 112
ai
W
12
ai
W D 3/2
J ~i! 5
~ K ~i!! 2 ~ 1 2 v 2!
E
1J pl~ i ! (A1.8)

NOTE 1—The two side planes and the two edge planes shall be parallel and perpendicular as applicable to within 0.5°.
NOTE 2—The machined notch shall be perpendicular to specimen length and thickness to within 62°.
FIG. A1.1 Recommended Single Edge Bend [SE(B)] Specimen

14
 E1820 − 23

FIG. A1.2 Definition of Area for J Calculation Using

the Basic Method

where K(i) is from A1.4.1, and where:

F
J pl~ i ! 5 J pl~ i21 ! 1 S
η pl~ i21 !
b ~ i21 ! DS A pl~ i ! 2 A
BN
pl~ i21 !
DG × (A1.9)
v pl(i) = plastic part of the load-line or crack mouth

C(i)
opening displacement = v(i) − (P(i) C(i)), and
= experimental compliance, (∆v/∆P)(i), corresponding

F 1 2 γ pl~ i21 ! S a ~ i ! 2 a ~ i21 !

b ~ i21 ! DG to the current crack size, ai.
NOTE A1.2—The point P(i), v(i) is the last data point recorded before the
i-th unloading begins and before any hold period for stress relaxation as
where: in accordance with 8.6.4.
ηpl(i−1) = 1.9, and For test methods that do not evaluate an experimental
γpl(i−1) = 0.9 load-line elastic compliance, the load-line compliance C(i) can
if the load-line displacement is used to measure Apl and, be determined from the following equation:

η pl 5 3.667 2 2.199 S D a ~ i21 !

W
10.437
a ~ i21 !
W S D 2

C ~i! 5
1
S S
EBe W 2 a i
D 2
× (A1.11)
and

γ pl 5 0.13112.131 S Da ~ i21 !
W
2 1.465 S D a ~ i21 !
W
2
F 1.193 2 1.98 S Dai
W
14.478
ai
W S D 2
2 4.443 S D ai
W
3
11.739 S DG
ai
W
4

if the crack mouth opening displacement is used to measure where:

Apl.
Be = B − (B − BN)2/B
In Eq A1.9, the quantity Apl(i) − Apl(i–1) is the increment of while for the crack mouth opening displacement case:

S D
plastic area under the chosen force versus plastic displacement
record between lines of constant plastic displacement at points 6S ai
C ~i! 5 × (A1.12)
i−1 and i shown in Fig. A1.3. The quantity Jpl(i) represents the EWBe W

F S D S D S D G
total crack growth corrected plastic J at point i and is obtained 2 3
ai ai ai 0.66
in two steps by first incrementing the existing J pl(i-1) and then 0.76 2 2.28 13.87 2 2.04 1
by modifying the total accumulated result to account for the
W W W ~ 1 2 a i /W ! 2
crack growth increment. Accurate evaluation of Jpl(i) from the where:
Eq A1.9 relationship requires small and uniform crack growth Be = B − (B − BN)2/B
increments consistent with the suggested elastic compliance
The compliance estimated using Eq A1.11 or Eq A1.12
spacing of Annex A8 and Annex A10. The quantity Apl(i) can
should be verified by calibrating against the initial experimen-
be calculated from the following equation:
tal force versus load-line displacement data to assure the
A pl~ i ! 5 A pl~ i21 ! 1 @ P ~ i ! 1P ~ i21 ! # @ v pl~ i ! 2 v pl~ i21 ! # /2 (A1.10) integrity of the load-line displacement measurement system.

15
 E1820 − 23

FIG. A1.3 Definition of Plastic Area for Resistance Curve

J Calculation

A1.4.3 Calculation of Crack Size—For a resistance curve where: J is defined in A1.4.2.1 with a = ao, the original crack
test method using an elastic compliance technique on single size, and then crack growth corrected using Annex A16 and:

S D S D S D
edge bend specimens with crack mouth opening displacements σ YS σ YS 2
σ YS 3
measured at the notched edge, the crack size is given as m 5 A 0 2 A 1*
σ TS
1A 2 *
σ TS
2 A 3*
σ TS
(A1.16)
follows:
ai with :
5 @ 0.999748 2 3.9504 u 1 2.9821 u 2
2 3.21408 u 3 A0 = 3.18-0.22 * (ao/W),
W
(A1.13) A1 = 4.32-2.23 * (ao/W),
A2 = 4.44-2.29 * (ao/W), and
4
1 51.51564 u 2 113.031 u 5 # A3 = 2.05-1.06 * (ao/W).
where: Calculation of δ requires σYS/σTS ≥ 0.5.
1 A1.4.5.2 Calculations of CTOD for the Resistance Curve

F G
u5 1/2 (A1.14) Test Method—For the resistance curve test method, calcula-
B e WEC i
11 tions of CTOD for any point on the force-displacement curve
S/4
are made from the following expression:
Ci = (∆vm/∆P) on an unloading/reloading sequence,
Ji
vm = crack mouth opening displacement at notched edge, δi 5 (A1.17)
m iσ Y
Be = B − (B − BN)2/B.
NOTE A1.3—Crack size on a single edge bend specimen is normally where Ji is defined in A1.4.2.2 with a = ai, the current crack
determined from crack mouth opening compliance. It can be determined
from load-line compliance if the correct calibration is available. size and:
A1.4.4 Other compliance equations are acceptable if the
resulting accuracy is equal to or greater than those described
m 5 A 0 2 A 1* S D
σ YS
σ TS
1A 2 * S D
σ YS
σ TS
2
2 A 3* S D
σ YS
σ TS
3
(A1.18)

and the accuracy has been verified experimentally.

with :
A1.4.5 Calculation of CTOD: A0 = 3.18-0.22 * (ai/W),
A1.4.5.1 Calculation of CTOD for the Basic Test Method— A1 = 4.32-2.23 * (ai/W),
For the basic test method, calculations of CTOD for any point A2 = 4.44-2.29 * (ai/W), and
on the force-displacement curve are made from the following A3 = 2.05-1.06 * (ai/W).
expression:
Calculation of δi requires σYS/σTS ≥ 0.5.
J
δ5 (A1.15)
mσ Y

16
 E1820 − 23

A2. SPECIAL REQUIREMENTS FOR TESTING COMPACT SPECIMENS

A2.1 Specimen A2.3 Specimen Preparation

A2.1.1 The standard compact specimen, C(T), is a single A2.3.1 For generally applicable specifications concerning
edge-notched and fatigue cracked plate loaded in tension. Two specimen size and preparation, see Section 7.
acceptable specimen geometries are shown in Fig. A2.1.
A2.3.2 All specimens shall be precracked in fatigue at a
A2.1.2 Alternative specimens may have 2 ≤ W/B ≤ 4 but
force value based upon the force Pm as follows:
with no change in other proportions.
0.4Bb2o σ Y
A2.2 Apparatus Pm 5 (A2.1)
2W1a o
A2.2.1 For generally applicable specifications concerning
See Section 7 for fatigue precracking requirements.
the loading clevis and displacement gage, see 6.2 and 6.5.2.

FIG. A2.1 Two Compact Specimen Designs That Have Been Used Successfully for Fracture Toughness Testing

17
 E1820 − 23
A2.4 Calculation where:
A2.4.1 Calculation of K—For the compact specimen at a ηpl (i –1) = 2.0 + 0.522 b(i−1)/W, and
force P(i), calculate K as follows: γ(i –1) = 1.0 + 0.76 b(i−1)/W.

K ~i! 5
P ~i!
~ BBN W ! 1/2
f S D
ai
W
(A2.2)
In Eq A2.9, the quantity Apl(i) − Apl(i-1) is the increment of
plastic area under the force versus plastic load-line displace-
ment record between lines of constant displacement at points
with:
i−1 and i shown in Fig. A1.3. The quantity Jpl(i) represents the
f S D ai
W
5 (A2.3) total crack growth corrected plastic J at point i and is obtained
in two steps by first incrementing the existing Jpl(i−1) and then

HS DF S D S D S D S D GJ
by modifying the total accumulated result to account for the
ai ai ai 2 ai 3
ai 4
21
W
0.88614.64
W
2 13.32
W
114.72
W
2 5.6
W
crack growth increment. Accurate evaluation of J pl(i) from the

S D
a i 3/2
above relationship requires small and uniform crack growth
12 increments consistent with the suggested elastic compliance
W
spacing of Annex A8 and Annex A10. The quantity A pl(i) can
A2.4.2 Calculation of J—For the compact specimen calcu- be calculated from the following equation:
late J as follows:
J 5 J el1J pl (A2.4)
@ P ~ i ! 1P ~ i21 ! # @ v pl~ i ! 2 v pl~ i21 ! #
A pl~ i ! 5 A pl~ i21 ! 1 (A2.10)
2
where:
where:
Jel = elastic component of J, and
Jpl = plastic component of J. vpl(i) = plastic part of the load-line displacement,
vi − P(i)CLL(i) , and
A2.4.2.1 J Calculations for the Basic Test Method—For the CLL(i) = experimental compliance, (∆v/∆P)i, corresponding
compact specimen at a point corresponding to ν, P on the to the current crack size, ai.
specimen force versus load-line displacement record, calculate NOTE A2.1—The point P(i), v(i) is the last data point recorded before the
as follows: i-th unloading begins and before any hold period for stress relaxation as
K 2~ 1 2 ν 2! in accordance with 8.6.4.
J5 1J pl (A2.5)
E For test methods that do not evaluate an experimental elastic
where: compliance, CLL(i) can be determined from the following
K is from A2.4.1 with a = ao, and equation:

J pl 5
η pl A pl
B Nb o
(A2.6) C LL~ i ! 5
1
S
W1a i
EBe W 2 a i DF
2

S D S D
2.1630112.219
ai
W
2 20.065
ai
W
2

where:
Apl = area shown in Fig. A1.2,
2 0.9925 S Dai
W
3

120.609 S Dai
W S DG
4

2 9.9314
ai
W
5

(A2.11)

BN = net specimen thickness (BN = B if no side grooves are where:

present),
bo = uncracked ligament, (W − ao), and ~ B 2 B N! 2
Be 5 B 2 (A2.12)
ηpl = 2 + 0.522bo/W. B
All basic test method J integral values shall be corrected for The load-line compliance estimated using Eq A2.11 should
crack growth using the following relationship (11):. be verified by calibrating against the initial experimental
J p⁄0 compliance to assure the integrity of the load-line displacement
S D
J 5 J e⁄0 1 (A2.7)
α 2 0.5 ∆a measurement system.
11
α10.5 b o In an elastic compliance test, the rotation corrected
with α = 0.9 for the C(T) specimen. compliance, Cc(i), described in A2.4.5 shall be used instead of
A2.4.2.2 J Calculation for the Resistance Curve Test CLL(i) in Eq A2.11.
Method—For the C(T) specimen at a point corresponding a (i), A2.4.3 Calculation of Crack Size—For a single specimen
v(i), and P(i) on the specimen force versus load-line displace- test method using an elastic compliance technique on the
ment record calculate as follows:
compact specimen with crack opening displacements measured
~ K ~i!! 2 ~ 1 2 ν 2! on the load-line, the crack size is given as follows:
J ~i! 5 1J pl~ i ! (A2.8)
E
a i /W 5 1.000196 2 4.06319u111.242u 2 2 106.043u 3 1464.335u 4

where K(i) is from A2.4.1, and: 2 650.677u 5 (A2.13)

J pl~ i ! 5 (A2.9)
where:

F J pl ~ i21 ! 1 S η pl ~ i21 !
b ~ i21 ! D A pl~ i ! 2 A pl~ i21 !
BN GF 1 2 γ ~ i21 ! S a ~ i ! 2 a ~ i21 !
b ~ i21 ! DG u5
1
@ B e ECc ~ i ! # 1/2 11
(A2.14)

18
 E1820 − 23

Cc(i) = specimen load-line crack opening elastic compliance where (Fig. A2.2):
(∆v/∆P) on an unloading/reloading sequence cor-
Ci = measured specimen elastic compliance, ∆vm/∆Pm, (at
rected for rotation (see A2.4.5),
Be = B − (B − BN)2/B. the load-line),
Cc(i) = corrected specimen elastic compliance, ∆vc/∆Pc (at
A2.4.4 The calculation of crack size values for C(T) speci- the load-line),
mens is a two-step procedure. First, values of uncorrected H* = initial half-span of the load points (center of the pin
crack size ai are obtained from measured values of load-line holes),
compliance Ci using Eq A2.13 and A2.14. Uncorrected crack Ri = radius of rotation of the crack centerline, (W + a )/2,
size values are then used to calculate the corresponding values where a is the updated crack size,
of the radius of rotation of the crack centerline, Ri, as follows: D = one half of the initial distance between the displace-
W1 a i ment measurement points,
Ri 5 (A2.15) θ = angle of rotation of a rigid body element about the
2
unbroken midsection line, or
A2.4.5 To account for crack opening displacement in C(T)

1 2
specimens, the crack size estimation shall be corrected for v m~i!

S D
D1
rotation. Compliance is corrected as follows: 2 D
θ i 5 arcsin 2 arctan , and (A2.17)
Ci =D 2
1R i 2 Ri

S DS D
C c~i! 5 (A2.16)
H* D
sinθ i 2 cos θ i sinθ i 2 cos θ i vm(i) = total measured load-line displacement at the begin-
Ri Ri
ning of the i-th unloading/reloading cycle,
vc = total corrected load-line displacement at the begin-
ning of the i-th unloading/reloading cycle.
A2.4.6 Other compliance equations are acceptable if the
resulting accuracy is equal to or greater than those described
and the accuracy has been verified experimentally.
A2.4.7 Calculation of CTOD:
A2.4.7.1 Calculation of CTOD for the Basic Test Method—
For the basic test method, calculations of CTOD for any point
on the force-displacement curve are made from the following
expression:
J
δ5 (A2.18)
mσ Y

where J is defined in A2.4.2.1 with a = ao, the original crack

size, and then crack growth corrected using Annex A16 and:

m 5 A 0 2 A 1* S D
σ YS
σ TS
1A 2 *
σ YS
σ TS S D 2
2 A 3* S D
σ YS
σ TS
3
(A2.19)

with: A0=3.62, A1 = 4.21, A2=4.33, and A3=2.00. Calcula-

tion of δ requires σYS/σTS ≥ 0.5.
A2.4.7.2 Calculation of CTOD for the Resistance Curve
Test Method—For the resistance curve test method, calcula-
tions of CTOD for any point on the force-displacement curve
are made from the following expression:
Ji
δi 5 (A2.20)
mσ Y

where J is defined in A2.4.2.2 with a = ai, the current crack

size, and,

m 5 A 0 2 A 1* S D
σ YS
σ TS
1A 2 *
σ YS
σ TS S D 2
2 A 3* S D
σ YS
σ TS
3
(A2.21)

with: A0=3.62, A1 = 4.21, A2=4.33, and A3=2.00. Calcula-

FIG. A2.2 Elastic Compliance Correction for Specimen Rotation tion of δi requires σYS/σTS ≥ 0.5.

19
 E1820 − 23

A3. SPECIAL REQUIREMENTS FOR TESTING DISK-SHAPED COMPACT SPECIMENS

A3.1 Specimen ments of circularity as well as width, W; crack size, a; and

A3.1.1 The standard disk-shaped compact specimen, thicknesses, B and BN , shall be made. Measure the dimensions
DC(T), is a single edge-notched and fatigue cracked plate BN and B to the nearest 0.05 mm (0.002 in.) or 0.5 %,
loaded in tension. The specimen geometry which has been used whichever is larger.
successfully is shown in Fig. A3.1. A3.4.1.1 The specimen blank shall be checked for circular-
ity before specimen machining. Measure the diameter at eight
A3.1.2 Alternative specimens may have 2 ≤ W/B ≤ 4 but equally spaced points around the circumference of the speci-
with no change in other proportions. men blank. One of these measurements shall lie in the intended
A3.2 Apparatus notch plane. Average these readings to obtain the diameter, D.
If any measurement differs from the average diameter, D, by
A3.2.1 For generally applicable specifications concerning more than 5 %, machine the blank to the required circularity.
the loading clevis and displacement gage see 6.2 and 6.5.2. Otherwise, D = 1.35 W.
A3.3 Specimen Preparation A3.4.1.2 Measure the width, W, and the crack size, a, from
the plane of the centerline of the loading holes (the notched
A3.3.1 For generally applicable specifications concerning edge is a convenient reference line but the distance from the
specimen size and preparation, see Section 7. centerline of the holes to the notched edge must be subtracted
A3.3.2 All specimens shall be precracked in fatigue at a to determine W and a). Measure the width, W, to the nearest
force value based upon the force Pm as follows: 0.05 mm (0.002 in.) or 0.5 %, whichever is larger.
0.4Bb2o σ Y A3.5 Calculation
Pm 5 (A3.1)
2W1a o
A3.5.1 Calculation of K—For the DC(T) specimen at a
See 7.4 for precracking requirements. force P(i), calculate K as follows:
A3.4 Procedure P ~i!
~ BBN W ! 1/2 ~ i !
K ~i! 5 f a /W (A3.2)
A3.4.1 Measurement— The analysis assumes the specimen
was machined from a circular blank, and, therefore, measure- where:

NOTE 1—All surfaces shall be perpendicular and parallel as applicable within 0.002 W TIR.
NOTE 2—The intersection of the crack starter notch tips on each surface of the specimen shall be equally distant within 0.005W from the centerline
of the loading holes.
NOTE 3—Integral or attached knife edges for clip gage attachment to the crack mouth may be used.
NOTE 4—For starter-notch and fatigue-crack configuration see Fig. 7.
NOTE 5—Required circularity measurements shall be made at eight equally spaced points around the circumference. One of these points shall be the
notch plane. Average the readings to obtain the radius. All values shall be within 5 % of the average.
FIG. A3.1 Disk-Shaped Compact Specimen, DC(T), Standard Proportions and Dimensions

20
 E1820 − 23

S D f
ai
W
5 (A3.3)
displacement record between lines of constant displacement at
points i−1 and i shown in Fig. A1.3. The quantity Jpl(i)

HS DF S D S D S D S D GJ
2 3 4
represents the total crack growth corrected plastic J at point i
ai ai ai ai ai and is obtained in two steps by first incrementing the existing
21 0.7614.8 2 11.58 111.43 2 4.08
W W W W W

S D 3/2
Jpl(i−1) and then by modifying the total accumulated result to
ai account for the crack growth increment. Accurate evaluation of
12
W Jpl(i) from the preceding relationship requires small and uni-
A3.5.2 Calculation of J—For the DC(T) specimen, calcu- form crack growth increments consistent with the suggested
late J as follows: elastic compliance spacing of Annex A8 and Annex A10. The
J 5 J el1J pl (A3.4) quantity Apl(i) can be calculated from the following equation:

where: @ P ~ i ! 1P ~ i21 ! # @ v pl~ i ! 2 v pl~ i21 ! #

A pl~ i ! 5 A pl~ i21 ! 1 (A3.10)
2
Jel = elastic component of J, and
Jpl = plastic component of J. where :
A3.5.2.1 J Calculation for the Basic Test Method—For the vpl(i) = plastic part of the load-line displacement,
DC(T) specimen at a point corresponding to ν(i), P(i) on the vi − P(i)CLL(i), and
specimen force versus load-line displacement record, calculate CLL(i) = experimental compliance, (∆v/∆P)i, corresponding
as follows: to the current crack size, ai.
K 2~ 1 2 ν 2! For test methods that do not evaluate an experimental elastic
J5 1J pl (A3.5)
E compliance, CLL(i) can be determined from the following
equation:
where K is from A3.5.1 with a = ao, and

1 2
2
a ~i!
η pl A pl 11
J pl 5 (A3.6) 1 W
B Nb o C LL~ i ! 5 × (A3.11)
EBe a ~i!
where: 12
W
Apl = area shown in Fig. A1.2,
BN = net specimen thickness (BN = B if no side grooves are
present),
F 2.046219.6496 S D a ~i!
W
2 13.7346
a ~i!
W S D 2

16.1748 S DG
a ~i!
W
3

bo = uncracked ligament, (W − ao), and

ηpl = 2 + 0.522bo/W. where:
Be = B − (B − BN)2/B.
All basic test method J integral values shall be corrected for
NOTE A3.1—The point P(i), v(i) is the last data point recorded before the
crack growth using the following relationship (11):
i-th unloading begins and before any hold period for stress relaxation as
J p⁄0 in accordance with 8.6.4.

