Number

2.4.52
Subject
Fracture Toughness of Resin Systems for Base
3000 Lakeside Drive, Suite 309S
Bannockburn, IL 60015-1249
Materials
Date Revision
07/13
IPC-TM-650 Originating Task Group

TEST METHODS MANUAL IPC 3-11

1 Scope This test method establishes a procedure for 3.1.5 Yield Stress The stress at fracture (slope of stress-
characterizing the toughness of the resin system materials strain curve is not required to be zero).
used in making laminates for the fabrication of printed wiring
boards. The single-edge-notch bending (SENB) geometry is 4 Test Samples
used to determine the critical-stress-intensity factor, K1c, and
the energy per unit area of crack surface or critical strain 4.1 Sample Construction The preferred finished sample is
energy release rate, G1c, at fracture initiation. This method a block of pure resin, free of contaminants and fully cured (not
assumes linear elastic behavior of the cracked specimen, so partially cured, not over-cured). Note: DSC may be used to
there are corresponding restrictions on the linearity of the evaluate a received sample’s degree of cure. TGA may be
load-displacement diagram. Use of this test method for used to check for the presence of residual solvents or other
printed wiring board laminate materials or other composites contaminants.
may not yield comparative results.
A heated hydraulic press may be required to prepare the
2 Applicable Documents sample. Attachment A is a method for making compression
molded thermoset neat resin castings. Size and occurrence of
2.1 ASTM Standards voids within the sample should be kept to an absolute
minimum (maximum void dimension 25 µm [0.001 in]; maxi-
D638 Test Method for Tensile Properties of Plastics
mum 5 voids/cc). Specimen block may be ground down to
D4000 Classification Systems for Specifying Plastic Materials the desired dimensions, and a mold shall not be used.
D5045 Standard Test Methods for Plane-Strain Fracture Default specimen dimensions should be 3.50 mm ± 0.05 mm
Toughness and Strain Energy Release Rate of Plastic thick, 12.7 mm wide (in general, the nominal width can be
Materials between 2X to 4X the thickness, but should be consistent)
and 55.88 mm long (length should be 4.4 times the width).
E399 Test Method for Linear-Elastic Plane-Strain Fracture However the absolute minimum thickness is 2.5 times the
Toughness K1c of Metallic Materials
square of the conditional or trial K1c (KQ) divided by the yield
E691 Practice for Conducting an Inter-Laboratory Study to stress (σy) of the material for the temperature and loading rate
Determine the Precision of a Test Method of the test.

The above should ensure that the sample is wide enough to

3 Terminology
ensure plane strain and sufficiently thick to avoid excessive
plasticity in the ligament. If non-linearity in loading still occurs,
3.1 Terms and Definitions (reference ASTM E399)
the width can be increased up to 4 times the thickness of the
3.1.1 Compact Tension Specimen geometry consisting of specimen. Polishing the sample (minimum 600 grit) is recom-
single-edge notched plate loaded in tension. mended to promote yielding in the tensile test, rather than
brittle fracture. Each of the thickness and width dimensions of
3.1.2 Critical Strain Energy Release Rate (G1c) Tough- the specimen should be measured in at least 3 locations to an
ness parameter based on energy required to fracture. accuracy of 0.1% and both dimensions shall be accurate to
within 1% of nominal. The average of these measurements will
3.1.3 Plane-Strain Fracture Toughness (K1c) Toughness be used in the calculations. At least 10 samples of each mate-
parameter indicative of material fracture resistance. rial are recommended for testing, allowing up to 5 samples for
developing sufficient skill in initiating consistent cracks and
3.1.4 Single-Edge Notched Bend Specimen geometry subsequently at least 5 samples meeting Section 4.2 criteria
consisting of center-notched beam. for acceptable fracture toughness measurements.