S D
J 5 J e⁄0 1 (A3.7)
α 2 0.5 ∆a The compliance estimated using Eq A3.11 should be verified
11
α10.5 b o by calibrating against the initial experimental compliance to
with α = 0.9 for the DC(T) specimen.
assure the integrity of the load-line displacement measurement
A3.5.2.2 J Calculation for the Resistance Curve Test system.
Method—For the DC(T) specimen at a point corresponding to In an elastic compliance test, the rotation corrected
ai, vi, and Pi on the specimen force versus load-line displace- compliance, Cc(i), described in A3.5.5 shall be used instead of
ment record, calculate as follows: CLL(i) given above.
~ K ~i!! 2 ~ 1 2 v 2! A3.5.3 Calculation of Crack Size—For a single-specimen
J ~i! 5 1J pl~ i ! (A3.8)
E
test method using an elastic compliance technique on DC(T)
where K(i) is from A3.5.1 and: specimens with crack opening displacements measured at the
J 5 (A3.9) load-line, the crack size is given as follows:
pl~ i !

F S D GF G
a ~i!
η ~ i21 ! A pl~ i ! 2 A pl~ i21 ! a 2 a ~ i21 ! 5 0.998193 2 3.88087u10.187106u 2 120.3714u 3 (A3.12)
~i! W
J pl~ i21 ! 1 1 2 γ ~ i21 !
b ~ i21 ! BN b ~ i21 !
245.2125u 4 144.5270u 5
where:
η(i−1) = 2.0 + 0.522 b(i−1)/W, and where:
γ(i−1) = 1.0 + 0.76 b(i−1)/W. 1
u5 (A3.13)
In the preceding equation, the quantity Apl(i) − Apl(i−1) is the @ ~ B e ECc ~ i ! ! 1/2 11 #
increment of plastic area under the force versus load-line

21
 E1820 − 23
where: vm(i) = total measured load-line displacement, at the begin-
Cc(i) = specimen crack opening compliance (∆v/∆P) on an ning of the i-th unloading/reloading cycle.
unloading/reloading sequence, corrected for rotation vc = total corrected load-line displacement at the begin-
(see A3.5.5), ning of the i-th unloading/reloading cycle.
Be = B − (B − BN)2/B.
A3.5.6 Other compliance equations are acceptable if the
A3.5.4 The calculation of crack size values for the DC(T) resulting accuracy is equal to or greater than those described
specimens is a two-step procedure. First, values of uncorrected and the accuracy has been verified experimentally.
crack size ai are obtained from measured values of load-line
compliance Ci using Eq A3.12 and Eq A3.13. Uncorrected A3.5.7 Calculation of CTOD:
crack size values are then used to calculate the corresponding A3.5.7.1 Calculation of CTOD for the Basic Test Method—
values of the radius of rotation of the crack centerline, Ri, as For the basic test method calculations of CTOD for any point
follows: on the force-displacement curve are made from the following
W1a i expression:
Ri 5 (A3.14)
2 J
δ5 (A3.16)
A3.5.5 To account for crack opening displacement in DC(T) mσ Y
specimens, the crack size estimation shall be corrected for
where J is defined in A3.5.2.1 with a = ao, the original crack
rotation. Compliance shall be corrected as follows:
size and then crack growth corrected using Annex A16 and:

S D S D S D
Ci

S DS D
C c~i! 5 (A3.15) σ YS σ YS 2
σ YS 3
H* D m 5 A 0 2 A 1* 1A 2 * 2 A 3* (A3.17)
sinθ i 2 cos θ i sinθ i 2 cos θ i σ TS σ TS σ TS
Ri Ri

where: with: A0=3.62, A1 = 4.21, A=4.33, and A3=2.00. Calculation

of δ requires σYS/σTS ≥ 0.5.
Ri = Radius of rotation of the crack centerline, (W + a)/2, A3.5.7.2 Calculation of CTOD for the Resistance Curve
where a is the updated crack size, Test Method—For the resistance curve test method, calcula-
Ci = measured specimen elastic compliance, ∆vm/∆Pm (at
tions of CTOD for any point on the force-displacement curve
the load-line),
Cc(i) = corrected specimen elastic compliance, ∆vc/∆Pc (at are made from the following expression:
the load-line) Ji
δ5 (A3.18)
H* = initial half-span of the load points (center of the pin mσ Y
holes),
D = one half of the initial distance between the displace- where J is defined in A3.5.2.2 with a = ai, the current crack
ment measurement points, size and,
θ = angle of rotation of a rigid body element about the
unbroken midsection line, or m 5 A 0 2 A 1* S D
σ YS
σ TS
1A 2 *S D
σ YS
σ TS
2
2 A 3* S D
σ YS
σ TS
3
(A3.19)

3= 4
v m~i!

S D
D1
2 D with: A0=3.62, A1 = 4.21, A2=4.33, and A3=2.00. Calcula-
θ i 5arcsin 2arctan , and
D 2 1R i 2 Ri tion of δ requires σYS/σTS ≥ 0.5.

A4. METHODS FOR EVALUATING INSTABILITY AND POP-IN

A4.1 Assessment of Force/Clip Gage Displacement A4.1.2 In the event that significant slow stable crack exten-
Records—The applied force-displacement record ob- sion precedes either unstable brittle crack extension or pop-in,
tained from a fracture test on a notched specimen will usually or a maximum force plateau occurs, the force-displacement
be one of the four types shown in Fig. A4.1. curves will be of the types shown in Fig. A4.1c, Fig. A4.1d,
respectively. These figures illustrate the values of P and ν to be
A4.1.1 In the case of a smooth continuous record in which
used in the calculation of δu.
the applied force rises with increasing displacement up to the
onset of unstable brittle crack extension or pop-in, and where A4.1.3 If the pop-in is attributed to an arrested unstable
no significant slow stable crack growth has occurred (see 3.2 brittle crack extension in the plane of the fatigue precrack, the
and Fig. A4.1a and Fig. A4.1b), the critical CTOD, δc, shall be result must be considered as a characteristic of the material
determined from the force and plastic component of clip gage tested.
displacement, νp , corresponding to the points Pc and νc. NOTE A4.1—Splits and delaminations can result in pop-ins with no

22
 E1820 − 23
arrested brittle crack extension in the plane of the fatigue precrack. ν1 = elastic displacement at pop-in No. 1 (see Fig. A4.2),
For this test method, pop-in crack extension in the plane of Pn = force at the nth pop-in, and
the fatigue precrack can be assessed by a specific change in νn = elastic displacement at the nth pop-in.
compliance. The following procedure may be used to assess NOTE A4.3—νn may be determined graphically or analytically (see Fig.
the significance of small pop-ins (see Fig. A4.1b and Fig. A4.2).
A4.1d). Referring to Figs. A4.1 and A4.2, measure the values yn = force drop at the nth pop-in, and
of Pc and νc or Pu and νu from the test record at points xn = displacement increase at the nth pop-in
corresponding to: (a) the earliest significant pop-in fracture, NOTE A4.4—Although an individual pop-in may be ignored on the basis
that is, for which F > 0.05 and (b) fracture, when pop-ins prior of these criteria, this does not necessarily mean that the lower bound of
to fracture may be ignored, that is, for which F < 0.05 as fracture toughness has been measured. For instance, in an inhomogeneous
follows: material such as a weld, a small pop-in may be recorded because of

S D
fortuitous positioning of the fatigue precrack tip. Thus, a slightly different
ν1 Pn 2 yn fatigue precrack position may give a larger pop-in, which could not be
F512 · (A4.1)
P 1 ν n 1x n ignored. In such circumstances the specimens should be sectioned after
testing, and examined metallographically to ensure that the crack tips have
where: sampled the weld or base metal region of interest (12).
F = factor representing the accumulated increase in compli- A4.1.4 The initial compliance C1 shall be determined by
ance and crack size due to all stable crack extensions, or constructing the tangent OA to the initial portion of the
pop-ins, or both, prior to and including the nth pop-in, force-clip gage displacement curve as shown in Fig. A4.3. The
and initial compliance C1 is the inverse of the slope of the tangent
n = sequential number (see Fig. A4.2) of the last of the line OA:
particular series of pop-ins being assessed.
NOTE A4.2—When only one pop-in occurs, n = 1. When multiple ∆ν g
C1 5 (A4.2)
pop-ins occur it may be necessary to make successive assessments of F ∆P
with n = 1, 2, 3, or more.

NOTE 1—Construction lines drawn parallel to the elastic loading slope to give νp, the plastic component of total displacement, νg.
NOTE 2—In curves b and d, the behavior after pop-in is a function of machine/specimen compliance, instrument response, and so forth..
FIG. A4.1 Types of Force versus Clip Gage Displacement Records

23
 E1820 − 23

NOTE 1—C1 is the initial compliance.

NOTE 2—The pop-ins have been exaggerated for clarity.
FIG. A4.2 Significance of Pop-In

FIG. A4.3 Determination of Initial Compliance

A5. METHOD FOR KIC DETERMINATION

A5.1 This annex has been removed from the standard. See
Test Method E399.

24
 E1820 − 23

A6. FRACTURE INSTABILITY TOUGHNESS DETERMINATION USING J

A6.1 This annex describes the method for characterizing value may be obtained, labeled JQu. In addition, part of an
fracture toughness values based on J, Jc, or Ju, for a fracture R-curve may be developed or the final point may be used in the
instability and the associated requirements for qualifying the evaluation of an initiation toughness value JIc (these are
data according to this test method. Data meeting all of the described in Annex A8 – Annex A11).
qualification requirements of 9.1 and those of this annex result
in qualified values of Jc or Ju. Data meeting the size require- A6.3.1 J is calculated at the final point where instability
ment result in a value of Jc that is insensitive to the in-plane occurs using the J formulas for the basic method including the
dimensions of the specimen. crack growth correction of Annex A16. This point is a Ju value.

A6.2 Fracture Instability Before Stable Tearing—When A6.3.2 Qualification of JQu as Ju—JQu = Ju if crack exten-
fracture occurs before stable tearing, a single-point toughness sion ∆ap ≥ 0.2 mm (0.008 in.) + JQu/2σY.
value may be obtained labeled Jc. A6.4 Significance of Jc and Ju—Values of JQc that meet the
A6.2.1 J is calculated at the final point of instability, using size criteria are labeled Jc and are considered to be insensitive
the J formulas for the basic method including the crack growth to the in-plane dimensions of the specimen. For ferritic steel
correction in Annex A16. This point is labeled JQc, a provi- specimens that have failed unstably by cleavage in the ductile
sional Jc value. to brittle transition, the analysis procedure of Test Method
A6.2.2 Qualification of JQc as Jc—JQc = Jc, a measure of E1921 is recommended. Values of JQc that do not meet validity
fracture toughness at instability without significant stable crack remain JQc and may be size-dependent. Ju is not considered to
extension that is independent of in-plane dimensions, provided be a size-insensitive property and therefore is not subject to a
the following two conditions are both met: (1) B, bo ≥ 100 size criterion. It is a characteristic of the material and specimen
JQc/σY, and (2) crack extension ∆ap < 0.2 mm + JQc/2σY. Note geometry and size. It signifies that at the test temperature the
that even if these conditions are met, Jc may be dependent on material is not completely ductile and can sustain only limited
thickness (length of crack front). R-curve behavior.
A6.3 Fracture Instability After Stable Tearing—When
fracture occurs after stable tearing crack extension ∆ap > 0.2
mm (0.008 in.) + JQc/2σY, a single-point fracture toughness

A7. FRACTURE INSTABILITY TOUGHNESS DETERMINATION USING CTOD (δ)

A7.1 This annex describes the method for characterizing Data that fail to meet the size criterion based on B or bo, but
fracture toughness values based on δ, δc, or δu for a fracture still meet the restriction on crack extension, are labeled δc.
instability and the associated requirements for qualifying the
data according to this test method. Data meeting all of the A7.3 Fracture Instability After Stable Tearing—When
qualification requirements of 9.1 and those in this annex result fracture occurs after stable tearing, crack extension ∆ap ≥ 0.2
in qualified values of δc or δu. Data meeting the size require- mm (0.008 in.) + δQc/1.4, a single-point fracture toughness
ment result in a value of δc* that is insensitive to in-plane value may be obtained, labeled δu. In addition, part of an
dimensions of the specimen. R-curve may be developed or the final point may be used in the
evaluation of an initiation toughness value (these are described
A7.2 Fracture Instability Before Stable Tearing—When in Annex A8 – Annex A11).
fracture occurs before stable tearing, a single-point toughness A7.3.1 δ is calculated at the final point where instability
value may be obtained labeled δc, the force Pc and the clip gage occurs, using the δ formulas for the basic method. This point is
displacement υc, for δc are indicated in Fig. 1. labeled δQu, a provisional δu value.
A7.2.1 δ is calculated at the final point, instability, using the A7.3.2 Qualification of δQu as δu—δQu = δu, if crack
δ formulas from Annex A1 – Annex A3. This point is labeled extension, ∆ap > 0.2 mm (0.008 in.) + δQu/1.4.
δQc, a provisional δc value.
A7.3.3 Significance of δc and δu—Values of δQc that meet
A7.2.2 Qualification of δQc =δc*, a fracture toughness value the qualification requirements are labeled δc* and are consid-
that is insensitive to the in-plane dimensions of the specimen, ered to be insensitive to the in-plane dimensions of the
if the following two conditions are met: (1) B, bo ≥ 300 δQc, specimen. Values of δQc that do not meet the size requirement
and (2) crack extension ∆ap < 0.2 mm (0.008 in.) + δQc/1.4. are labeled δc and may be size-dependent. δu is not considered

25
 E1820 − 23
to be a size-insensitive property and, therefore, is not subject to temperature the material is not completely ductile and can
a size criterion. It is a characteristic of the material and sustain only limited R-curve behavior.
specimen geometry and size. It signifies that at the test

A8. J-R CURVE DETERMINATION

NOTE A8.1—Annex A8 – Annex A11 cover methods for evaluating applied to account for geometry changes due to deformation
toughness for stable tearing. for the compact, C(T), and disk-shaped compact, DC(T),
A8.1 This method describes a single-specimen technique specimens.
for determining the J-R curve of metallic materials. The J-R A8.2.3 If an elastic compliance method is used, the unload/
curve consists of a plot of J versus crack extension in the reload sequences should be spaced with the displacement
region of J controlled growth. The J-R curve is qualified interval not to exceed 0.01bo, the average being about 0.005bo.
provided that the criteria of 9.1 and A8.3 are satisfied.
The use of larger increments between unloadings will lead to
A8.2 J Calculation: less accurate J-R curves although the result will be conserva-
tive. If an initiation value of toughness is being evaluated, more
A8.2.1 J can be calculated at any point on the force versus
load-line displacement record using the equations suggested in unload/reload sequences may be necessary in the early region
the calculation section of Annex A1 – Annex A3 for the of the J-R curve.
different specimen geometries. A8.3 Measurement Capacity of Specimen:
A8.2.2 If a resistance curve method is used, the values of
crack size are calculated using the compliance equations A8.3.1 The maximum J-integral capacity for a specimen is
described in Annex A1 – Annex A3 (or an alternative method given by the smaller of the following:
for measuring crack size). The rotation correction shall be J max 5 b o σ Y /10, or

FIG. A8.1 Typical J-R Curve

26
 E1820 − 23
J max 5 Bσ Y /10. A8.4.1 The J-integral values and the corresponding crack
A8.3.2 The maximum crack extension capacity for a speci- extension values must be plotted as shown in Fig. A8.1. If an
men is given by the following: elastic compliance method is used, shift the J-R curve accord-
ing to the procedure described in A9.3. The J-R curve is
∆a max 5 0.25 b o
defined as the data in a region bounded by the coordinate axes
A8.4 Constructing the J-R Curve: and the Jmax and ∆amax limits given in A8.3.1 and A8.3.2.

A9. JIc and KJIc EVALUATION

A9.1 Significance—The property JIc determined by this A9.4 If the optically measured crack size, ao, differs from
method characterizes the toughness of a material near the onset aoq by more than the larger of 0.01W or 0.5mm, the data set is
of crack extension from a preexisting fatigue crack. The JIc not adequate according to this test method.
value marks the beginning stage of material crack growth
resistance development, the full extent of which is covered in A9.5 Evaluate the final Ji values using the adjusted aoq of
Annex A8. JIc is qualified provided that the criteria of 9.1 and A9.3.3 and the equations of the applicable Annex A1, Annex
A9.9 and A9.10 are satisfied. A2, or Annex A3.

A9.2 J Calculation: A9.6 Calculation of an Interim JQ:

A9.2.1 Calculations of the J integral are made using the A9.6.1 Basic Procedure—For each specimen, calculate ∆a
equations in Annex A1 – Annex A3. as follows:
A9.2.2 A collection of standard data sets, E1820/ ∆a 5 a p 2 a o (A9.2)
1–DS1(2016)-E1820/9–DS9(2020), is available for verifying Resistance Curve Procedure—For each ai value, calculate a
computer algorithms developed to implement the calculations corresponding ∆ai as follows:
to evaluate JIc. See 2.2.
∆a i 5 a i 2 a 0q (A9.3)
A9.3 Corrections and Adjustments to Data: Plot J versus ∆a as shown in Fig. A9.1. Determine a
A9.3.1 If the basic method is used, calculate crack growth construction line in accordance with the following equation:
corrected J values using the procedure of Annex A16. J 5 2σ Y ∆a (A9.4)
A9.3.2 If an elastic compliance method is used, a correction A9.6.2 Plot the construction line, then draw an exclusion
is applied to the estimated ∆ai data values to obtain an line parallel to the construction line intersecting the abscissa at
improved aoq. This correction is intended to obtain the best 0.15 mm (0.006 in.). Draw a second exclusion line parallel to
value of aoq, based on the initial set of crack size estimates, ai, the construction line intersecting the abscissa at 1.5 mm (0.06
data. For data generated using the basic procedure of 8.4, no in.). Plot all J − ∆ a data points that fall inside the area enclosed
adjustments to the crack size and crack extension data are by these two parallel lines and capped by Jlimit = boσY ⁄ 7.5.
necessary. To evaluate JIc using data from the basic procedure,
proceed to A9.6. A9.6.3 Plot a line parallel to the construction and exclusion
lines at an offset value of 0.2 mm (0.008 in.).
A9.3.3 Adjustment of aoq—The value of JQ is very depen-
dent on the aoq used to calculate the ∆ai quantities. The value A9.6.4 At least one J–∆a point shall lie between the
obtained for aoq in 8.6.3.1 might not be the correct value and 0.15-mm (0.006-in.) exclusion line and a parallel line with an
the following adjustment procedure is required. offset of 0.5 mm (0.02 in.) from the construction line as shown
A9.3.3.1 Identify all Ji and ai points that were determined in Fig. A9.2. At least one J−∆a point shall lie between this
before the specimen reached the maximum force for the test. 0.5-mm offset line and the 1.5-mm (0.06-in.) exclusion line.
Use this data set of points to calculate a revised aoq from the Acceptable data are shown in Fig. A9.2. The other J−∆a points
following equation: can be anywhere inside the region of qualified data.
J A9.6.5 JIc is determined by fitting a power law to selected
a 5 a oq1 1BJ2 1CJ3 (A9.1) data that fall in the region defined in A9.6.2.
2σ Y

The coefficients of this equation shall be found using a least

squares fit procedure, see Appendix X1.
J 5 C1 S D
∆a
k
C2
(A9.5)

A9.3.3.2 If the number of points used in A9.3.3.1 to To select the (∆a, J) data points to be used in determination
determine aoq is less than 8 or of these 8 there are less than 3 of the power law coefficients, C1 and C2, start with the last
between 0.4 JQ and JQ or the correlation coefficient of this fit point that falls within the limits defined in A9.6.2 (designate
is less than 0.96, the data set is not adequate to evaluate any the index of this point as n). Working back toward the start of
toughness measures in accordance with this test method. the test one unload/reload at a time, find the first point that falls

27
 E1820 − 23

FIG. A9.1 Definition of Construction Lines for Data Qualification

backup or other anomalies at the start of the test cause a point with low J
to fall to the right of the 0.15 mm exclusion line, thereby either causing C2
> 1 or providing an unreasonably low determination of JQ.
A9.6.6 The intersection of the regression line of A9.6.5 with
the 0.2-mm offset line defines JQ and ∆aQ. To determine this
intersection the following procedure is recommended.
A9.6.6.1 As a starting point estimate an interim JQ(1) = JQ(i)
value from the data plot of Fig. A9.1.
A9.6.6.2 Evaluate ∆a(i) from the following:
J Q~i!
∆a ~ i ! 5 10.2 mm ~ 0.008 in.! (A9.7)
2σ Y

A9.6.6.3 Evaluate an interim JQ(i+1) from the following

power law relationship:

J Q ~ i11 ! 5 C 1 S D
∆a ~ i !
k
C2

(A9.8)

where k = 1.0 mm or 0.0394 in.