Material in this Test Methods Manual was voluntarily established by Technical Committees of IPC. This material is advisory only
and its use or adaptation is entirely voluntary. IPC disclaims all liability of any kind as to the use, application, or adaptation of this Page 1 of 8
material. Users are also wholly responsible for protecting themselves against all claims or liabilities for patent infringement.
Equipment referenced is for the convenience of the user and does not imply endorsement by IPC.
 IPC-TM-650
Number Subject Date
2.4.52 Fracture Toughness of Resin Systems for Base Materials 07/13

Revision

4.2 Sample Preparation The required square notch is cut 5.2 Displacement Measurement
in the center of the sample width, within ± 0.25 mm, using a
diamond saw or similar stress-free method to form a smooth 5.2.1 The displacement measurement using an internal
bottom of the notch. The slot width should be 0.25 mm displacement transducer having sufficient precision shall be
[0.00984 in.]. performed using the machines stroke or position transducer.
The fracture test displacement data shall be corrected for
The crack itself is initiated on each sample with a new 0.23
system compliance, loading pin penetration and specimen
mm [0.009 in] thick ultra-sharp carbon steel razor blade compression by performing a calibration of the testing sys-
(example: http://www.mcmaster.com/#3962a4/=3qpeql). It tems as described in ASTM D-5045.
is recommended that the blade be refrigerated or cooled in
liquid nitrogen or in dry ice shortly before use. The razor blade 5.2.2 The displacement measurement using an external dis-
is then carefully tapped using a small weighted hammer with placement transducer having sufficient precision shall be
sufficient force and control for the crack to initiate on the first performed with the transducer located between the top and
or second try. A new, cool or cold razor blade is recom- bottom plates, and as close as possible to the load point on
mended for reducing the force needed for crack initiation. A the specimen to ensure displacement accuracy.
few specimens in every test lot should first be sacrificed for
operator practice at crack initiation, precisely determining the 5.3 Yield Stress
hammer force needed for that sample lot to avoid only mak-
ing indentations. The depth of the natural crack generated by 5.3.1 The yield stress, σy, is determined by the material’s
tapping shall be a least twice the width of the machined maximum load in an uniaxial tensile test. Using a constant
notch (3X the width of the notch is ideal). stroke rate uniaxial tensile test, the loading time to yield shall
be within ± 20 percent of the actual loading time observed in
The total depth of the notch plus the depth of the crack shall
the fracture test. A zero slope to the stress-strain curve is not
be half the thickness of the sample, within ± 5 percent. There-
required. If a tensile test cannot be performed, then use 0.7
fore the depth of the square notch should be 45 percent of
times the compressive yield stress as an approximation.
the sample width minus 0.75 mm, ± 0.13 mm. The crack
shall be sufficiently sharp to ensure that a minimum value
6 Procedure
of toughness is obtained during the subsequent 3-point
bending. The actual depths are measured after fracture within 6.1 Test Preparation The specimens and all testing shall
0.5 % accuracy at three locations; at the center of the crack be performed at 23 °C ± 3 °C. The actual temperature of the
front, and at the end of the crack front on each surface of the specimen shall be recorded. The relative humidity should be
specimen. The average of these three measurements, which between 30 % and 60 % RH, and shall be recorded.
should be fairly uniform, shall be used in the calculations.
Cracks or breaks should be resin-resin, not between resin and 6.2 Displacement Correction Specimen shall be identi-
filler. cal to the specimen prepared for fracture testing, except with-
out the notch or crack in the middle. This specimen shall be
5 Equipment/Apparatus or Material used for single notch bend testing (reference ASTM D5045).

5.1 Test Machine 6.3 Testing

5.1.1 The testing machine used shall be a constant dis- 6.3.1 The notched specimen that has been pre-cracked is
placement rate device; an electromechanical screw-driven subjected to loading at a loading rate of 5.0 mm per minute.
machine, or a closed loop feedback-controlled servo-
hydraulic load frame. The stationary and moving rollers used 6.3.2 The test is performed and the load versus loading
for the 3-point loading (typically two under each end, and one point displacement curve is obtained. In the ideal case, there
on top in the middle of the specimen block opposite the is an abrupt drop of load to zero at the instant of crack growth
crack) shall each be large enough to avoid excessive inden- initiation. If this occurs, then determine the trial K1c or KQ from
tation of the plastic, however the roller diameter should not the maximum load. Typically there will be a noticeable devia-
exceed the overall thickness of the specimen. tion from linearity prior to fracture.