FIG. A9.2 Definition of Regions for Data Qualification
A9.6.6.4 Increment i and return to A9.6.6.2 and A9.6.6.3 to
get ∆a(i) and interim JQ(i+1) until the interim JQ from two
to the left of the 0.15 mm exclusion line (designate the index successive iterations values converge to within 60.1 %.
of this point as m). Perform a power law regression analysis A9.6.6.5 Any other numerical method may be used, as long
using the method of least squares using all data points between as the accuracy requirement of A9.6.6.4 is met.
indices m+1 and n. The power law is linearized for regression NOTE A9.2—Examples of numerical methods that can be used are:
analysis in the following form:: bisection method, false position (regula falsi), secant method, Newton

S D ∆a method.
lnJ 5 lnC 1 1C 2 ln (A9.6) NOTE A9.3—The user is reminded that, irrespective of the 60.1 %, or
k
better, accuracy achieved in the iterative process described in A9.6.6.4, the
where k = 1.0 mm or 0.0394 in. Use only the data which typical scatter band of JIc fracture toughness measurements measured in
accordance with this test method will be much larger than 60.1 %.
conform to the requirements stated in the previous sections.
Draw the regression line as illustrated in Fig. A9.1. A9.6.6.6 Project the intercepts of the power law curve with
NOTE A9.1—This data selection procedure avoids the situation where the 0.15-mm (0.006-in.) and the 1.5-mm (0.06-in.) exclusion

28
 E1820 − 23
lines vertically down to the abscissa. This indicates ∆amin and value measured at the last unloading can be taken as JQ and
∆alimit , respectively. Eliminate all data points that do not fall shall meet the requirements of A9.10.1 and A9.10.2, in
between ∆amin and ∆alimit as shown in Fig. A9.1. Also eliminate addition to the main body requirements, to be qualified as JIc.
all data points which lie above the limiting J capacity where A9.8.1 For ferritic steel specimens that have failed unstably
Jlimit = boσY ⁄ 7.5. The region of qualified data is shown in Fig. by cleavage in the ductile to brittle transition, the analysis
A9.2. procedure of Test Method E1921 is recommended.
A9.6.6.7 At least five data points must remain between
∆amin, ∆alimit, and Jlimit. Data point spacing must meet the A9.9 Qualification of Data—The data shall satisfy the
requirements of A9.6.4. If these data points are different from requirements of 9.1 and all of the following requirements to be
those used in A9.6.6 to evaluate JQ, obtain a new value of JQ qualified according to this test method. If the data do not pass
based only on qualified data. these requirements no fracture toughness values can be deter-
mined according to this test method.
A9.7 If the specimen fails by instability or the onset of
A9.9.1 The power coefficient C2 of A9.6.5 shall be less than
cleavage without a J-∆a point exceeding the 0.5 mm offset line,
1.0.
an alternative data set can be used to obtain JQ. In this case at
least 4 data points shall be in Region A of Fig. A9.2 and at least A9.9.2 For the Resistance Curve Procedure the following
one of these points shall fall between the 0.2 mm (0.008 in.) additional requirements must be satisfied:
offset line and the 0.5 mm (0.02 in.) offset line. The data A9.9.2.1 If an elastic compliance method is used, aoq shall
available is fit with the same power law procedure of A9.6.6.1 not differ from ao by more than the larger of 0.01W or 0.5 mm.
– A9.6.6.4 and JQ is evaluated at the intersection of the linear A9.9.2.2 The number of data available to calculate aoq shall
regression line of A9.6.5 and the 0.2 mm offset line. be ≥8; the number of data between 0.4JQ and JQ shall be ≥ 3;
NOTE A9.4—In the case of ductile instability, more stable tests can be and the correlation coefficient of the least squares fit of
achieved by stiffening the test machine, generally by reducing the length A9.3.3.1 shall be greater than 0.96.
of the load train, especially by removing any unnecessary alignment
fixtures or threaded connections, or both. If the test machine is servo A9.10 Qualification of JQ as JIc—JQ = JIc , a size-
controlled, controlling using the CMOD gage signal rather than the stroke independent value of fracture toughness, if:
signal can also improve the stability of the test.
A9.10.1 Thickness, B > 10 JQ ⁄ σY,
A9.8 If the specimen fails unstably without a J-∆a data A9.10.2 Initial ligament, bo > 10 JQ ⁄ σY,
point beyond the 0.2 mm offset line of Fig. A9.2, the maximum
J value measured is evaluated using Annex A6. Additionally, A9.11 Evaluation of KJIc—Calculate KJIc = √(E'JIc) using E'
for all specimens that do not fail by cleavage instability, the J = E/(1−ν2) and the qualified JIc of A9.10.

A10. METHOD FOR δ-R CURVE DETERMINATION

A10.1 This annex describes a single-specimen technique for being about 0.005 W. If an initiation value of toughness is
determining the δ-R curve of metallic materials. The δ-R curve being evaluated, more unload/reload sequences may be neces-
consists of a plot of δ versus crack extension. To measure the sary in the early region of the δ-R Curve.
δ-R curve the resistance curve procedure of 8.6 must be used.
The δ-R curve is qualified provided that the criteria of 9.1 and A10.3 Measurement Capacity of a Specimen:
A10.3 are satisfied. A10.3.1 The maximum δ capacity for a specimen is given as
follows:
A10.2 δ Calculation:
δ max 5 b o /10m
A10.2.1 δ can be evaluated at any point along the force
where m is defined in Annex A1 – Annex A3 for the different
versus load-line displacement record using the equations sug-
specimen geometries.
gested in the calculation section of Annex A1 – Annex A3 for
the different specimen geometries. A10.3.2 The maximum crack extension capacity for a
specimen is given as follows:
A10.2.2 The values of crack size are calculated using the
∆a max 5 0.25 b o .
compliance equations described in Annex A1 – Annex A3. The
rotation correction shall be applied to account for geometry A10.4 Constructing the δ-R Curve:
changes due to deformation for the compact, C(T), and
A10.4.1 The δ values and the corresponding crack extension
disk-shaped compact, DC(T), specimens.
values must be plotted as shown in Fig. A10.1. A δ-R curve is
A10.2.3 The unload/reload sequences should be spaced established by smoothly fitting the data points in the region
with the displacement interval less than 0.01 W, the average bounded by the coordinate axes and the δmax and ∆amax limits.

29
 E1820 − 23

FIG. A10.1 Typical δ-R Curve

A11. METHOD FOR δIc DETERMINATION

A11.1 Significance—The value of CTOD, δIc, determined is < 0.96, the data set is not adequate to evaluate any toughness
by this method characterizes the fracture toughness of materi- measures in accordance with this method.
als near the onset of stable crack extension from a preexisting
fatigue crack. δIc is qualified provided that the criteria of 9.1 A11.4 If the optically measured crack size, ao, differs from
and A11.9 and A11.10 are satisfied. aoq by more than the larger of 0.01 W or 0.5 mm, the data set
is not adequate in accordance with this method.
A11.2 δ Calculation—Calculations of δ are made using the
equations in Annex A1 – Annex A3. A11.5 Evaluate the final δi values using the adjusted aoq of
A11.3.2.1 and the equations of the applicable Annex A1,
A11.3 Corrections and Adjustments to Data: Annex A2, or Annex A3.
A11.3.1 A correction is applied to the estimated ai data A11.6 Calculation of an Interim δQ:
values to obtain an improved aoq. This correction is intended to
obtain the best value of aoq, based on the initial set of crack size A11.6.1 Basic Procedure—for each specimen, calculate ∆a
estimates, ai, data. For data generated using the basic proce- as follows:
dure of 8.4, no adjustments to the data are necessary. To ∆a 5 a p 2 a o (A11.2)
evaluate δIc using data from the basic procedure, proceed to
Resistance Curve Procedure—for each ai value, calculate a
A11.6.
corresponding ∆ai as follows:
A11.3.2 Adjustment of aoq—The value of δQ is very depen- ∆a i 5 a i 2 a 0q (A11.3)
dent on the aoq used to calculate the ∆ai quantities. The value
obtained for aoq in 8.6.3.1 might not be the correct value, and Plot δ versus ∆a as shown in Fig. A11.1. Draw a construction
the following adjustment procedure is required. line in accordance with the following equation:
A11.3.2.1 Identify all δi and ai points that were determined δ 5 1.4 ∆a (A11.4)
before the specimen reached the maximum force for the test.
Use this data set of points to calculate a revised aoq from the A11.6.2 Plot the construction line. Draw an exclusion line
following equation: parallel to the construction line intersecting the abscissa at 0.15
mm (0.006 in.) as shown in Fig. A11.1. Draw a second
δ exclusion line intersecting the abscissa at 1.5 mm (0.06 in.).
a 5 a oq1 1Bδ 2 1Cδ 3
(A11.1)
1.4 Plot all δ-∆ap data points that fall inside the area enclosed by
The coefficients of this equation shall be found using a least these two parallel lines and capped by δlimit = bo ⁄ 7.5m, where
squares fit procedure, see Appendix X1. m is defined in Annex A1 – Annex A3 for the different
A11.3.2.2 If the number of points used in A11.3.2.1 to specimen geometries.
calculate aoq is less than 8, or of these 8 there are less than 3 A11.6.3 One δ-∆ap point must lie between the 0.15-mm
between 0.4δQ and δQ, or the correlation coefficient of this fit (0.006-in.) exclusion line and a parallel line with an offset of

30
 E1820 − 23

FIG. A11.1 Definition of Construction Lines for Data Qualification

0.5 mm (0.02 in.) from the construction line. One δ-∆ap point
must lie between a line parallel to the construction line at an
δQ 5 C1 S D
∆a
k
C2
(A11.5)
offset of 0.5 mm (0.020 in.) and the 1.5-mm exclusion line.
Acceptable data are shown in Fig. A11.2 with at least one point where k = 1 mm (or 0.0394 in.) depending upon units used.
in Region A and at least one point in Region B. The other δ-∆ap This power law can be determined by using a method of least
points can be placed anywhere inside the region of qualified squares to determine a linear regression line of the following
data. form:

A11.6.4 Plot a line parallel to the construction line and

exclusion lines at an offset value of 0.2 mm (0.008 in.).
lnδ 5 lnC 1 1C 2 ln S D∆a
k
(A11.6)

A11.6.5 To establish a crack initiation measurement point Use only the data that conform to the criteria stated in the
under dominant slow-stable crack growth, a power law curve previous sections. Plot the regression line as illustrated in Fig.
fitting procedure shall be used. This has the following form: A11.1.
A11.6.6 The intersection of the regression line of A11.6.4
with the 0.2 mm offset line defines δQ and ∆aQ. To determine
this intersection the following procedure is recommended.
A11.6.6.1 As a starting point, estimate an interim
δQ(1) = δQ(1) value from the data plot of Fig. A11.1.
A11.6.6.2 Evaluate ∆a(1) from the following:
δ Q~1!
∆a ~ 1 ! 5 10.2 mm ~ 0.079in.! (A11.7)
1.4

A11.6.6.3 Evaluate an interim δQ(i+1) from the following

power law relationship:

δ Q ~ i11 ! 5 C 1 S D
∆a ~ 1 !
k
C 2

(A11.8)

where:
k = 1.0 mm or 0.0394 in.
A11.6.6.4 Increment i and return to A11.6.6.2 and A11.6.6.3
to get ∆a(i) and interim δQ(i + 1) until the interim δQ values from
FIG. A11.2 Definition of Regions for Data Qualification two successive iterations converge to within 60.1 %.

31
 E1820 − 23
A11.6.6.5 Any other numerical method may be used, as controlled, controlling using the CMOD gage signal rather than the stroke
long as the accuracy requirement of A11.6.6.4 is met. signal can also improve the stability of the test.

NOTE A11.1—Examples of numerical methods that can be used are: A11.8 If the specimen fails unstably without a δ-∆a data
bisection method, false position (regula falsi), secant method, Newton point beyond the 0.2 mm offset line of Fig. A11.2, the
method. maximum δ value measured is evaluated using Annex A7.
NOTE A11.2—The user is reminded that, irrespective of the 60.1 %, or
Additionally, for specimens that do not fail by cleavage
better, accuracy achieved in the iterative process described in A11.6.6.4,
the typical scatter band of δIc fracture toughness measurements measured instability, the δ value measured at the last unloading can be
in accordance with this Test Method will be much larger than 60.1 % taken as δQ and shall meet only the requirements of A11.9.1 to
A11.6.6.6 Project the intercepts of the power law curve with be qualified as δIc.
the 0.15-mm (0.006-in.) and the 1.5-mm (0.06-in.) exclusion A11.8.1 For ferritic steel specimens that have failed unsta-
lines vertically down to the abscissa. This indicates ∆amin and bly by cleavage in the ductile to brittle transition, the analysis
∆alimit, respectively. Eliminate all data points that do not fall procedure of Test Method E1921 is recommended.
between ∆amin and ∆alimit as shown in Fig. A11.1. Also
eliminate all data points which lie above the limiting δ capacity A11.9 Qualification of Data—The data shall satisfy the
where δlimit = bo/7.5m, where m is defined in Annex A1 requirements of 9.1 and all of the following requirements to be
through Annex A3 for the different specimen geometries. qualified according to this method. If the data do not pass these
A11.6.6.7 At least five data points must remain between requirements, no fracture toughness values can be determined
∆amin, ∆alimit , and δlimit. Data point spacing must meet the according to this method.
requirements of A11.6.3. If these data points are different from A11.9.1 The power coefficient C2 of A11.6.5 shall be less
those used in A11.6.6 to evaluate δQ, obtain a new value of δQ than 1.0.
based only on qualified data. A11.9.2 For the Resistance Curve Procedure the following
additional requirements must be satisfied:
A11.7 If the specimen fails by instability or the onset of
A11.9.2.1 aoq shall not differ from ao by more than the
cleavage without a δ-∆a point exceeding the 0.5 mm offset line,
greater of 0.01W or 0.5 mm.
an alternative data set can be used to obtain δQ. In this case at
A11.9.2.2 The number of data available to calculate aoq
least 4 data points shall be in Region A of Fig. A11.2 and at
shall be ≥ 8; the number of data between 0.4δQ and δQ shall be
least one of these points shall fall between the 0.2 mm (0.008
≥ 3; and the correlation coefficient of the least squares fit of
in.) offset line and the 0.5 mm (0.02 in.) offset line. The data
A11.6.5 shall be greater than 0.96.
available is fit with the same power law procedure of A11.6.6.1
– A11.6.6.4 and δQ is evaluated at the intersection of the linear A11.10 Qualification of δQ as δIc:
regression line of A11.6.5 and the 0.2 mm offset line. δQ = δIc, a size-independent value of fracture toughness, if:
NOTE A11.3—In the case of ductile instability, more stable tests can be
achieved by stiffening the test machine, generally by reducing the length A11.10.1 The initial ligament, bo ≥ 10mδQ,
of the load train, especially by removing any unnecessary alignment where m is defined in Annex A1 through Annex A3 for the
fixtures or threaded connections, or both. If the test machine is servo different specimen geometries.

A12. COMMON EXPRESSIONS

NOTE A12.1—Annex A12 covers miscellaneous information. A12.1.1 The elastic stress intensity factor for a specimen is
expressed as follows:
A12.1 Stress-Intensity Factor:

TABLE A12.1 Parameters for Stress-Intensity Factors

Specimens
SE(B) C(T) DC(T)
ξ 3(S/W) (a/W)1/2 2 + a/W 2 + a/ W
ζ 2(1 + 2a /W) (1 − a/W)3/2 (1 − a/W)3/2 (1 − a/W)3/2
C0 1.99 0.886 0.76
C1 −2.15 4.64 4.8
C2 6.08 −13.32 −11.58
C3 −6.63 14.72 11.43
C4 2.7 −5.6 −4.08
Limits 0 # a /W # 1 0.2 # a/W# 1 0.2 # a/W # 1
S/W = 4 H/ W = 0.6 D/W = 1.35
Refs (13) (13, 14) (15)

32
 E1820 − 23
TABLE A12.2 Parameters for Compliance Expressions
Specimen SE(B) C(T) DC(T)
Location vLL vLL vLL
Y S sW1ad sW1ad
A0
sW2ad
1.193
s W2ad
2.163
sW2ad
2.0462
A1 −1.980 12.219 9.6496
A2 4.478 −20.065 −13.7346
A3 −4.433 −0.9925 6.1748
A4 1.739 20.609 0
A5 0 −9.9314 0
Limits 0 # a /W # 1 0.2 # a/W# 0.2 # a/W #
0.975 0.8
Refs (16) (17) (18)

Pf~ a/W ! v
K5 (A12.1) C5 5 (A12.2)
~ BBN W ! 1/2 P

where: Y2
F S D S D S D S D S DG
a a 2
a 3
a 4
a 5

S D S DF S D S D S D S DG
A 0 1A 1 1A 2 1A 3 1A 4 1A 5
a ξ a a 2
a 3
a 4
B eE W W W W W
f 5 C 0 1C 1 1C 1C 3 1C
W ζ W 2
W W 4
W
A12.2.2 Be = B − (B − BN) 2/B for all cases and the other
A12.1.2 The parameters for f(a/W) are listed in Table A12.1. parameters for compliance are listed in Table A12.2.
A12.2 Compliance from Crack Size:
A12.2.1 Compliance, C , of a specimen is expressed as a
function of crack size as follows:

A13. METHOD FOR RAPID LOADING KIC DETERMINATION

A13.1 This annex has been removed from the standard. See
Test Method E399.

A14. SPECIAL REQUIREMENTS FOR RAPID-LOAD J-INTEGRAL FRACTURE TOUGHNESS TESTING7

A14.1 Scope exceeding those of conventional (static) testing. Standard

A14.1.1 This annex covers the determination of the rate fracture toughness test specimens are prepared as described in
dependent JIc(t) and the J-integral versus crack growth resis- this method, tested under rapid-load or drop weight conditions,
tance curve (J-R(t) curve) for metallic materials under condi- and a J-R(t) curve is calculated. From this J-R(t) curve a JQ(t)
tions where the loading rate exceeds that allowed for conven- can be evaluated using Section 9 of this method. If unstable
tional (static) testing, see 8.4.2. fracture intervenes, a JQc(t) can be evaluated at the onset of
A14.1.2 This international standard was developed in ac- unstable behavior as in the static case.
cordance with internationally recognized principles on stan- A14.2.1.1 Force, load-line displacement, and time are re-
dardization established in the Decision on Principles for the corded for each test. The force versus displacement curve
Development of International Standards, Guides and Recom- resulting from each test is analyzed to ensure that the initial
mendations issued by the World Trade Organization Technical portion of the curve is sufficiently well defined that an
Barriers to Trade (TBT) Committee. unambiguous curve can be determined from the J(t) versus
A14.2 Summary of Requirements crack size (a(t)) data. In addition, a minimum test time is
calculated from the specimen stiffness and effective mass that
A14.2.1 Special requirements are necessary for J-integral sets a maximum allowed test rate for the material and geometry
fracture toughness testing of metallic materials at loading rates being tested. At times less than the minimum test time a
7
significant kinetic energy component is present in the specimen
This test method is an Annex to ASTM E1820. It is under the jurisdiction of
ASTM Committee E08 on Fatigue and Fracture and is the direct responsibility of relative to the internal energy, and the static J integral
Subcommittee E08.08 on Elastic-Plastic and Fracture Mechanics Technology. equations presented in this method are not accurate. Evaluation

33
 E1820 − 23
of a JQ(t) or JQc(t) at a time less than the minimum test time is A14.3.1.10 JQ(t)[FL−1]—In J integral fracture toughness
not allowed by this method. testing, the provisional, rate dependent, J integral at the onset
A14.2.1.2 Evaluation of the J-R(t) curve requires estimation of stable crack extension as defined in this annex.
of crack extension as a function of load-line displacement or A14.3.1.11 dJ/dt [FL−1T−1]—In J integral fracture testing,
time using the normalization method of Annex A15. An elastic the rate of change of the J integral per unit time. Two loading
compliance method cannot be used. A multiple specimen rate quantities are defined in this method, (dJ/dt)I measured
method can be used to evaluate JQ(t) from a series of tests, before JQ(t), and(dJ/dt)T measured after JQ(t), as defined by
which can be corrected using Annex A16 and assembled into a this annex.
J-R(t) curve. The J-R(t) curve is valid if it meets the require-
ments of this method. A14.4 Significance and Use
A14.2.1.3 All of the criteria for the static JIc, Jc , and J-R A14.4.1 The significance of the static J-R curve, JIc, and Jc
curve evaluations apply to the rapid load J integral fracture properties applies also to the case of rapid loading. The J
toughness test. The rapid load J integral resistance curve is integral fracture toughness of certain metallic materials is
denoted J-R(t), the stable initiation property JIc(t), and the sensitive to the loading rate and to the temperature of test. The
unstable initiation property by Jc (t), where the time to reach J-R(t) curve and JIc(t) properties are usually elevated by higher
the instant corresponding to JQ in milliseconds is indicated in test rates while Jc(t) can be dramatically lowered by higher test
the brackets. rates.

A14.3 Terminology A14.5 Apparatus

A14.5.1 Loading—Two types of high rate loading systems
A14.3.1 Definitions: are anticipated. Servohydraulic machines with high flow rate
A14.3.1.1 The definitions given in Terminology E1823 are servovalves and high capacity accumulators, or alternatively,
applicable to this annex. drop weight impact machines can be used. On-specimen force
A14.3.1.2 The definitions given in Section 3 of this method measurements are recommended for high rate tests. Remote
are applicable. force cells or other transducers can be used for high rate tests
A14.3.1.3 Rapid Load—In J integral fracture testing, any if the requirements of this annex are met. Strain gage bridges
loading rate such that the time taken to reach Pm (see 7.4.4) is are recommended for on-specimen force measurement, as
less than 0.1 minutes. shown in Figs. A14.1 and A14.2. For each specimen type, four
A14.3.1.4 Minimum Test Time, tw(t)—In J integral fracture gages are connected to construct a four-arm bridge and
testing, the minimum time to the rate dependent JQ(t) or JQc(t) calibrated statically before the rapid load test (see A14.5.4).
accepted by this method (19). Test times less than tw will lead Strain gages with grid patterns of approximately 0.25B are
to inaccurate J integral results since large kinetic energy recommended. For SE(B) specimens, gages should be posi-
components will be present. In this method: tioned on the specimen mid-plane at the specimen span
2π quarterpoints. For C(T) specimens, the gages should be posi-
tw 5 (A14.1) tioned on the specimen upper and lower surfaces near the
=k s /M eff specimen mid-plane with the gage edge at least 0.1W behind
where: the initial crack, ao.
ks = specimen load-line stiffness, (N/m), A14.5.2 Servohydraulic Testing Fixtures—The fixtures used
Meff = effective mass of the specimen, taken here to be half for static fracture toughness tests generally require some
of the specimen mass (kg). modification for rapid load tests. Slack grip fixtures are often
A14.3.1.5 Test Time, tQ(t)[T]—In J integral fracture testing, necessary to reduce the applied force oscillation and to allow
the observed time to the rate dependent JQ(t). the actuator to accelerate before force is applied to the
specimen. Soft metal absorbers are generally used in drop
A14.3.1.6 Jc(t)[FL−1]—In J integral fracture testing, the
tower tests to reduce the inertial shock caused by the impact of
rate dependent J integral at the onset of fracture instability
the test machine striker on the specimen surface.
prior to the onset of significant stable tearing crack extension,
Both initial and final crack sizes are required by the
see 3.2.12, as defined in this annex.
normalization method of J-R(t) curve development of Annex
A14.3.1.7 JQc(t)[FL −1]—In J integral fracture testing, the A15. The high rate test must be stopped abruptly to obtain a
provisional rate dependent J integral at the onset of fracture limited specimen deformation and a crack extension increment
instability prior to the onset of significant stable tearing crack satisfying the requirement of A15.1.1. Rigid stop block fixtures
extension, as defined in this annex. can be used to obtain the abrupt stop. In some cases a ramp and
A14.3.1.8 Ju(t)[FL−1]—In J integral testing, the rate depen- hold or square wave command signal can be used to obtain
dent J integral at the onset of fracture instability after signifi- limited specimen deformation for the specimen test.
cant stable tearing crack extension, see 3.2.13, as defined in A14.5.3 Drop Tower Testing Fixtures—Special fixtures are
this annex. necessary for drop tower testing according to this standard.
A14.3.1.9 JIc(t)[FL −1]—In J integral testing, the rate de- Recommended fixtures for SE(B) and C(T) specimens are
pendent J integral at the onset of stable crack extension as shown in Figs. A14.3 and A14.4, respectively (20). Stop block
defined in this annex. fixtures are required to obtain a limited extent of stable crack

34
 E1820 − 23

FIG. A14.1 Strain Gages Mounted on SE(B) Specimen for Measurement of Transmitted Force

FIG. A14.2 Strain Gages Mounted on C(T) Specimen for Measurement of Transmitted Force

35
 E1820 − 23

FIG. A14.3 Test Fixture for Drop Tower SE(B) Specimens

FIG. A14.4 Test Fixtures for Drop Tower C(T) Specimens

36
 E1820 − 23
growth for J-R(t) curve development. Soft metal absorbers are vantages are that the output may be non-linear, and the signal
recommended to reduce the initial shock resulting from the conditioners used with these transducers are often the limiting
impact of the drop tower striker on the specimen surface. A component in frequency response of the displacement mea-
high frequency load-line displacement transducer and signal surement system. Capacitive transducers have been designed to
conditioner is required for drop tower tests. fit in the notch of the C(T) specimen as shown in Fig. A14.5.
A14.5.4 Force Transducers—If remote force transducers Fiber-optic transducers have been used to measure load-line
are used, they shall meet the requirements of Practice E4. displacement of SE(B) specimens. If the load-line displace-
Requirements on the measured initial specimen stiffness and on ment is measured relative to the test fixture, care must be taken
the force and displacement signal smoothness are presented in to account for the effects of fixture compliance and brinnelling
A14.7.4. Static calibration of the on-specimen strain gage on the measured displacement, as discussed in 8.3.1.1.
bridge should be done over a force range from 20 to 100 % of A14.5.6 Signal Conditioners—The user is referred to Guide
the final precracking force. At least five force calibration values E1942 for a detailed discussion of requirements for data
shall be used, spaced evenly over this interval, and at least two acquisitions systems. The signal conditioner must have suffi-
repeat data sets are required. The applied force shall exceed 1⁄4 cient bandwidth to capture the transducer signal without
of the calibrated range of the reference force cell used. The introducing distortion.
on-specimen, transmitted force measuring system shall be A14.5.6.1 Signal conditioners shall have a frequency band-
accurate to within 2 % of the final precracking force over the width in excess of 10/tQ for the force signal and 2/t Q for the
calibration range. displacement signal(s). The more stringent requirement on the
A14.5.5 Displacement Transducers—The transducer shall force signal is necessary to obtain an accurate measurement of
have response characteristics that allow it to follow the motion the elastic component of the J integral near crack initiation. No
of the specimen while not introducing excessive mechanical “phase shifting” of transducer signals is allowed by this
noise into the measured displacement. method. The bandwidth required to accurately capture a signal
A14.5.5.1 Cantilever beam displacement gages such as of that frequency will depend on the type of low-pass filter in
those used in static fracture toughness testing may be suitable the signal conditioner, and the tolerable error. If a low-pass
for rapid-load testing (see 6.2). The cantilever beam displace- filter is present in the measurement system it should not
ment gage described in Annex A1 of Test Method E399 has introduce more than 0.5 % measurement error, see Guide
been used successfully at loading times (tQ) slightly less than 1 E1942.
ms. A14.5.7 Data Sampling—The user is referred to Guide
A14.5.5.2 Gap measuring transducers that use either capaci- E1942 for a detailed discussion of requirements for data
tance or optical means to measure displacement have also been acquisitions systems. The rate at which an analog signal is
used successfully in rapid-load testing (20). These transducers sampled to create a digital signal shall be high enough to
have the advantage that they can be rigidly attached to the ensure that the peak value is accurately captured. The rate of
specimen, and the vibration characteristics of the transducer data acquisition shall result in the time per data set being less
generally do not affect the measured displacement. The disad- than tQ/50.

FIG. A14.5 High Rate Capacitance COD gage and C(T) Specimen with Attachment Holes

37
 E1820 − 23

FIG. A14.6 Evaluation of tQ and the Test Rates (dJ/dt)I and (dJ/dt)T

A14.6 Procedure JIc) and the corresponding force PQ and time tQ. If a ductile
A14.6.1 Follow the procedure of Sections 7 and 8 to prepare instability occurs so that the final stable crack size af cannot be
and test specimens. The following items are additional steps determined, the normalization method cannot be used to
necessary for high rate testing. develop the J-R(t) curve or the corresponding JQ for this test
specimen.
A14.6.2 Calculate tw, the minimum test time from Eq
A14.1. The loading rate is optional but the time to reach JQ(t) A14.6.5.1 The dynamic yield strength and dynamic ultimate
or JQc(t) shall not be less than tw. tensile strength at the relevant strain rate are required for the
evaluation of JQ. An approximate equivalent strain rate to be
A14.6.3 For each test, force and load-line displacement are used for dynamic tensile testing shall be obtained from (21, 22)
required as functions of time. Additional crack opening dis-
2σ YS
placement data, electric potential data, or both, can be acquired ε̇ 5 (A14.2)
t QE
as well if desired.
A14.6.4 Install and align the specimen in the test fixtures, where σYS and E are values corresponding to quasistatic
establish the test temperature, conduct the test at the desired strain rates and evaluated at the temperature of the fracture
test rate, collect and store the data required. Remove the test toughness test and tQ is the time to fracture from A14.6.
specimen from the fixture and mark the extent of the ductile A14.6.5.2 If a pop-in is present, refer to Annex A4 to assess
crack growth according to 8.5.3, break the specimen open its significance. If the pop-in is significant, Jc(t) or Ju(t) values
according to 8.5.4 to expose the fracture surface, and measure corresponding to the point of onset can be calculated using
the initial crack size ao, and the final crack size af according to Annex A6. If fracture instability occurs without significant
8.5.4. ductile crack extension, Jc(t) or Ju(t) values corresponding to
A14.6.5 If the specimen is characterized by ductile upper the point of onset can be calculated as defined in Annex A6. If
shelf behavior, the normalization method of Annex A15 can be fracture instability follows significant ductile crack extension,
used to develop the J-R(t) curve for the test specimen. A the J-R(t) and JIc(t) can be determined providing that af is
multi-specimen method can also be used with J evaluated using distinguishable. The validity of the J-R(t) curve and JIc(t) are
the basic method relationships corrected for crack extension subject to the requirements of Annex A8 and Annex A9, and
using Annex A16. Using Section 9, calculate JQ (the tentative Section 9.

38
 E1820 − 23
A14.7 Qualification of the Data specimen within 610 %. Additionally, the measured force
A14.7.1 Test equipment, specimen geometries, specimen displacement data in the region between 0.3∆LLF and 0.8∆LLF
fixture alignment, and measured data must meet all require- should remain within the bounds of the parallel lines con-
ments of Sections 6 – 9, except as specifically replaced in structed on Fig. A14.7. If these requirements are not met, slack
A14.5. Additional requirements specified here are necessary grips or impact absorbers must be added or modified or the test
for high rate testing. rate reduced to obtain a smoother data set that can be qualified
according to this method.
A14.7.2 All of the test equipment requirements of A14.5
shall be met. A14.7.5 If tQ < tw, the test data are not qualified according
to this method. A slower loading rate must be used, or the
A14.7.3 Plot the J integral versus the time as shown in Fig. specimen geometry changed to decrease tw for the test to be
A14.6. If fracture instability occurs, calculate J based on a o qualified according to this method.
using the basic analysis procedure and plot the data up to and
including JQc(t) or JQu(t). Use a linear regression analysis to A14.7.6 If the normalization method of Annex A15 is used
evaluate (dJ/dt)I as shown in the example of Fig. A14.5 using to obtain JIc, the J resistance curve, or both, at least one
the data from 0.5JQ(t) to JQ(t), from 0.5JQc(t) to JQc(t), or from confirmatory specimen must be tested at the same test rate and
0.5 JQu(t) to JQu(t), as the case may be. Extrapolate this line to under the same test conditions. From the normalization method
the abscissa to evaluate the quantity tQ, as shown in Fig. A14.6. the load-line displacement corresponding to a ductile crack
A14.7.3.1 A second loading rate, (dJ/dt )T, is defined as the extension of 0.5 mm shall be estimated. The additional
slope of the J versus time data beyond maximum force, as specimen shall then be loaded to this load-line displacement
shown in Fig. A14.6, over the range from JQ to JQ + 0.5(Jmax level, marked, broken open and the ductile crack growth
−JQ) or the end of test, if fracture instability occurs. measured. The measured crack extension shall be 0.5 6 0.25
mm in order for these results to be qualified according to this
A14.7.4 Plot force versus load-line displacement for the method.
time interval 0 ≤ t ≤ tQ, as shown schematically in Fig. A14.7.
Use a linear regression analysis to evaluate the initial specimen A14.8 Qualifying the High Rate Results
stiffness ks using data over the range from 20 % to 50 % of the
maximum force measured in the test. Plot this best fit line on A14.8.1 All qualification requirements of 9.1, Annex A6,
the figure, and also plot two parallel lines of the same slope Annex A8, Annex A9, and A14.7 must be met to qualify the
with the y-intercept offset by 610 % of Pmax as shown in Fig. J-R(t) curve, JQ(t) as JIc(t), or JQc(t) as Jc(t) according to this
A14.7. Locate the final crossover ∆LLF. method. If the normalization method of Annex A15 is used, the
A14.7.4.1 For this data set to be qualified according to this additional requirements of this annex shall also be met.
method, the compliance, 1/ks, shall agree with the predictions A14.8.2 The maximum crack extension capacity for a
of Eq A2.11 for the C(T) specimen and Eq A1.11 for the SE(B) specimen to qualify the J-R(t) curve is given by the following:

FIG. A14.7 Force Smoothness Verification Schematic

39
 E1820 − 23
∆a max 5 0.15b o (A14.3) measurement, precision of the force measurement, as well as
the precision of the recording devices used to produce the force
A14.9 Report displacement record used to calculate J and crack size. For the
A14.9.1 The report shall include all the items of Section 10 test rates allowed by this annex, if the procedures outlined in
as well as the following: this annex are followed, the force and load-line displacement
A14.9.1.1 The minimum test time, tw, according to A14.6.2. can be measured with an precision comparable with that of the
A14.9.1.2 The PQ and tQ, corresponding to the calculated static loading as described in the main body. If the normaliza-
JQ(t) or JQc(t). tion function method of Annex A15 is used, the crack size and
A14.9.1.3 The (dJ/dt)I, (dJ/dt)T values, or both. crack extension information must be inferred from initial and
A14.9.1.4 If JIc(t) is being reported, the final crack exten- final crack size measurements. The requirement for the addi-
sion obtained on the confirmatory specimen of A14.7.6 shall be tional specimen to be tested near to the point of crack initiation
reported. has been added to validate the JIc(t) measurement. A round
robin used to evaluate the overall test procedures of this
A14.10 Precision and Bias method is reported in (23).
A14.10.1 Precision—The precision of J versus crack A14.10.2 Bias—There is no accepted “standard” value for
growth is a function of material variability, the precision of the measures of elastic-plastic fracture toughness of any material.
various measurements of linear dimensions of the specimen In absence of such a true value, any statement concerning bias
and testing fixtures, precision of the displacement is not meaningful.

A15. NORMALIZATION DATA REDUCTION TECHNIQUE

A15.1 Scope of 4 mm or 15 % of the initial uncracked ligament. Open

A15.1.1 The normalization technique can be used in some source software that can be used to perform the analysis
cases to obtain a J-R curve directly from a force displacement described in this annex is discussed in (27)
record taken together with initial and final crack size measure- A15.2.2 Each force value Pi up to, but not including the
ments taken from the specimen fracture surface. Additional maximum force Pmax, is normalized using:
restrictions are applied (see A15.3) which limit the applicabil- Pi

S D
ity of this method. The normalization technique is described P Ni 5
W 2 a bi η pl (A15.1)
more fully in Herrera and Landes (24) and Landes, et al. (25), WB
W
Lee (26), and Joyce (23). The normalization technique is most
valuable for cases where high loading rates are used, or where where abi is the blunting corrected crack size at the ith data
high temperatures or aggressive environments are being used. point given by:
In these, and other situations, unloading compliance methods Ji
are impractical. The normalization method can be used for a bi 5 a o 1 (A15.2)
2 σY
statically loaded specimens if the requirements of this section
are met. The normalization method is not applicable for low with Ji calculated from:
toughness materials tested in large specimen sizes where large K i2 ~ 1 2 v 2!
amounts of crack extension can occur without measurable Ji 5 1J pli (A15.3)
E
plastic force line displacement.
where Ki and Jpli are calculated as in Annex A1 and Annex
A15.1.2 This international standard was developed in ac-
A2 for each specimen type using the crack size ao.
cordance with internationally recognized principles on stan-
dardization established in the Decision on Principles for the A15.2.3 Each corresponding load-line displacement is nor-
Development of International Standards, Guides and Recom- malized to give a normalized plastic displacement:
mendations issued by the World Trade Organization Technical v pli v i 2 P i C i
Barriers to Trade (TBT) Committee. v' pli 5 5 (A15.4)
W W

A15.2 Analysis where Ci is the specimen elastic load-line compliance based

A15.2.1 The starting point for this analysis is a force versus on the crack size abi, which can be calculated for each
load point displacement record like that shown in Fig. A15.1. specimen type using the equations of Annex A1 and Annex A2.
Also required are initial and final physical crack sizes optically A15.2.4 The final measured crack size shall correspond to a
measured from the fracture surface. This procedure is appli- crack extension of not more than 4 mm or 15 % of the initial
cable only to Test Method E1820 specimen geometries defined uncracked ligament, whichever is less. If this crack extension
in Annex A1 – Annex A3, with 0.45 ≤ ao/W ≤ 0.70 and cannot is exceeded, this specimen cannot be analyzed according to this
be used if the final physical crack extension exceeds the lesser annex.

40
 E1820 − 23

FIG. A15.1 Typical Force versus Displacement Curve

A15.2.5 The final force displacement pair shall be normal- tangency point. If less than ten data pairs are available for this
ized using the same equations as above except that the final fit, including the final measured data pair, this method cannot
measured crack size, af, is used. Typical normalized data are be used.
shown in Fig. A15.2. A15.2.8 An iterative procedure is now used to force PNi,
A15.2.6 A line should be drawn from the final force vpli /W, ai data to lie on Eq A15.5. This involves adjusting the
displacement pair tangent to the remaining data as shown in crack size of each data set to get the normalized force and
Fig. A15.2. Data to the right of this tangent point shall be displacement pair defined in A15.2.2 and A15.2.3 to fall on the
excluded from the normalization function fit. Data with vpli/W function defined in Eq A15.5. To do so, start at the first data
≤ 0.001 shall also be excluded from the normalization function point with νpli/W ≥ 0.002, normalize the force and displace-
fit. ment using the initial measured crack size ao, and compare the
normalized force with the result of the normalization function
A15.2.7 If at least ten data pairs conform with A15.2.6, the
of A15.2.7. Adjust the crack size until the measured PNi and the
data of Fig. A15.2 can be fit with the following required
functional value of P N are within 60.1 %. Each subsequent
analytical normalization function:
data set is treated similarly. If each step is started with the crack
a1b v' pl1c v' pl2 size resulting from the previous data set, only small, positive
PN 5 (A15.5)
d1v' pl adjustments of crack size are necessary, and the process of
where a, b, c, and d are fitting coefficients. This function can obtaining the crack sizes corresponding to each data set is
be fitted to the data of Fig. A15.1 using standard curve fitting relatively rapid.
packages available as part of computer spreadsheet programs A15.2.8.1 The data of Fig. A15.1, normalized and adjusted
or separately. An example fit for the data of Fig. A15.2 is to fit the normalization function of Fig. A15.3, is shown in Fig.
shown in Fig. A15.3. The normalization function shall fit all A15.4.
the data pairs described above (including the final pair) with a A15.2.9 Since force, load-line displacement, and crack size
maximum deviation less than 1 % of the PN at the final point. estimates are now available at each data point, the standard
Data should be evenly spaced between vpli /W = 0.001 and the equations of Annex A1 and Annex A2 are used to evaluate the

41
 E1820 − 23

FIG. A15.2 Normalized Force versus Displacement Curve Showing Points up to Maximum Force and the Final Data Point

FIG. A15.3 The Normalization Function Shown Fitted to the Normalization Data

42
 E1820 − 23

FIG. A15.4 Data is Adjusted, Defining the Crack Size Necessary to Place All Points on the Analytical Normalization Function
(Only a portion of the data is shown for clarity)

FIG. A15.5 The Resulting J-R Curve for this Specimen

J integral at each data point, resulting in a J-R curve as shown

43
 E1820 − 23
in Fig. A15.5. A JIc value can now be evaluated from this J-R A15.5 Precision and Bias
curve using the method of Section Annex A9. A15.5.1 Precision—The precision of the J resistance curve
is a function of material variability, the precision of the various
A15.3 Additional Requirements measurements of linear dimensions of the specimen and testing
fixtures, precision of the displacement measurement, precision
A15.3.1 Requirements presented in 9.1, Annex A8, and
of the force measurement, as well as the precision of the
Annex A9 shall be met to qualify a J-R curve or a JIc value
recording devices used to produce the force displacement
obtained by the normalization method. Additional require-
record used to calculate J and crack size. For the test rates
ments specific to the use of the normalization method are allowed by this annex, if the procedures outlined in this annex
presented below. are followed, the crack size throughout the fracture toughness
A15.3.2 If the normalization method is used to obtain JIc, at test can be measured with a precision comparable with that of
least one additional, confirmatory specimen shall be tested at the unloading compliance procedure described in the main
the same test rate and under the same test conditions. From the body. A round robin describing the use of the normalization
normalization method the load-line displacement correspond- procedure on rapidly loaded SE(B) and C(T) specimens is
ing to a ductile crack extension of 0.5 mm shall be estimated. presented in (23). A requirement for the testing of a confirma-
tory specimen tested near the point of stable crack initiation is
The additional specimen shall then be loaded to this load-line
present to validate the JIc measurement.
displacement level, marked, broken open and the ductile crack
growth measured. The measured crack extension shall be 0.5 6 A15.5.2 Bias—Crack sizes generally vary through the thick-
0.25 mm in order for these results, and hence the JIc value, to ness of fracture toughness specimens. A nine point average
be qualified according to this method. procedure based on optical measurements obtained from the
post-test fracture surface is generally used to give a reportable
A15.4 Report crack size. Different measurements would be obtained using
more or less measurement points. Alternative crack sizes can
A15.4.1 Section 10 describes the reporting requirements for be estimated using compliance methods, which obtain different
this method. If the normalization method is used, the following average crack size estimates for irregular crack front shapes.
additional items shall be reported. Stringent crack front straightness requirements are present in
A15.4.2 If the normalization function is used the coeffi- this standard to minimize differences caused by these effects.
cients of the fit shall be reported as well as the maximum The normalization method acts to interpolate between optically
deviation of the fit and the number of data used. measured crack average lengths measured at the start and end
of the stable resistance curve fracture toughness test. This
A15.4.3 If JIc is reported, the accuracy of the confirmatory method has been demonstrated in (23) to give results consistent
specimen of A15.3.2 shall be reported. with those obtained by unloading compliance procedures.

A16. EVALUATION OF CRACK GROWTH CORRECTED J-INTEGRAL VALUES

A16.1 This annex had been removed from the standard.

Crack growth correction of J-integral values is addressed in
Annex A1–Annex A3.

A17. FRACTURE TOUGHNESS TESTS AT IMPACT LOADING RATES USING PRECRACKED CHARPY-TYPE SPECIMENS

A17.1 Scope properties determined are comparable to conventional large-

A17.1.1 This Annex specifies requirements for performing scale fracture mechanics results when the validity criteria of
and evaluating instrumented impact tests on precracked Annex A8 – Annex A11 and Annex A14 are met. However,
Charpy-type specimens using a fracture mechanics approach. because of the small absolute size of the Charpy specimen, this
Minimum requirements are given for measurement and record- is often not the case. Nevertheless, the values obtained can be
ing equipment such that similar sensitivity and comparable used in research and development of materials, in quality
measurements are achieved. Dynamic fracture mechanics control and service evaluation and to establish the relative

44
 E1820 − 23
variation of properties with test temperature and loading rate A17.4.3 Falling weight testing machines, which may be
measured on precracked Charpy-type specimens. spring assisted, are allowed. The striker is normally instru-
mented to provide force/time or force/time and force/
A17.2 Principle displacement records.
A17.2.1 This Annex prescribes impact bend tests which are A17.4.4 Other testing machines which comply with the
performed on fatigue precracked Charpy-type specimens to calibration and other requirements of Test Method E2298 are
obtain dynamic fracture mechanics properties of materials. not excluded.
This Annex extends the procedure for V-notch impact bend
tests in accordance with Test Methods E23, and may be used A17.4.5 Requirements on Absorbed Energy—The reliability
for evaluation of the Master Curve in accordance with Test of instrumented force values on which these tests are based
Method E1921. Instrumented testing machines are required in depends on the quality of the acquisition system and the
order to utilize this Annex, together with ancillary instrumen- calibration of the instrumented striker. The calibration of the
tation and recording equipment in accordance with Test striker shall be performed in accordance with Test Method
Method E2298. The characteristic fracture toughness param- E2298. Additionally, for each test in which the entire force
eters depend on material response reflected in the force/time signal has been recorded (that is, until the force returns to the
diagrams described in Table A17.1 and Fig. A17.1. Note that baseline), one of the following requirements shall be met:
only Type I diagrams can be linearly fit up to fracture. (a) the difference between KV and Wt shall be within
615 % of KV or 61 J, whichever is larger, or
NOTE A17.1—The symbol used in these Test Methods for force is P, (b) the difference between KV and Wt shall not exceed
while Test Method E2298 uses F. Therefore the parameters Pmax , Pbf, Pgy 625 % or 62 J, whichever is larger.
used in the following sections correspond to the E2298 parameters Fm,
Fbf, Fgy.
For every test that fulfills requirement (b), but not (a), force
values may be adjusted using an iterative procedure until the
A17.3 Specimen Size, Configuration, and Preparation equivalence KV = Wt is achieved (28). If the difference between
KV and Wt exceeds 625 % of KV or 62 J, whichever is larger,
A17.3.1 Specimens shall be prepared in accordance with the
the test shall be discarded and the user shall check and if
dimensions of the type A Charpy impact specimens of Test
necessary repeat the instrumented striker calibration. If record-
Methods E23, with or without the 2.0 mm V-notch, followed
ing of the entire force signal for an individual test is not
by fatigue precracking.
achieved (for example due to the specimen being ejected from
A17.3.2 Fatigue precracking shall be conducted in accor- the machine without being fully broken), the user shall
dance with 7.4. demonstrate conformance of the testing system using at least
five specimens of the same test series, for which the entire
A17.3.3 Specimens are fatigue precracked to produce an
force signal has been recorded, that fulfil one of the above
initial crack size ao in the range 0.45 < ao/W < 0.70.
requirements. Otherwise, conformance shall be demonstrated
A17.3.4 Side-grooving of the specimens in accordance by testing at least five additional non-precracked or precracked
with 7.5 is recommended. Charpy specimens, and showing that in all cases the difference
between KV and Wt is within 615 % of KV or 61 J, whichever
A17.4 Apparatus is larger. If this requirement is not met but the difference
A17.4.1 The preferred testing apparatus is the instrumented between KV and Wt does not exceed 625 % of KV or 62 J, the
Charpy pendulum impact testing machine according to Test force adjustment described above shall be applied.
Method E2298, modified to have a variable pendulum release NOTE A17.2—From a theoretical point of view, KV is expected to be
position. slightly higher than Wt, the difference being due to vibrational energy
losses and other smaller contributions such as secondary impacts between
A17.4.2 Other pendulum machines may be used, with either striker and specimen. For more insight on the difference between KV and
fixed anvil/moving striker or fixed striker/moving anvil, and Wt, see reference (29).
fixed or moving test specimen. The pendulum release position
for such machines is normally variable, and the striker or anvils A17.5 Test Procedures and Measurements
are normally instrumented to provide force/time or force/ A17.5.1 Tests are performed in a manner similar to the
displacement records. standard Charpy impact test of Test Methods E23 and the

TABLE A17.1 Fracture Toughness Properties to be Determined

Material response/fracture behavior Corresponding J-R curve Characteristic
diagram Parameters
type (See Fig. A17.1)
Linear-elastic I ... JcdX, KJcd,X
Elastic-plastic, unstable fracture with ∆a < 0.2 mm II ... Jcd,X (B)
Elastic-plastic, unstable fracture with 0.2 mm # ∆a # 0.15 (W–a0) II ... Jud,X (B, ∆a)
Elastic-plastic, unstable fracture with ∆a $ 0.15 (W–a0) III Jd –∆a JQd,X or JIcd,X
Elastic plastic; no unstable fracture IV Jd –∆a JQd,X or JIcd,X

45
 E1820 − 23

FIG. A17.1 Typical Force-time Diagrams (Schematic)

instrumented impact test of Test Method E2298, especially indicated by the pointer on the analogue scale. From this, the
with regard to the pendulum hammer and the handling of corresponding impact velocity is calculated as:

Œ
pre-cooled or pre-heated specimens.
MC 2 KV0
A17.5.2 Data recording—The force/displacement diagram v o 5 v os (A17.1)
MC
is recorded according to Test Method E2298, from which the where vos is the maximum pendulum velocity corresponding
key data values Pmax, Pbf, Wm, and Wt are determined. In to MC, the full pendulum capacity. A reduced velocity (1 to
addition to the procedures of Test Method E2298, the following 2 m/s) can be advantageous, especially for brittle materials,
procedures are provided concerning impact velocity, available since it reduces the effect of oscillations by lowering their
energy and time to fracture. These data form the basis for relative amplitude and by increasing their number within the
fracture time tf (see A17.5.2.2).
evaluation of toughness parameters according to A17.6 –
A17.9. A17.5.2.2 Time to Fracture—When the time tf to initiate
A17.5.2.1 Impact velocity and available energy—This stan- unstable fracture is less than the minimum test time tw of
dard applies to any impact velocity v0, provided the time to A14.3.1.4, the instant of crack initiation is not detectable in the
fracture fulfills the requirements of A17.5.2.2. Impact veloci- force signal with adequate accuracy because of oscillations
ties for pendulum or falling weight testing machines can be (see Fig. A17.1, Type I), and fracture toughness cannot be
varied by adjusting the striker release height. The impact evaluated using this test method.
velocity v0 for a pendulum machine can be determined as
A17.5.3 Recording Apparatus—Refer to Section 7 of Test
follows: set the pointer to the end-of-scale position as in a
Method E2298.
conventional Charpy test in accordance with Test Methods
E23, release the pendulum from the appropriately reduced A17.5.4 Execution of the Test—Refer to Section 9 of Test
height, with no specimen in place. Read the energy KV0 (in J) Method E2298.

46
 E1820 − 23
A17.5.5 Evaluation of the Force-Displacement Curve— of A14.3.1.4, fracture toughness shall be evaluated according
Refer to Section 11 of Test Method E2298. to the quasi-static approach of Annex A6 and Annex A7.
Impact velocity may be reduced in order for the time to fracture
A17.5.6 Calculation of fracture parameters— The value of
to fulfil the requirements of A14.3.1.4 and A17.5.2.
J-integral at unstable fracture, Jcd (force-time diagrams Type I
and II in Fig. A17.1) or Jud (force-time diagram of Type III in A17.6.3 Stable Crack Extension—In the case of stable crack
Fig. A17.1), or at test termination, Jd (force-time diagram of extension as in Fig. A17.1 (Types III or IV), either multi
Type IV in Fig. A17.1) shall be calculated using the appropriate specimen or single-specimen techniques described in A17.7
formulas Eq A1.4-A1.6 for the Basic Test Method. In and A17.8, respectively, are to be used to determine the J R
particular, the specimen elastic compliance C0 is required to curve. The determination of characteristic fracture toughness
evaluate the plastic component of the area under the force- values from dynamic J-R curves is described in A17.9.
displacement curve (Fig. A1.2). This can be obtained using the
A17.6.4 Loading Rate— As indicated in Table A17.1, frac-
following theoretical expression:
ture toughness values shall be stated with the corresponding
C 0,th 5 C S 1C M (A17.2) loading rate added in parentheses. The latter may be estimated
where CS is the specimen compliance calculated using Eq as follows:
A1.11 and CM is the impact machine compliance. This latter Type I:
can be measured with unnotched specimens using one of the
methods described in (30). Alternatively, if CM is not K Jcd
K̇ 5 (A17.4)
available, C0 can be estimated as the reciprocal of the initial tf
elastic slope (C0,exp), by fitting force-displacement data be- Types I and II:
tween the second oscillation (that is, discarding the first iner- J cd
tia peak) and the onset of general yielding. If both C0,exp and J̇ 5 (A17.5)
tf
C0,th are available, C0,th shall be used and the difference be-
tween the two values shall be within 615 %. Values of or
stress intensity factor shall be obtained from the correspond-
J ud
ing J-integral values using: J̇ 5

Œ
tf
EJd Types III and IV:

S D
K Jd 5 (A17.3)
1 2 v2 P max·v o ao
Calculated KJd values at the onset of cleavage fracture, KJcd, J̇ 5 ·η (A17.6)
B N · ~ W 2 a o ! pl W
can be used to calculate the reference temperature, T0, in In alternative to Eq A17.6, the procedures given in A14.7.3
accordance with Test Method E1921, provided all relevant and A14.7.3.1 can also be used for calculating (dJ/dt)I and
requirements are met. (dJ/dt)T respectively. For practical purposes, the loading rate
A17.5.7 Crack Size Measurements—Original crack size and shall be indicated using its order of magnitude (for example,
final physical crack size shall be measured in accordance with the stress intensity factor corresponding to a loading rate of
8.5. 4 × 105 MPa√m/s shall be indicated as KJcd5).
A17.5.8 Multiple Specimen Tests—To determine dynamic A17.6.5 Dynamic Tensile Properties— The dynamic yield
J-R curves by multi-specimen techniques, the fracture process and ultimate tensile stresses at the relevant strain rate may be
is interrupted over a range of stable crack extension values, that required for certain evaluation procedures and validity checks.
are combined to obtain a single J-R curve. This procedure is An approximate equivalent strain rate for the fracture mechan-
described in A17.7. ics test, which can be used for dynamic tensile testing, may be
calculated from (21, 22):
A17.5.9 Single Specimen Tests—It is also possible to esti-
mate dynamic J-R curves from an individual specimen using σ YS
ε̇ 5 2 (A17.7)
the Normalization Data Reduction technique, as described in t̄·E
A17.8. where: σYS and E are values corresponding to quasistatic
strain rates (that is, conforming to the requirements of Test
A17.6 Analysis of Results Methods E8/E8M) and evaluated at the temperature of the
fracture mechanics test; t̄ is the time to fracture in the case
A17.6.1 Fracture Behavior—The adequacy of fracture of small scale yielding (Type I in Fig. A17.1), or the time
toughness parameters depends on the fracture behavior of the interval of the initial linear part of the force-time record in
test specimen as reflected in the force-displacement diagrams the case of distinct elastic-plastic material behavior (Types
described in Table A17.1. Therefore the measured force dis- II-IV in Fig. A17.1). Eq A17.7 provides a general estimate
placement or force-time diagram shall be assigned to one of the of strain rate values associated with fracture in the test
diagram types shown in Fig. A17.1, using the indications specimen.
provided in Table A17.1.
A17.7 Determination of J-R curves at Impact Loading
A17.6.2 Unstable Fracture—In the case of unstable fracture
Rates by Multiple Specimen Methods
as in Fig. A17.1 (Types I or II), the applicable evaluation
method depends on the oscillations superimposed on the force A17.7.1 The following methods make it possible to deter-
signal. If time to fracture is more than the minimum test time mine fracture toughness parameters in those cases where stable

47
 E1820 − 23
crack extension occurs, Fig. A17.1 (Types III and IV). The corresponding ∆ap represent points on the cleavage J-∆a curve
multi-specimen procedure involves loading a series of nomi- which can be analyzed in accordance with Annex A8 and
nally identical specimens to selected displacement levels, Annex A9. Differences between the temperatures of the various
resulting in corresponding amounts of stable crack extension. resistance points can be neglected, provided they don’t exceed
Each specimen tested provides one point on the resistance 50°C. The requirements of Annex A8 and Annex A9 shall be
curve. The requirements and procedures of Annex A8-Annex satisfied in order to obtain a valid J-R curve. Details of this
A11 concerning number and spacing of data points shall be method are given in (31).
fulfilled.
A17.7.4 The user is warned that results obtained using the
A17.7.2 Low-blow Test—This test procedure is intended to Low-blow or Stop Block methods, in which the loading rate is
limit the impact energy W0 of the pendulum hammer or drop progressively reduced down to zero, may differ from results
weight so that it is sufficient to produce a certain stable crack obtained using tests leading to specimen fracture, such as the
extension, but not sufficient to fully break the specimen. By Cleavage J-R curve method.
selecting different energy levels in a series of tests on nomi-
nally identical specimens, a series of different crack extensions A17.8 J-R Curve Determination by Single Specimen Meth-
∆ai are produced. From the corresponding J-values, J-∆a ods
curves are constructed. A17.8.1 The Normalization Data Reduction (NDR) tech-
nique can be applied to a Low-blow test performed in
NOTE A17.3—An alternative method is the Stop Block approach,
whereby the hammer swing is arrested by using stop blocks, thus avoiding accordance with A17.7.1, provided the measured crack exten-
complete fracture of the specimen. sion does not exceed 15 % of the initial uncracked ligament.
The provisions of Annex A15 apply, including the additional
A17.7.2.1 The following procedure is recommended:
requirements of A15.3. A study published in (32) shows that
(1) Prepare 7 – 10 specimens to nominally the same initial
for two steels and two test temperatures, NDR single-specimen
crack length a0.
results are in good agreement with multiple specimen data
(2) Perform a full blow instrumented impact bending test
generated using the Low-blow technique.
on one of the specimens. Evaluate the energy at maximum
force and the total fracture energy, Wm and Wt, in accordance A17.9 Determination of Fracture Toughness Near the On-
with Test Method E2298. set of Stable Crack Extension
(3) Determine the energy spacing as ∆W0 = 2Wm/N, where
N is the number of available specimens. A17.9.1 From J-R curves determined according to A17.7 or
(4) Perform an impact test by setting the release position of A17.8, fracture toughness values near the onset of stable crack
the pendulum hammer, or the height of the drop weight, such extension can be determined in conformance to Annex A9.
that W0 = 2Wm/N. Avoid a second impact between the striker Specimen qualification in accordance with Annex A9 require-
and the test specimen. ments will be difficult to achieve if the specimen undergoes
(5) Repeat the test on the remaining specimens, increasing significant plasticity during crack extension because of the
the impact energy W0 by the amount ∆W0 =2Wm/N at each test. relatively small size of the specimen. In this case, values of JQd
(6) In order to mark the crack extension, post-test fatigue cannot be regarded as material properties independent of
cycling or heat tinting may be used. specimen size and their use in safety assessments may result in
(7) Break all specimens open after testing. Care is to be non conservative results. Nevertheless, these values can be
taken to minimize post-test specimen deformation. Cooling used for research and development of materials, in quality
ferritic steels may help to ensure brittle behavior during control and service evaluation and to establish the variation of
specimen opening. properties with test temperature.
(8) Measure a0 and ∆ap = ∆ai (where “i” is the test index, A17.9.2 The construction line for JQ calculation shall have
with 1 ≤ i ≤ N-1) in accordance with 8.5. the following expression, see also Eq A9.4:
(9) Calculate Ji according to A1.4.2.1. J 5 2σ Yd ∆a , (A17.8)
(10) Plot the resulting N-1 pairs of values (Ji, ∆ai ) in a J-∆a where σYd, the dynamic effective yield strength, is calculated
diagram and determine the J-R curve according to Annex A8 using the following relationship (33):
and JQd,X (a provisional value of JIcd,X) according to Annex A9.
2.58P Y W
A17.7.2.2 The differences in impact velocity and loading σ Yd 5 , (A17.9)
B ~ W 2 a 0! 2
rate between the various tests are small enough to have a where PY is the average between the force at general yield,
negligible influence on the results and can be ignored, provided Pgy, and the maximum force Pmax, determined from the
velocity and loading rate do not vary by more than a factor 3 force/displacement diagram in accordance with Test Method
between the minimum and maximum values. E2298.
NOTE A17.4—For side-grooved specimens, B = BN.
A17.7.3 Cleavage J-R curve Method—This test method can
only be used for steels that exhibit a brittle-ductile transition. A17.10 Report
The test temperature is varied within the ductile-to-brittle
A17.10.1 In addition to the information listed in Section 10
transition zone so that stable crack extensions of varying
of the main body, the test report shall include the following.
lengths ∆ap are obtained from tests terminated by cleavage
fracture. Jud values calculated according to A1.4.2 and the A17.10.2 Identification and type of testing apparatus.

48
 E1820 − 23
A17.10.3 Striker impact velocity vo (A17.5.3). (1) value of KJcd obtained, if applicable,
A17.10.4 Nominal energy of the striker at velocity vo. (2) value of K̇ obtained, if applicable (only order of
magnitude),
A17.10.5 Absorbed energy KV according to Test Method (3) value of J obtained, if applicable,
E2298. (4) value of J̇ obtained, if applicable (only order of
A17.10.6 Calibration of the instrumented striker. magnitude),
(5) type of force-time diagram, with reference to Fig.
A17.10.7 Details of force adjustment in accordance with
A17.1, Types I – IV,
A17.4.5, if applicable.
(6) for diagrams of Types III or IV, final crack extension,
A17.10.8 Specimen elastic compliance (theoretical or and
experimental, or both) and, if available, machine compliance. (7) a copy of the test record.
A17.10.9 Time to fracture or time at test termination, as A17.10.11 In case of J-R curve determination, values of J
appropriate. and ∆a in tabular form and values of JQd,X or JIcd,X obtained.
A17.10.10 Fracture parameters determined as: A17.10.12 Dynamic tensile properties used, if applicable.

A18. GUIDELINES FOR THE USE OF DIRECT CURRENT ELECTRIC POTENTIAL DIFFERENCE METHODS FOR THE DE-
TERMINATION OF CRACK SIZE

A18.1 Applications potential due to a combination of material resistivity and

A18.1.1 Electric Potential Difference (EPD) procedures for geometric changes that are unrelated to crack extension. These
crack size determination are applicable to metallic materials in changes shall be properly accounted for in order to determine
a wide range of testing environments. The Direct Current EPD crack extension in ductile materials.
(DCEPD) single-specimen technique relies on simple calibra- A18.3.3 Changes in the specimen or instrumentation may
tion functions for standard specimen geometries, and can yield also result in proportional changes in the measured voltage. For
a higher density of data points to define a J-R curve than example, a 1 °C change in specimen temperature can result in
typically achievable using elastic compliance procedures. a noticeable change in the DCEPD signal due to the change in
Specimen geometries for fracture toughness testing explicitly the material’s electric resistivity. Also, some materials exhibit
covered in this annex are the compact, C(T), specimen and the
time-dependent resistivity changes while at elevated tempera-
single-edge bend, SE(B), specimen.
tures (34). Moreover, system-level drift occurs when the
A18.2 Principles response of system components such as amplifiers, relays,
data-acquisition devices, or power supplies change over time.
A18.2.1 Inferring crack extension from electric potential
difference measurements relies on the principle that the electric These phenomena can all produce apparent crack size changes
field in a cracked specimen with a current flowing through it is that introduce errors between the predicted and actual crack
a function of specimen geometry, and in particular of the crack size. One way to compensate for these effects is to normalize
size. For a constant current flow, the electric potential differ- DCEPD measurements using additional voltage measurements
ence across the crack plane will increase with increasing crack taken at a reference location. The reference location may be
size due to modification of the electric field and associated either on the test specimen or on an additional specimen of the
perturbation of the current streamlines. The change in DCEPD same material in the same environment, and powered by the
can be related to crack size through a suitable calibration same electric current source as the test specimen. If the
relationship. DCEPD may be used for both specimen prepara- reference measurements are taken on an additional specimen,
tion (fatigue precracking) and testing (crack initiation and this specimen may be fatigue precracked or blunt-notched. If
stable crack propagation). the reference measurements are taken directly on the test
specimen, the location must be chosen so that the reference
A18.3 Basic Methods voltage is not affected by crack propagation and plastic
A18.3.1 Both direct current (DC) and alternating current deformation. Since all material and instrument variations are
(AC) techniques have been used to measure crack size in also included in the reference measurements, this normaliza-
laboratory specimens. However, this annex only addresses DC tion process will eliminate them. The reference voltage should
techniques since these are more commonly used than AC be of the same order of magnitude as the primary DCEPD
methods. measurements to avoid introducing excessive noise in the
A18.3.2 In the DC method, a constant current is passed measurement signal.
through the specimen resulting in a potential difference across NOTE A18.1—The effects of material and instrument variation are often
the crack plane. The plasticity associated with elastic-plastic not as critical to toughness testing as to tests which require a longer
fracture of ductile materials can increase the measured electric duration, such as fatigue crack growth rate tests.

49
 E1820 − 23
A18.3.4 The DCEPD method can be applied using equip- relative stability of the power supply should be equal to the
ment found in most testing laboratories, as shown in Fig. effective resolution of the voltage measurement system; that is,
A18.1. A computerized system for control and acquisition is if the voltage measurement system can effectively resolve one
strongly recommended. High-gain, low-noise amplification part in 103 of the output voltage from the specimen (including
(typically 30 to 40 dB power gain) may be used to increase the electric noise, inherent inaccuracies such as nonlinearity, and
measured EPD data (typically on the order of 100 µV or less at so forth), then the power supply should be stable to one part in
the specimen) up to voltage levels that are readily recorded 103.
with digital data acquisition instrumentation. Since the test
loading rate must take into account both the response time of A18.5 Voltage Measurement Equipment
the measurement system and the voltage resolution, the provi- A18.5.1 Voltage measurements shall be made with any
sions of this annex only apply to quasi-static tests in accor- equipment that has sufficient resolution, accuracy, and stability
dance with 8.4.2. characteristics. The DCEPD method requires equipment ca-
A18.3.5 The DCEPD method is susceptible to thermoelec- pable of measuring small changes in DC voltage (for example,
tric effects (35) which produce potentials in addition to those 0.05 µV to 0.5 µV) with relatively low DC signal to AC RMS
due to the sample electric field. These can be a substantial noise ratios. Although there are a variety of ways to implement
fraction of the total measured voltage. Since the thermoelectric the voltage measurement system, three commonly used sys-
effects are present even without the input current, it is possible tems are: amplifier/autographic recorder, amplifier/
to account for them by subtracting voltage measurements taken microcomputer analog-to-digital converter, and digital
with the current off from the measurements made with the voltmeter/microcomputer. The use of analog or digital filtering,
current on. An alternative method (Reversing DCEPD, or or both to reduce measurement noise is recommended.
RDCEPD) corrects for the thermoelectric effects by taking
A18.5.2 Autographic recorders are commonly available
voltage measurements while reversing the direction of current
with suitable sensitivity and can be used to record the output
flow. Positive measurements of DCEPD are taken, followed by
voltage directly from the specimen. A preamplifier can be used
reversing the current flow direction and then taking negative
to boost the direct voltage output from the specimen before
DCEPD measurements. The corrected DCEPD signal is given
recording. Another common technique uses a preamplifier to
by one-half of the difference between positive and negative
boost the direct output from the specimen to a level that can be
potential readings. This procedure is repeated for each mea-
digitized using a conventional analog-to-digital (A/D) con-
surement (36, 37).
verter and microcomputer. A third method makes use of a
NOTE A18.2—When using the RDCEPD method, appropriate consid- digital voltmeter with a digital output capability. The advantage
eration should be given to specific features, such as settling time after of this type of system is that all the sensitive analog circuits are
current reversal and averaging of multiple voltage readings. The specific
contained within a single instrument.
conditions to be chosen should account for the characteristic of the
equipment used, see also refs. (36, 37).
A18.6 Gripping Considerations
A18.4 Current Generating Equipment A18.6.1 The DCEPD method for crack size determination
A18.4.1 Any suitable constant current supply may be used relies on a current of constant magnitude passing through the
which has sufficient short- and long-term stability. The current specimen when the voltage is measured. During such potential
shall be stable to one part in 103. For optimum conditions, the measurements, it is essential that no portion of the applied

FIG. A18.1 Schematic Diagram of a DCEPD System with Reference Specimen

50
 E1820 − 23
current be shunted in a parallel circuit through the test probe wires to a more permanent connection into the data
machine. Shunting of current through the test machine, the test acquisition system is recommended. Calibration relationships
fixtures or the clip gage will cause under-prediction in the used to relate DCEPD measurements and crack size can be
crack size measurements. If current shunting is suspected, it is sensitive to the location of voltage wires on the test specimen.
recommended that the specimen be electrically insulated from For a 25-mm thick C(T) specimen, attachment within 60.5
the test fixture. Leaving one contact point between the speci- mm of the locations for which the calibration relationship was
men and the test machine may be required to interrupt current developed will provide adequate accuracy in crack size mea-
shunting while maintaining a ground connection to the surement.
specimen, in order to avoid a floating ground. A18.7.4.1 When testing C(T) specimens of materials with a
high toughness-to-strength ratio, such as austenitic stainless
NOTE A18.3—One common practice to ensure that there are no current
shunting paths through the test cell is to measure the voltage across the steels, large plastic strain can occur in the small section of
crack before and after the specimen is mounted in the fixture. The two material between the pin hole and the notch, which can
measurements will be equal if there are no appreciable current paths other introduce significant apparent crack extension measurements.
than through the test specimen. Alternative current paths are more likely To mitigate this, it is recommended to attach the voltage wires
while force is applied to the test specimen as this can improve the electric
contact between the specimen and the test frame.
as shown in Fig. A18.2 (38).
NOTE A18.4—To electrically insulate the specimen from the test fixture,
a thin layer of insulating tape (such as ultra-high molecular weight plastic)
A18.8 Resolution of Electric Potential Systems
may be placed at the specimen contact points. This tape strain hardens, A18.8.1 The effective resolution of DCEPD measurements
and after an initial loading, has only a small effect on system compliance depends on a number of factors including voltmeter resolution,
and may be reused several times before replacement.
amplifier gain, current magnitude, specimen geometry, voltage
A18.7 Wire Selection and Attachment measurement and current input wire locations, and electric
resistivity of the specimen material. Herein, effective resolu-
A18.7.1 Careful selection and attachment of current input
tion is defined as the smallest change in crack size which can
and voltage measurement wires can avoid many problems
be distinguished in actual test operation, not simply the best
associated with the electric potential difference method. This is
resolution of the recording equipment. For common laboratory
particularly important in aggressive test environments such as
specimens, a direct current in the range of 5 A to 50 A and
elevated temperature where the strength, melting point, and
voltage resolution of about 60.1 µV or 60.1 % of the initial
oxidation resistance of the wires must be taken into account.
voltage will yield a resolution in crack size better than 0.1 % of
A18.7.2 Depending on the specimen size and configuration, the specimen width. For highly conductive materials (that is,
the voltage measurement wires and other optional measure- aluminum, copper) or lower current levels, or both, the
ment transducers can be mounted either before or after placing resolution would decrease, while for materials with a higher
the specimen in the test machine. This is mostly a matter of resistivity (that is, titanium, nickel) resolutions better than 0.01
operator convenience (see Note A18.3). % of the specimen width have been achieved.
A18.7.3 Current Input Wires—Selection of the current input NOTE A18.5—The following is an example of the magnitude of
wires shall be based on current carrying ability, and ease of voltages as measured at 20 °C on a non-standard C(T) specimen (thickness
attachment (weldability, connector compatibility). Wires shall B = 7.7 mm, width W = 50 mm, a/W = 0.22) for a direct current of 10 A
be of sufficient gage to carry the required current under the test (39):
conditions and may be mechanically fastened or welded to the Approximate DCEPD Approximate Change in
Measured at 10 A Crack Size for 1 µV
specimen. Symmetric locations relative to the crack plane and Change
crack centerline are required. Material in DCEPD
Aluminum 0.1 mV 300 µm
A18.7.4 Voltage Measurement Wires—Voltage wires shall Steel 0.6 mV 50 µm
be as fine as possible to allow precise location on the specimen Titanium 3.5 mV 9 µm
and minimize stress on the wire during testing which could A18.9 Techniques to Reduce Voltage Measurement Scatter
cause detachment. For readily weldable materials, attaching
probe wires using resistance spot welding works well. This A18.9.1 Because of the low level of voltages which must be
attachment method allows the leads to be accurately located, measured with the DCEPD method, a number of procedures
uses little specimen space, and is relatively durable. For should be followed to improve voltage measurement precision.
materials of low weldability (for example, certain aluminum A18.9.2 Induced EMF—Voltage measurement lead wires
alloys), lead wires may be fastened using mechanical fasteners should be as short as possible and should be twisted together to
provided that the size of the fastener is accounted for when reduce stray voltages induced by changing magnetic fields.
determining the location of voltage sensing leads. Voltage Holding them rigid also helps reduce the stray voltages which
sensing wires shall be located diagonally across the starter can be generated by moving the wires through any static
notch or crack tip in order to average measurements of magnetic fields that may exist near the test frame. In addition,
non-uniform crack fronts. The probe wires are typically of very routing the voltage measurement leads away from the motors,
small diameter (for example, 30 gauge or 0.25 mm diameter) transformers, or other devices which produce strong magnetic
so a terminal strip near the specimen for transition from the fields is recommended.

51
 E1820 − 23

FIG. A18.2 Recommended voltage wire locations for materials with high toughness-to-strength ratio

A18.9.3 Electric Grounding—In order to mitigate the intro- NOTE A18.7—The input current duty cycle is the ratio between the time
duction of ground-loops, all devices used in the electric the current is on and the total duration of the test, expressed in percentage.
potential difference system, including test controllers, data- A18.9.6 Selection of Input Current Pulse Duration—When
acquisition devices, power supplies, amplifiers, and relays, etc., using pulsed current, the duration of the input current pulse
shall be placed at a common electric ground. should be short enough to minimize specimen heating, and
A18.9.4 Thermal Effects—For DC systems, thermoelectric long enough to allow current stabilization (depending on the
voltage measurement and correction are critically important. A characteristics of the DC current source) and stable DCEPD
minimum number of connections should be used and main- measurements.
tained at a constant temperature to minimize thermoelectric A18.9.7 Selection of Input Current Magnitude—The choice
effects (see A18.3.5). of current magnitude is an important parameter: too low a
A18.9.4.1 All measuring devices (amplifiers/preamplifiers, value may not produce measurable output voltages; too high a
voltmeters, analog-to-digital converters) and the specimen value may cause excessive specimen heating or arcing (40).
itself shall be maintained at a constant temperature. Enclosures A18.9.7.1 To minimize these problems, current densities
to ensure constant temperatures throughout the test are gener- should be kept to the minimum value which can be used to
ally beneficial. produce the required crack size resolution. The maximum
NOTE A18.6—If there are concerns about signal fluctuations and/or current that can be used with a particular specimen can be
drifts due to thermoelectric effects, it is helpful to monitor the stability of determined by monitoring the specimen temperature while
the DCEPD signals while the specimen is at the test temperature but increasing the current in steps, allowing sufficient time for the
before actually starting the test. If the stability is not deemed adequate, the
use of a reference potential (see A18.3.3) or a tighter control of the test
specimen to thermally stabilize. Particular care should be
temperature, or both will be beneficial. exercised when testing in vacuum, as convection currents are
not available to help maintain the specimen at the test
A18.9.5 Selection of Input Current Duty Cycle—The choice
temperature.
of input current duty cycle (see Note A18.7) shall be the result
of a compromise. A 100 % duty cycle corresponds to a constant A18.9.8 DC Current Stabilization Period—Allow a suffi-
applied current throughout the test. This option allows for cient stabilization period after turning the DC current either on
continuous monitoring of the DCEPD signal, however it may or off before making a voltage measurement. Most solid-state
require a reduction of the input current magnitude and an power sources can stabilize the output current within a period
associated drop in sensitivity in order to avoid significant of 1 or 2 s for a step change in output. However, this should be
specimen heating. A low-duty current cycle will minimize verified for each particular specimen and experimental con-
specimen heating, allowing for higher input current magnitude figuration. Measurements of DCEPD shall be taken once the
and for correction of thermal effects. current is stabilized.

52
 E1820 − 23
A18.10 Effects of Plasticity on DCEPD CMOD (for SE(B) specimens) or LLD (for C(T) specimens) as
A18.10.1 Plastic deformation of the specimen can increase a function of DCEPD, as shown in Fig. A18.3. The value of
measured electric potential differences due to changes in DCEPD corresponding to a change in slope corresponds to V0.
material resistivity and specimen geometry (other than crack On the force versus displacement record of the test, the point
extension). It is therefore necessary to distinguish between corresponding to the change in slope shall fall between the
changes in DCEPD due to plasticity and changes due to crack elastic limit (end of the initial linear portion of the curve) and
extension. the maximum recorded force.
A18.10.2 Two methods to account for plasticity-induced NOTE A18.8—The specific procedure used to identify the point corre-
DCEPD changes are described in A18.11 for the determination sponding to the change in slope is left to the user. Different methods,
of crack size from DC electric potential differences. Both ranging from purely visual to analytical approaches, have been used (43).
methods require establishing a value of DCEPD labelled V0. NOTE A18.9—For some high-toughness materials, it might be difficult
Depending on the method used, V0 is either a fixed value that to unequivocally identify the point corresponding to the initiation of crack
corresponds to the DCEPD at the initiation of ductile crack extension, particularly when the point of slope change is not clearly
extension (reference method, A18.11.1), or a force-dependent detectable (37). In these instances, it may be useful to run some tests
interrupted before and just after the onset of crack extension to validate the
value that corresponds to the start of the test plus any DCEPD determination of V0 for a given material, preferably before the actual
increase associated with the elastic deformation of the speci- scheduled tests and after the confirmatory specimen mentioned in
men (alternative method, A18.11.2). A18.16.3.
A18.10.3 The analytical or experimental calibrations used A18.11.2 Alternative Method: Determination of V0 from the
for calculating crack size (A18.4-A18.5) do not account for the Force versus DCEPD Record—Construct a plot of force as a
effects of plasticity on the measured DCEPD. Within the function of DCEPD, as shown in Fig. A18.4. Using force as the
requirements of this annex, it is assumed that most, if not all, dependent variable and DCEPD as the independent variable,
of the significant plasticity in the specimen occurs prior to the linearly fit the data corresponding to the steeply rising part of
initiation of crack extension. For the reference method in the record using the equation V0 = CF + D, where F is applied
A18.11.1, the change in DCEPD measured prior to the onset of force and C, D are fitting coefficients. Another key value of
crack extension shall be ignored, and the remainder of the
DCEPD, Vel, corresponds to the upper deviation from linearity
DCEPD change shall be used to establish the ductile crack
in the force versus DCEPD curve (Fig. A18.4).
extension portion of the J-R curve. This method involves the
analysis of a plot of crack-mouth opening displacement or NOTE A18.10—Experience has shown that the alternative method in
load-line displacement as a function of DCEPD (41, 42). The A18.11.2 tends to yield unconservative results when applied to high-
alternative method in A18.11.2, which has been shown to work toughness, high-ductility steels such as austenitic steels, where significant
satisfactorily for several steels (43), uses a plot of force as a deformation occurs prior to crack initiation (37). For such materials, the
use of the reference method (A18.11.1) is recommended.
function of DCEPD (44).
NOTE A18.11—The non-linear portion of the curve shown in Fig. A18.4
A18.11 Methods for determining V0 at the very beginning of the test, which is often observed, corresponds to
the rapid increase of DCEPD after the specimen is loaded, caused by the
A18.11.1 Reference Method: Determination of V0 from the separation of the fatigue precrack surfaces. This portion should not be
CMOD/LLD versus DCEPD Record: Construct a plot of included in the linear fit.

FIG. A18.3 Example of DCEPD versus CMOD or LLD Record for a Structural Steel

53
 E1820 − 23

FIG. A18.4 Example of Force versus DCEPD Test Record for a Structural Steel

A18.12 Consideration of Crack-Tip Blunting in the Calcu- size a0. The approximate blunting-corrected initial crack size
lation of Crack Size for V = Vel, a0,bl, is given by:
A18.12.1 Depending on the method used for the establish- J Vel
a 0,bl 5 a 0 1 (A18.5)
ment of V0, different approaches shall be used to account for 2σ Y
crack-tip blunting before the onset of ductile crack extension.
with JVel = J-integral corresponding to Vel and a0. For V >
A18.12.2 When using the reference method (A18.11.1), the Vel, ai shall be calculated from DCEPD using a specific ana-
crack size before the onset of ductile crack extension (V ≤ V0) lytical or experimental relationship, for example one of those
shall be calculated as: described in A18.14 or A18.15.
Ji A18.13 Charateristics of DCEPD versus Crack Size Rela-
a i 5 a 01 (A18.1)
2σ Y tionships
where Ji is calculated using a0 in accordance with Annex A18.13.1 Some closed-form solutions for the relationship
A1 for SE(B) specimens or Annex A2 for C(T) specimens, between DCEPD and crack size are described in A18.14 and
respectively, and σY is the effective yield strength at the test A18.15 for C(T) and SE(B) specimens, respectively. These
temperature. The blunting-corrected initial crack size at the relationships are sensitive to the location of current and voltage
initiation of ductile tearing (V = V0), a0,bl, is given by: wires, but they are not sensitive to test temperature, as long as
J V0 temperature variations are accounted for (see Note A18.6). The
a 0,bl 5 a 0 1 (A18.2)
2σ Y relationships are also insensitive to specimen material and
dimensions, provided that all dimensions vary in a proportional
where JV0 is the J-integral value corresponding to V0 and manner.
a0. For V > V0, ai shall be calculated from DCEPD using a
specific analytical or experimental relationship, for example A18.13.2 When testing alternative geometries or with dif-
one of those described in A18.14 or A18.15. ferent current/voltage wire locations than those shown in
A18.12.3 When using the alternative method (A18.11.2), A18.14 and A18.15, the relationship between DCEPD and
the crack size corresponding to V ≤ Vel shall be approximated crack size may be derived empirically or by numerical mod-
by: elling (finite elements). Any such relationship shall be experi-
mentally verified using post-test fracture surface measurements
J el,i
a i 5 a 01 (A18.3) (that is, by comparing measured and predicted final crack sizes
2σ Y or extensions). It must be emphasized that a specific relation-
where Jel,i is calculated as: ship between crack size and DCEPD is only valid for a specific
combination of current and voltage wire locations.
K i2 ~ 1 2 v 2 ! NOTE A18.12—The placement of voltage wires is usually a compromise
J el,i 5 (A18.4)
E between good sensitivity to crack size changes and freedom from errors
caused by minor variations in lead location from specimen to specimen.
and Ki is calculated according to A1.4.1 for SE(B) speci- Lead locations near the crack tip yield better sensitivity to changes in
mens or A2.4.1 for C(T) specimens, using the original crack crack size or initiation of crack extension. The difficulty with this type of

54
 E1820 − 23
arrangement is that the electric field is, in general, highly non-uniform in initial crack size, calculated by means of Eq A18.2 for the
the near-tip region. Thus, minor variations in lead placement from one reference method or Eq A18.5 for the alternative method.
specimen to the next may produce significant differences in measured
voltage for the same crack size (45). In most cases, those positions which A18.13.6 For calibration functions which are derived nu-
give greatest sensitivity to crack size changes also have the greatest merically or empirically, for example Eq A18.9, Eq A18.10,
sensitivity to variations in lead wire positioning.
NOTE A18.13—Current input wire locations also represent a compro-
and Eq A18.13, the value of ā is usually predefined, so the
mise between uniformity and sensitivity. Placement of the current input corresponding value of V̄ can be calculated from V0 and a0,bl,
wires near the crack tip region focuses the current streamlines there, using, for example, Eq A18.10, Eq A18.12 and Eq A18.14.
resulting in increased sensitivity to crack initiation. Placement of the
current leads remote from the crack plane tends to provide a more uniform A18.13.7 The equations listed in the following sections
current field for the measurement of crack extension (see, for example, (A18.4 for C(T) specimens and A18.5 for SE(B) specimens)
Figs. A18.5-A18.9). have been derived under DC conditions for sharp cracks in the
A18.13.3 A closed-form analytical solution for the relation- respective specimen geometries using either closed-form or
ship between voltage difference and crack size was established experimental calibrations.
by Johnson (46) for the M(T), middle-cracked tension, speci- A18.13.8 Regardless of the relationship between DCEPD
men. It has been shown that this relationship may also be and crack size used, the use of a reference probe is recom-
applied to a wide range of specimen geometries, including mended (see A18.3.3). When employing such a reference
C(T) and SE(B) specimens (47). probe, the DCEPD measurements made for crack size deter-
A18.13.4 General Form of the Relationships— mination shall be divided by the ratio Vref/Vref,0,
Relationships between crack size and DCEPD are generally of
where:
the form:

SD
Vref = the reference probe voltage measured at the same
a V time as the DCEPD test specimen probe voltage,
5f , (A18.6)
W V̄ and
or alternatively Vref,0 = the initial reference probe voltage.
a 2 ā
ā
5f S D
V 2 V̄
V̄
, (A18.7) A18.14 Examples of Relationships between DCEPD and
Crack Size for C(T) Specimens
where: A18.14.1 A closed-form analytical expression (46) that can
V = DCEPD signal, be used for the C(T) geometry is the following:

S D
3 46 4
V̄ = DCEPD signal corresponding to a specific crack size ā, πy
a = crack size, and cosh
a 2 2W

S D
W = specimen width. 5 cos21

5 3
W π πy
cosh
A18.13.5 In some analytically derived calibration functions, V 2W

S D
cosh cosh21
for example, Eq A18.8, it is possible to select convenient V0 πa 0,bl
cos
values of V̄ and ā such that, V̄ = V0 and ā = a0,bl, where V0 is 2W
obtained using the reference method (A18.11.1) or the alter- (A18.8)
native method (A18.11.2), and a0,bl is the blunting-corrected

FIG. A18.5 Possible C(T) Specimen configuration for Eq. A18.8, with 2y = W/3.

55
 E1820 − 23

FIG. A18.6 Schematic of C(T) Specimen DC Current and Potential Lead Connections for Eq. A18.9

FIG. A18.7 Alternative C(T) Specimen DC Potential Lead Connections for Eq. A18.11.

FIG. A18.8 Schematic of SE(B) Specimen DC Potential Lead Connections for Eq. A18.8

where y is the half-distance between the voltage measure- ments points. A possible configuration for a C(T) specimen

56
 E1820 − 23

FIG. A18.9 Alternative SE(B) Specimen DC Potential Lead Connections for Eq. A18.13

with 2y = W/3 is shown in Fig. A18.5. where V̄ corresponds to a ≈ 0.5W, and is calculated from V0
A18.14.2 Another closed-form analytical relationship for using the following relationship:
the C(T) specimen, developed from finite element analyses and
V0

S D
verified through experimental techniques (48), is the following: V̄ 5 (A18.12)
a 0,bl 2.8523
a
W
5 B 0 1B 1
V
V̄
SD SD SD
1B 2
V
V̄
2
1B 3
V
V̄
3
(A18.9)
3.4916
W
10.5

NOTE A18.15—When the reference method (A18.11.1) is used, Eq

A18.12 yields a unique value of V̄ for the whole test. Conversely when
a the alternative method (A18.11.2) is used, since V0 is force-dependent, Eq
for 0.24 # # 0.7
W A18.12 provides a different value of V̄ for each data point.
where: A18.15 Examples of Relationships between DCEPD and
V = DCEPD signal, Crack Size for SE(B) Specimens
V̄ = DCEPD signal corresponding to ā = 0.241W
B0 = -0.5051, A18.15.1 The same closed-form expression provided in
B1 = 0.8857, A18.14.1, Eq A18.8, has been found to apply to SE(B)
B2 = -0.1398, and specimens when the current input leads and voltage wires are
B3 = 0.0002398. at the locations shown in Fig. A18.8.
V0 A18.15.2 An empirical relationship for the SE(B) specimen
S D S D S D
V̄ 5 2 3 (A18.10)
a 0,bl a 0,bl a 0,bl was developed based on data from (49), which involved
A 0 1A 1 1A 2 1A 3
W W W current input leads and voltage wires located as shown in Fig.
where: A18.9. The following power-law relationship was obtained
A0 = 0.5766 over the range 0.45 ≤ a/W ≤ 0.8:
A1 =
A2 =
A3 =
1.9169,
-1.0712, and
1.6898.
a
w F
5 0.4512 S V
V̄
2 0.5 DG 0.4688
(A18.13)

NOTE A18.14—When the reference method (A18.11.1) is used, Eq

where V̄ corresponds to ā ≈ 0.5W , and is calculated from
A18.10 yields a unique value of V̄ for the whole test. Conversely, when V0 using the following relationship:
the alternative method (A18.11.2) is used, since V0 is force-dependent, Eq
A18.10 provides a different value of V̄ for each data point. V0

S D
V̄ 5 2,1331 (A18.14)
a 0,bl
A18.14.3 An empirical relationship for the C(T) specimen 2.2163 10.5
W
was developed based on data from (49), which involved
current inputs at the W/4 position, as shown in Fig. A18.7. The
following power-law relationship was obtained in the range
NOTE A18.16—When the reference method (A18.11.1) is used, Eq
0.45 ≤ a/W ≤ 0.8: A18.14 yields a unique value of V̄ for the whole test. Conversely, when the
a
w F
5 0.2864 S V
V̄
2 0.5 DG 0.3506
(A18.11)
alternative method (A18.11.2) is used, since V0 is force-dependent, Eq
A18.14 provides a different value of V̄ for each data point.

57
 E1820 − 23
A18.16 Validity Requirements A18.17.2 Each predicted crack size for V > V0 (reference
A18.16.1 The requirements of 9.1.4, 9.1.5.1, A9.9, A9.10, method, A18.12.2) or V > Vel (alternative method, A18.12.3)
A11.9, and A11.10 shall be fulfilled. shall be adjusted using:
A18.16.2 The final predicted crack extension, ∆ap predicted, a p 2a 0,bl
a i,adj 5 ~ a 2 a 0,bl ! 1a 0,bl ,
a p, predicted 2 a 0,bl i
(A18.15)
and the measured physical crack extension, ∆ap, shall not differ
by more than 0.25 ∆ap for ∆ap < 0.3 bo, and by more than 0.075
bo thereafter. where ai is the unadjusted crack size.
A18.16.3 At least one additional, confirmatory specimen A18.18 Report
shall be tested under the same test conditions as the remaining
A18.18.1 Section 10 describes the reporting requirements
specimens. The value of DCEPD corresponding to a crack
for this method. If the DCEPD method is used, the following
extension of 0.5 mm shall be estimated from the analytical or
additional items shall be reported.
experimental relationship used. The additional, confirmatory
specimen shall then be tested to this DCEPD value, marked as A18.18.2 Details of the analytical or experimental relation-
prescribed in 8.5.1, broken open and the ductile crack exten- ship used to infer crack size from DCEPD measurements.
sion shall be measured. The measured crack extension shall be
0.5 6 0.25 mm in order for the subsequent tests to be qualified A18.18.3 If the reference method (A18.11.1) has been used,
according to this method. DCEPD versus CMOD or LLD record for each test, indicating
NOTE A18.17—The confirmatory specimen should be tested before the the point of slope change used to determine V0. Additionally,
scheduled tests, as it can provide an indication of V0 (constant or an explanation on how this point was established (visual
force-dependent), which can be used for a preliminary analysis of the examination, use of a slope determination algorithm, etc.) shall
following tests. be included.
A18.17 Crack Size Adjustment A18.18.4 If the alternative method (A18.11.2) has been
A18.17.1 After checking that all validity requirements in used, force versus DCEPD record for each test, showing the
A18.16 are fulfilled, predicted crack sizes shall be adjusted linear fit established in the initial portion of the test.
based on the measured original crack size, a0, and the mea- Additionally, an explanation on how the linear fit was estab-
sured physical crack extension, ∆ap, in order to improve the lished (visual examination, use of a slope determination
overall accuracy of the predictions. algorithm, etc.) shall be included.

APPENDIXES

(Nonmandatory Information)

X1. FITTING OF EQUATION A9.1

X1.1 To fit Eq A9.1 to the Ji, ai data using the method of X1.2 This equation can be set up and solved using a
least squares, the following equation must be set up and solved standard spreadsheet or using a mathematical analysis program
for aoq, B, and C: like MathCad, Maple, or Mathematica.

58
 E1820 − 23

X2. GUIDELINES FOR MEASURING THE FRACTURE TOUGHNESS OF MATERIALS WITH SHALLOW CRACKS

X2.1 Significance and Use crack extension here defined as occurring at ∆ap = 0.2 mm
X2.1.1 Fracture toughness measurements may be made (0.008 in) + JIcSC()/2σY .
using specimens with relatively shallow cracks, 0.05 < a/W < X2.2.1.5 JcSC() [FL-1]—The property Jc determined by this
0.45, which are not permitted by the standard test method. The test method characterizes the fracture toughness of materials at
resulting measures of fracture toughness, designated JxSC() and fracture instability prior to the onset of significant stable
δxSC(), will be dependent upon the size of the specimen and the tearing crack extension, ∆ap < 0.2 mm (0.008 in) + JcSC()/2σY,
crack length. The fracture toughness determined from speci- in a specimen with a shallow crack.
mens with shallow cracks is usually non-conservative when X2.2.1.6 JuSC() [FL-1] The quantity Ju determined by this
compared to the fracture toughness determined from standard, test method measures fracture instability after the onset of
deep crack specimen geometries and may exhibit considerably significant stable tearing crack extension, ∆ap ≥ 0.2 mm (0.008
more scatter, particularly in the ductile to brittle transition in) + JuSC()/2σY, in a specimen with a shallow crack.
region for ferritic materials. The J resistance curves determined
according to this appendix are not corrected for crack growth X2.3 Specimen Size and Configuration
and will be non-conservative relative to crack growth corrected X2.3.1 Recommended Specimen—The recommended speci-
resistance curves. men is a single-edge notch, bend specimen SE(B), similar to
X2.1.2 This appendix is provided to give recommended that shown in Fig. A1.1.
procedures for conducting fracture toughness tests of speci- X2.3.1.1 The initial crack size to specimen width shall be
mens containing shallow cracks. Special requirements for the 0.05 ≤ a/W ≤ 0.45 and the specimen width to thickness shall be
instrumentation, specimen, testing procedure and data analysis in the range 1 ≤ W/B ≤ 4.
are described. X2.3.1.2 The narrow notch configuration of Fig. 6 is rec-
ommended; however, the notch opening at the front face of the
X2.1.3 Particular care must be exercised when using these specimen may need to be modified from the dimensions shown
non-standard measures of fracture toughness for structural in Fig. 7, particularly for specimens with a/W < 0.2. The notch
integrity assessments. The user is cautioned that differences opening should be made as small as practical to minimize the
may exist between laboratory test and field conditions and that influence of the machined notch on the elastic compliance of
the fracture toughness of a shallow crack specimen may be the specimen and the fracture response of the specimen.
strongly influenced by the size of the crack and the specimen Notches produced using the wire electrical discharge machin-
geometry. ing process with a wire diameter less than 0.25 mm (0.010 in.)
usually produce satisfactory results.
X2.2 Terminology X2.3.1.3 An alternative method for producing a shallow
X2.2.1 All of the following parameters describe various crack specimen involves machining an SE(B) specimen with
measures of fracture toughness determined using specimens an over-sized W dimension. A fatigue crack is extended from
containing shallow cracks. These parameters are similar to the the starter notch and then the specimen is remachined to
corresponding parameters for standard specimens except that remove the starter notch leaving a specimen with only a fatigue
they include the subscript SC(a0/W) to indicate a shallow crack crack. Integral knife edges may subsequently be machined into
specimen. The number in the parentheses is the original crack the specimen.
size to specimen width ratio for the shallow crack specimen. X2.3.1.4 Integral knife edges or features for the seating of
X2.2.1.1 δIcSC() [L]— is a value of CTOD near the onset of the crack mouth opening displacement gage may be particu-
slow stable crack extension in a specimen with a shallow crack, larly advantageous for specimens containing shallow cracks.
here defined as occurring at ∆ap = 0.2 mm (0.008 in) + 0.7δIc. Suggested configurations are shown in Fig. X2.1. The square
X2.2.1.2 δSC() [L]—is the value of CTOD at the onset of notch configuration along with the ring-type crack mouth
unstable crack extension (see 3.2.36) or pop-in (see 3.2.22) in opening displacement gage of Fig. X2.2 is well-suited to small
a specimen with a shallow crack when ∆ap < 0.2 mm (0.008 in) specimens and shallow cracks. The integral knife edges shown
+ 0.7δcSC(). δcSC() corresponds to the force Pc and clip-gage in Fig. X2.1 may not be suitable for very small cracks, a
displacement vc (see Fig. 1). < 2 mm (0.079 in.).
X2.2.1.3 δuSC()[L]—is the value of CTOD at the onset of
unstable crack extension (see 3.2.36) or pop-in (see 3.2.22) in X2.4 Apparatus
a specimen with a shallow crack when the event is preceded by X2.4.1 Apparatus is required for the measurement of the
∆ap≥0.2 mm (0.008 in) + 0.7δuSC(). δuSC() corresponds to the applied force and the crack mouth opening displacement.
force Pu and the clip gage displacement vu (see Fig. 1). It may Standard force transducers as described in 6.3 are satisfactory.
be size dependent and a function of test specimen geometry. It Load-line displacement measurements are not required for the
can be useful to define limits on ductile fracture behavior. shallow crack SE(B) specimens.
X2.2.1.4 JIcSC() [FL-1] - The property JIc determined by this X2.4.1.1 Crack Mouth Opening Displacement Gages—The
test method characterizes the fracture toughness of materials in standard gage described in 6.2.2 may be suitable for measuring
a specimen with a shallow crack near the onset of stable tearing the CMOD on specimens which have a notch opening that is

59
 E1820 − 23

FIG. X2.1 Recommended Notch Configurations with Integral Features for Mounting Crack Mouth Opening Displacement Gages. These
Notch Configurations Are Only Recommended for a $ 2mm (0.079 in.)

All dimensions are in mm.

FIG. X2.2 Alternative crack opening displacement gage suitable for use with very narrow Notches.

60
 E1820 − 23
large enough to accommodate the arms of the gage. For small K ~i!2 ~ 1 2 ν 2!
specimens and for specimens with very shallow cracks, an J5 1J pl (X2.2)
E
alternative gage design such as the ring gage in Fig. X2.2 is where K(i) is from A1.4.1 and
recommended. Alternative means for measuring CMOD may η CMOD~ i21 ! @ A CMODpl~ i ! 2 A CMODpl~ i21 ! #
be required for specimens with very shallow cracks, a < 2 mm J pl~ i ! 5 J pl~ i21 ! 1 (X2.3)
B N b ~ i21 !
(0.079 in.).
X2.4.1.2 The bend-test fixture recommended in 6.5.1 is In Eq X2.3, the quantity ACMODpl(i) - ACMODpl(i-1) is the
suitable for testing SE(B) specimens with shallow cracks. increment of area under the force versus plastic component of
CMOD record between lines of constant displacement at i-1
X2.5 Specimen Preparation and i as shown in Fig. A1.3. The quantity ACMODpl(i) can be
evaluated from the following equation:
X2.5.1 The requirements of Section 7 are generally appli-
cable with the following notable exceptions. ~ P i 1P i21 !
A CMODpl~ i ! 5 A CMODpl~ i21 ! 1
2 ~ v pl~ i ! 2 v pl~ i21 ! ! (X2.4)
X2.5.1.1 Crack Starter Notch Configuration—Only the
straight through notch configuration of Fig. 5 is recommended where:
for the shallow crack SE(B) specimen. vpl(i) = plastic part of the CMOD = v(i)- PiCi and
X2.5.1.2 Fatigue crack length—The crack size, total length Ci = slope (∆vm/∆P)i of the current unload/reload cycle.
of the starter notch plus the fatigue crack, shall be between In Eq X2.3, ηCMOD is a function of crack size and is given
0.05W and 0.45W. by the following expression:

SD SD
X2.5.1.3 Fatigue Loading Requirements—In order to pro- 2
mote early fatigue crack initiation it is recommended that the a a
η CMOD ~ i21 ! 5 3.667 2 2.199 10.437 (X2.5)
specimen be statically preloaded in such a way that the notch w w
tip is compressed in a direction normal to the intended crack X2.7.2.3 For a resistance curve test method using an elastic
plane (not to exceed a force equal in magnitude to 2Pf). The compliance technique with crack mouth opening displacement
fatigue crack shall be grown a minimum of 1.5× the size of the measured at the notched edge of a shallow crack specimen with
plastic zone resulting from the compression preload, rp, where: 0.05 ≤ a/W ≤ 0.45, the crack length is given as follows (50):

rp 5 S D
1 K
3π σ ys
2
(X2.1)
a
W
5 1.01878 2 4.5367u19.0101u 2 2 27.333u 3 174.4u 4 2 71.489u 5
with K evaluated using the expression in A1.4.1 and the (X2.6)
maximum compressive force used to preload the notch.
where:
X2.6 Procedure u = 1
X2.6.1 The requirements of Section 8 for the SE(B) speci-
men are generally applicable for conducting the tests. The
F B e WECi
S/4 G 1/2
11

resistance curve procedure is recommended. It may be neces-

Ci = (∆vm/∆P) on an unloading/reloading sequence, and
sary to use unload/reload cycles near the maximum recom-
Be = = B - (B-BN)2/B.
mended range of either 50 % of Pf or 50 % of the current force,
whichever is smaller, in order to get accurate estimates of the X2.7.3 Calculation of CTOD:
specimen compliance. X2.7.3.1 For the shallow crack SE(B) specimen, calcula-
tions of CTOD for any point on the force-displacement record
X2.6.2 The user is cautioned that specimens with shallow are made using A1.4.5 except that J values shall not be crack
cracks can store greater amounts of elastic energy than the growth corrected when using equations Eq A1.15 or Eq A1.17.
standard deeply cracked specimen. If the specimen fails in an
unstable manner, the broken halves of the specimen may be X2.8 Analysis of Results
forcefully ejected from the testing machine and suitable
restraints should be fashioned. X2.8.1 The data shall meet the following requirements to be
qualified according to this method. If the data do not pass these
X2.7 Calculation requirements no fracture toughness measures can be deter-
mined according to this method.
X2.7.1 Calculation of K—The stress intensity factor, K, is X2.8.1.1 All requirements on the test equipment in 6 or as
calculated using the expression in A1.4.1. modified in X2.4 shall be met.
X2.7.2 Calculation of J: X2.8.1.2 All requirements on machining tolerance and pre-
X2.7.2.1 J Calculation for the Basic Method— J is calcu- cracking in Section 7 or as modified in X2.5 shall be met.
lated according to A1.4.2 except that the crack growth correc- X2.8.1.3 All requirements on fixture alignment, test rate,
tion of Annex A16 shall not be employed because it is not and temperature stability and accuracy in 8 shall be met.
applicable to shallow cracks. X2.8.1.4 The crack size requirements in 9.1.4 and 9.1.5
X2.7.2.2 J Calculations for the Resistance Curve Method— shall be met for shallow crack fracture toughness tests.
At a point corresponding tov ~ i ! , P ~ i ! on the force versus crack- X2.8.2 Fracture Toughness Calculation—The reported frac-
mouth opening displacement record, calculate J as follows: ture toughness values for shallow crack specimens shall

61
 E1820 − 23
include the subscript SC(a0/W) where a0/W is replaced by the significant stable tearing, proceed to Annex A6 and Annex A7
original crack size to specimen width ratio. When the test to evaluate toughness values and then go to Annex A8 and
terminates with fracture instability, evaluate whether the frac- Annex A10 to develop R-curves. Proceed to Annex A9 and
ture occurred before significant stable tearing or after signifi- Annex A11 to develop initiation values of toughness.
cant stable tearing. The beginning of significant stable tearing X2.8.2.1 The maximum crack extension capacity for a
is defined in A6.3 and A7.3. For fracture instability, proceed to specimen in A8.3.2 and A10.3.2 is limited to ∆amax = 0.1b0 for
Annex A6 and Annex A7 to evaluate the toughness values in a shallow crack specimen.
terms of J or CTOD. For fracture instability occurring after

X3. RECOMMENDED BEST PRACTICES FOR ESTIMATING CRACK SIZE USING ELASTIC COMPLIANCE

X3.1 This test method does not prescribe specific practices defined as the point at which the crack opening displacement
for performing elastic compliance estimates of crack size. This first decreases. If the specimen is held at constant displacement
appendix provides recommendations for analyzing the compli- to allow stress relaxation per 8.6.4, the displacement decrease
ance data from unload/reload sequences and provides a method would occur after completing the hold period. The force at the
for quantifying the uncertainty in JIc due to uncertainty in the start of the unload is P1. The end of an unload is defined as the
compliance estimate. point at which crack opening displacement first increases again
to start reloading. The force at the end of the unload is P2 and
X3.2 It is not uncommon for there to be some non-linearity
in the test record at the start and end of an unload; therefore, ∆P=(P1 – P2). Selection of data for linear regression of the
the following recommendations are made to exclude data near unload should start with the first point that falls below
the displacement reversals at the start and end of an unload P1-0.05∆P (point a in Fig. X3.1) and stop with the last point
from the compliance calculation. The start of an unload is before the force first falls below P1 – 0.95∆P (point b in Fig.

FIG. X3.1 Unload-reload sequence for compliance measurements (offset exaggerated for clarity)

62
 E1820 − 23
X3.1). Selection of data for regression of the reload should start ance and can be estimated from the non-dimensionalized
when the force exceeds P1 – 0.95∆P (point c in Fig. X3.1), and root-mean-square (rms) of the standard error of the compliance
stop with the last point before the force exceeds P1-0.05∆P for all (∆a, J) data pairs used in the power law fit of section
(point d in Fig. X3.1). After the data first crosses a limit to start A9.6.5. The non-dimensionalized root-mean-square of the
an unload or reload, all subsequent data is included in the standard error of the compliance is calculated as follows:
regression analysis until the data first crosses the other limit to X3.5.1 Calculate the root-mean-square of the standard error
end the unload or reload. of the compliances (from Eq X3.1) corresponding to the n (∆a,
X3.2.1 Nonlinearity in the test record can result from stress J) points selected for the power law fit in A9.6.5.

Œ
relaxation and ductile tearing among other possible causes. It
Σ i51
n
~ s β 1 ! i2
may be necessary to exclude data from a larger region than the e5 (X3.2)
upper and lower 5% of the load range as specified in X3.2. A n
plot of the residuals from the regression analysis may be useful X3.5.2 Non-dimensionalize the rms standard error, e, from
to determine an appropriate range of linear data to include in X3.5.1 by multiplying it by the product of the elastic modulus
the compliance estimate. of the material, the specimen effective thickness, and the
specimen width (EBeW), then dividing by 0.0254 mm (0.001
X3.3 Conduct linear regression on the data selected for the
in).
unload and reload separately. The independent variable (x) is
taken as the force and the dependent variable (y) is displace- EB e W
ẽ 5 e (X3.3)
ment so that the slope from the regression analysis is the 0.0254 mm ~0.001 in.!
compliance. It is recommended that crack sizes be calculated X3.5.3 Analysis of data from a round robin has shown that
using just the unload compliances for each unload/reload the uncertainty in JIc due to noise in the unload/reload data is
sequence. As an alternative it is permissible to calculate crack less than 4% of JIc when the non-dimensionalized rms standard
sizes using the reload compliances or to average the unload and error of the compliances, ẽ , for all points used in the power law
reload compliances and calculate crack sizes from the average. fit is less than 400 (51). It is recommended that the non-
NOTE X3.1—Averaging the unload and reload compliances may give a dimensionalized rms standard error of the compliance corre-
different result than performing regression on all of the unload and reload sponding to the unload/reload points used in the power law fit
data as a single data set. This distinction is important when quantifying the
uncertainty in JIc. not exceed 400. If this value is exceeded, the data acquisition
system should be checked to make sure enough data is being
X3.4 For each linear regression, also calculate the standard collected during an unload (reload), digital resolution is
error of the compliance using equation Eq X3.1. This is a sufficient, and noise is not excessive. See Guide E1942 for
measure of the uncertainty in the slope measurement and can guidance on evaluating a data acquisition system. See also the
be used to estimate uncertainty in JIc. transducer requirements in 6.2.3 and 6.3.2. The limits used for
data selection in X3.2 should also be checked to make sure data

!
1
Σ @ y i 2 ~ β̂ 0 1 β̂ 1 x i ! #
2
is being selected from the linear region of the unload (reload).
n22
S β1 5 (X3.1) The load train alignment, movement of pins in clevis holes, and
Σ ~ x i 2 x̄ ! 2
seating of clip gage on knife edges should also be checked for
where: these are possible sources of nonlinearity in the unload
sβ1 = standard error of the compliance (reload). It is possible that reloads may have consistently lower
β0 = estimate of intercept standard errors, or vice versa, so examining the standard errors
β̂1 = estimate of compliance for both will help determine what data to use for crack size
x̄ = mean of x-values determination in order to reduce uncertainty in the resulting JIc.
xi = force values NOTE X3.2—The standard error of the compliance can be decreased by
yi = displacement values increasing the number of points acquired during an unload (reload).
n = number of points used in compliance calculation
X3.6 Report the non-dimensionalized rms standard error of
If the average compliance is used to calculate the crack size,
all points used in the power-law fit of A9.6.5; whether unload,
calculate the root-mean square of the standard errors of the
reload or the average of unload and reload was used for crack
unload and the reload compliances to obtain a standard error of
size estimates; and the average number of points in the
the average compliance.
regression analyses of the unloadings to determine the com-
X3.5 Uncertainty in JIc is related to uncertainty in compli- pliances.

63
 E1820 − 23
REFERENCES

(1) Rice, J. R., Journal of Applied Mechanics, Vol. 35, 1968, p. 379. (19) Nakamura, T., Shih, C. F., and Freund, L. B., “Analysis of a
(2) Joyce, J. A., and Tregoning, R. L., “Development of Consistent Size Dynamically Loaded Three-Point-Bend Ductile Fracture Specimen,”
Criteria for ASTM Combined Fracture Mechanics Standards,” Fa- Engineering Fracture Mechanics, Vol 25, No. 3, 1986, pp. 323–339.
tigue and Fracture Mechanics, Vol 30, ASTM STP 1360, P. C. Paris, (20) Joyce, J. A., and Hackett, E. M., “An Advanced Procedure for J-R
et al, eds., ASTM, West Conshohocken, PA, 1999. Curve Testing Using a Drop Tower,” Nonlinear Fracture Mechanics:
(3) Joyce, J. A., “Manual on Elastic-Plastic Fracture: Laboratory Test Vol 1–Time Dependent Fracture, ASTM STP 995, A. Saxena, J. D.
Procedures,” ASTM Manual Series MNL 27, 1996. Landes, and J. L. Bassani, Eds., ASTM, Philadelphia, 1989, pp.
(4) Dawes, M. G.,“ Elastic-Plastic Fracture Toughness Based on the COD 298–317.
and J-Integral Concepts,” Elastic-Plastic Fracture, ASTM STP 668, J. (21) Irwin, G. R., " Crack-Toughness Testing of Strain-Rate Sensitive
D. Landes, J. A. Begley, and G. A. Clarke, eds., ASTM, 1979, pp. Materials", Journal of Engineering for Power, Trans. ASME, Oct
307–333. 1964, pp. 444-450.
(5) Willoughby, A. A., and Garwood, S. J., “On the Loading Compliance (22) Shoemaker, A. K., "Factors Influencing the Plane-Strain Crack
Method of Deriving Single Specimen RCurves in Three Point Toughness Values of a Structural Steel," Transactions of the ASME,
Bending,” Elastic-Plastic Fracture: Second Symposium, Volume II— Journal of Basic Engineering, Sep 1969, pp. 506-511.
Fracture Resistance Curves and Engineering Applications, ASTM (23) Joyce, J. A., “Analysis of the E08.02 High Rate Round Robin,”
STP 803, C. F. Shih and J. P. Gudas, eds., ASTM, 1983, pp. Journal of Testing and Evaluation, JTEVA, Vol 29. No. 4. July 2001,
II-372–II-397. pp. 329–351.
(6) Hackett, E. M., and Joyce, J. A., “Dynamic J-R Curve Testing of a (24) Herrera, R., and Landes, J. D., “Direct J-R Curve Analysis: A Guide
High Strength Steel Using Key Curve and Multi-Specimen to the Methodology,” Fracture Mechanics: Twenty-First Symposium,
Techniques,” Fracture Mechanics, Seventeenth Volume, ASTM STP ASTM STP 1074, J.P. Gudas, J.A. Joyce, and E.M. Hackett, Eds.,
905, ASTM, 1986, pp. 741–774. ASTM, Conshohocken, PA, 1990, pp 24–43.
(7) Hellman, D., Rohwerder, G., and Schwalbe, K. H.,“ Development of (25) Landes, J. D., Zhou, Z., Lee, K., and Herrera, R., “Normalization
a Test Setup for Measuring Deflections of Single Edge Notched Bend Method for Developing J-R Curves with the LMN Function,”
(SENB) Specimens,” Journal of Testing and Evaluation, Vol 12 , No. Journal of Testing and Evaluation, JTEVA, Vol 19, No. 4, July 1991,
1, January 1984, pp. 62–64. pp. 305–311.
(8) KarisAllen, K. J. and Mathews, J. R., “The Determination of Single (26) Lee, Kang, “Elastic-Plastic Fracture Toughness Determination Un-
Edge-Notched Bend Specimen Load-Line Displacement from Re- der Some Difficult Conditions,” Ph.D Dissertation, The University of
motely Located Sensors in Elastic-Plastic Fracture Testing,” Journal Tennessee, Knoxville, August, 1995.
of Testing and Evaluation, JTEVA, Vol 22, No. 6, Nov. 1994, pp. (27) Linares, A.E., et al., “Using Automated J-R Curve Analysis Software
581-583. To Simplify Testing And Save Time,” Advanced Materials &
(9) Begley, J. A., Clarke, G. A., and Landes, J. D., “Results of an ASTM Processes, Feb/Mar 2019, Vol. 177, No.2, p27.
Cooperative Test Program on the JIc Determination of HY-130 Steel,” (28) Lucon, E., “On the Effectiveness of the Dynamic Force Adjustment
Journal of Testing and Evaluation, Vol 8, No. 5, September 1980. for Reducing the Scatter of Instrumented Charpy Results,” Journal
(10) Gudas, J. P., and Davis, D. A., “Evaluation of the Tentative JI-R of ASTM International, Volume 6, Issue 1 (January 2009).
Curve Testing Procedure by Round Robin Tests of HY-130 Steel,” (29) Manahan, M. P., and Stonesifer, R. B., "The Difference Between
Journal of Testing and Evaluation, Vol 10, No. 6, November 1982, Total Absorbed Energy Measured Using an Instrumented Striker and
pp. 252–262. that Obtained Using an Optical Encoder," Pendulum Impact Testing:
(11) Wallin, K., and Laukkanen, A., “Improved Crack Growth Correc- A Century of Progress, ASTM STP 1380, T. A. Siewert, ed., ASTM,
tions for J-R Curve Testing,” Engineering Fracture Mechanics, 71, 2000, pp. 181-197.
2004 , pp. 1601–1614. (30) Ireland, D. R., " Procedures and Problems Associated with Reliable
(12) Pisarski, H. G., “Measurement of Heat Affected Zone Fracture Control of the Instrumented Impact Test," Instrumented Impact
Toughness,” Paper TS31, Steel in Marine Structures, C. Noordhoek Testing, ASTM STP 563, T.S.,DeSisto, ed., ASTM, 1974, pp.3-29.
and J. de Back, eds., Proceedings of 3rd International ECSC (31) Böhme, W., " Experience with Instrumented Charpy Tests obtained
Conference, Delft, June 15-18, 1987, Elsevier Science Publishers by a DVM Round Robin and Further Developments," ESIS 20, E.
B.V., Amsterdam, 1987, pp. 647–656. van Walle, ed., MEP Publications, London, 1996, pp. 1-23.
(13) Srawley, J. E., “Wide Range Stress Intensity Factor Expressions for (32) Lucon, E., “ The Use of the Normalization Data Reduction Tech-
ASTM E399 Standard Fracture Toughness Specimens,” Interna- nique for Measuring Upper Shelf Toughness Under Impact Loading
tional Journal of Fracture, Vol 12, June 1976, pp. 475–476. Rates”,” Journal of ASTM International, Vol. 8, No. 10, (pending
(14) Newman, J. C., Jr., “Stress Analysis of the Compact Specimen release) November 2011.
Including the Effects of Pin Loading,” ASTM STP 560, 1974, pp. (33) Lucon, E., “Estimating dynamic ultimate tensile strength from
105–121. instrumented Charpy data,” Materials and Design, Volume 97, May
(15) Newman, J. C., Jr., “Stress Intensity Factors and Crack Opening 2016, pp. 437-443.
Displacements for Round Compact Specimens,” International Jour- (34) Hartman, G. A., and Johnson, D. A., “D-C Electric Potential Method
nal of Fracture, 17(6), December 1981, pp. 567–578. Applied to Thermal/Mechanical Fatigue Crack Growth,” Experimen-
(16) Underwood, J. H., Kapp, J. A., and Baratta, F. I., “More on tal Mechanics, March 1987, pp. 106–112.
Compliance of the Three-Point Bend Specimen,” International (35) Pollock, D. D., “Thermoelectricity, Theory, Thermometry, Tool,”
Journal of Fracture, Vol 28, 1985, pp. R41–R45. ASTM STP 852, ASTM, 1985.
(17) Saxena, A., and Hudak, S. J., “Review and Extension of Compliance (36) Catlin, W. R., Lord, D. C., Prater, T. A., and Coffin, L. F., “The
Information for Common Crack Growth Specimens,” International Reversing D-C Electrical Potential Method,” Automated Test Meth-
Journal of Fracture, Vol 14, October 1978, pp. 453–468. ods for Fracture and Fatigue Crack Growth, ASTM STP 877,
(18) Yoon, K. K., Gross, L. B., Wade, C. S., and VanDerSluys, W. A., Cullen, W. H., Landgraf, R. W., Kaisand, L. R., and Underwood, J.
“Evaluation of Disk-Shaped Specimen for Determining J-R Curves,” H., Eds., ASTM, 1985, pp. 67–85.
Fracture Mechanics: 26th Volume, ASTM STP 1256, W. G. Reuter, J. (37) Bicego, V., Liviero, D., Fossati, C., and Lucon, E., “J-R curve testing
H. Underwood, and J. C. Newman, eds., American Society for utilizing the reversing direct current electrical potential drop
Testing and Materials, Philadelphia, 1995. method,” Application of Automation Technology to Fatigue and

64
 E1820 − 23
Fracture Testing, ASTM STP 1092, Braun A.A., Ashbaugh, E., and initiation of growth using the DC potential drop and the partial
Smith F.M., Eds., American Society for Testing and Materials, 1990, unloading methods,” ASTM STP 856, 1985, pp. 338-362
pp. 143-166. (45) Aronson, G. H., and Ritchie, R. O., “Optimization of the Electrical
(38) Tarnowski, K. M., Davies, C. M., Dean, D. W., and Nibkin, K. M., Potential Technique for Crack Monitoring in Compact Test Pieces
“The Influence of Plasticity on Crack Length Measurements Using Using Finite Element Analysis,” Journal of Testing and Evaluation,
the Potential Drop Technique,” Evaluation of Existing and New JTEVA, Vol. 7, No. 4, July 1979, pp. 208–215.
Sensor Technologies for Fatigue, Fracture and Mechanical Testing, (46) Johnson, H. H., “Calibrating the Electric Potential Method for
ASTM STP 1584, Kang J., Jablonski, D., and Dudzinski, D., Eds., Studying Slow Crack Growth,” Materials Research and Standards,
American Society for Testing and Materials, 2015, pp. 73-96. Vol. 5, No. 9, Sept. 1965, pp. 442–445.
(39) Donald, J. K., and Ruschau, J., “Direct Current Potential Difference (47) Tarnowski, K. M., Nibkin, K. M., Dean, D. W., and Davies, C. M.,
Fatigue Crack Measurement Techniques,” Fatigue Crack Measure- “A Unified Potential Drop Calibration Function for Common Crack
ment: Techniques and Applications, K. J. Marsh, R. A. Smith, and R. Growth Specimens,” Experimental Mechanics, published on line 25
O. Ritchie, Eds., EMAS, Warley UK, 1991, pp. 11-37. May 2018, https://doi.org/10.1007/s11340-018-0398-z.
(40) Dover, W. D., et al., “A.C. Field Measurement—Theory and (48) Hicks, M. A., and Pickard, A. C., “A Comparison of Theoretical and
Practice,” from The Measurement of Crack Length and Shape Experimental Methods of Calibrating the Electrical Potential Drop
During Fracture and Fatigue, Beevers, C. J., Ed., EMAS, Cradley Technique for Crack Length Determination,” Int. Journal of
Heath, UK, 1980, pp. 222-260. Fracture, No. 20, 1982, pp. 91–101.
(41) Wilkowski, G. M., Wambaugh, J. O., and Prabhat, K., “Single (49) Hackett, E. M., Kirk, M. T., and Hays, R. A., “An Evaluation of J-R
Specimen J-R Curve Evaluations Using the DCEP Method and a Curve Testing of Nuclear Piping Materials Using the Direct Current
Computerized Data Acquisition System,” Fracture Mechanics: Fif- Potential Drop Technique,” NUREG/CR-4540, U.S. Nuclear Regu-
teenth Symposium, R. J. Sanford, Ed., ASTM, 1984, pp. 553-576. latory Commission, August 1986.
(42) Marschall, C. W., Held, P. R., Landow, M. P., and Mincer, P. N., “Use (50) Joyce, J.A., " J Resistance Curve Testing of Short Crack Bend
of the Direct-Current Electric Potential Method to Monitor Large Specimens Using Unloading Compliance," Fracture Mechanics,
Amounts of Crack Growth in Highly Ductile Metals,” Fracture Twenty-Second Symposium, Vol. 1, ASTM STP 1131, American
Mechanics: Twenty-First Symposium, ASTM STP 1074, J. P. Gudas, Society for Testing and Materials, West Conshohocken, PA, 1992,
J. A. Joyce, and E. M. Hackett, eds., ASTM, 1990, pp. 581-593. pp. 904-926.
(43) Lucon, E., “ASTM E08.07.09 Analytical Round-Robin on the Use of (51) Graham, Stephen M., “Uncertainty in Ductile Fracture Initiation
DC Electrical Potential Difference for the Measurement of Crack Toughness (JIc) Resulting From Compliance Measurement,” Appli-
Size in Ductile Fracture Testing,” Materials Performance and cation of Automation Technology in Fatigue and Fracture Testing
Characterization, Vol. 7, No. 2, April 2018. and Analysis, STP 1571, Peter C. McKeighan and Arthur A. Braun,
(44) Schwalbe, K.-H., Hellmann, D., Heerens, J., Knaack, J., and Müller- Eds., pp. 134 – 152, ASTM International, West Conshohocken, PA
Roos, J., “Measurement of stable growth including detection of 2014.

SUMMARY OF CHANGES

Committee E08 has identified the location of selected changes to this standard since the last issue
(E1820 – 22ɛ1) that may impact the use of this standard. (Approved February 1, 2023.)

(1) Section A9.6.6.4 was revised. Section A9.6.6.5, Note A9.2, (2) Section A11.6.6 was revised. Section A11.6.6.5, Note
and Note A9.3 were added. A11.1, and Note A11.2 were added.

Committee E08 has identified the location of selected changes to this standard since the last issue (E1820 – 21)
that may impact the use of this standard. (Approved November 1, 2022.)

(1) Fig. A14.6 was revised.

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.ast-
m.org). Permission rights to photocopy the standard may also be secured from the Copyright Clearance Center, 222 Rosewood Drive,
Danvers, MA 01923, Tel: (978) 646-2600; http://www.copyright.com/

65