Page 2 of 8
 IPC-TM-650
Number Subject Date
2.4.52 Fracture Toughness of Resin Systems for Base Materials 07/13

Revision

6.3.3 The load corresponding to a 2.5 % apparent incre-

ment of crack extension is established by a specified deviation
from the linear portion of the record. The K1c value is calcu-
lated from this load by equations that have been established
on the basis of elastic stress analysis on other specimens suf-
ficiently large to show linear elastic behavior. B' Pmax
6.3.4 After breaking, the interface should appear smooth B
and glossy. A layered or hazy surface indicates deformation PQ
as a failure mechanism rather than cracking, and invalidates
the test results. After testing, measure the depth of the crack
from the notch depth at three locations along the width of the P
sample. Also measure the depth of the crack from the top of
the sample (at the same three locations). θ'
6.3.5 Determining the load displacement area or G 1c
requires an accurate integration of the load versus loading
point displacement curve, including an accurate displacement θ Notes:
determination using the displacement transducer. • Compliance (C) = tan θ =
inverse slope of line AB
7 Analysis Fracture toughness testing is recommended to
• 1.05 C = tan θ'
be performed at least twice per year.

7.1 Calculation and Interpretation of Results (reference

ASTM D5045, Section 9). A u
2-4-52-7-1.eps

7.1.1 In order to establish that a valid K1C has been deter-

Figure 7-1 Determination of C and PQ
mined, it is first necessary to calculate a conditional result, KQ,
which involves a construction on the test record, and to then
7.1.3 Calculate KQ in accordance with the procedure for
determine whether this result is consistent with the size of the
single edge notch bending in 7.1.4. For this calculation, a
specimen in accordance with 7.1.6. The procedure is given in
value of a, which is the total crack length after both notching
7.1.2 through 7.1.8.
and pre-cracking, but before fracture, is best determined from
7.1.2 Load the specimen and obtain a P (load) versus u the fracture surface after testing. An average value is used,
(displacement) plot (see Figure 7-1). but the difference between the shortest and longest length
should not exceed 10 %. Take care that it is the original crack
Draw a best straight line (AB) to determine the initial compli-
which is being observed, since slow growth can occur prior to
ance, C. C is given by the reciprocal of the slope of line (AB).
catastrophic fast fracture.
Draw a second line (AB’) with a compliance 5 % greater than
that of line (AB). If the maximum load that the specimen was 7.1.4 (Reference ASTM D5045, Section A1.4). The general
able to sustain, Pmax, falls within lines (AB) and (AB’), use formula for KQ calculation of bend specimens is given in [Ref.
Pmax to calculate KQ. If Pmax falls outside line (AB) and line 3]. The general principles of the bend-test fixture are illustrated
(AB’), then use the intersection of line (AB’) and the load curve
in Figure 7.2.
as PQ. Furthermore, if Pmax/PQ <1.1, use PQ in the calcula-
tion of KQ. However, if Pmax/PQ >1.1, the test is invalid.

Page 3 of 8
 IPC-TM-650
Number Subject Date
2.4.52 Fracture Toughness of Resin Systems for Base Materials 07/13

Revision

W B
a

0.6D D

S = 4W
Support Rollers

Displacement Transducer

W/2 < D < W

Busses For Rubber Bands (see ASTM E399)

B
2-4-52-7-2.eps

Figure 7-2 Bending Rig with Transducer for Single-Edge-Notch Bending (SENB)

This fixture is designed to minimize frictional effects by allow- where (0 < x < 1):
ing the support rollers to rotate and move apart slightly as the ƒ(x) = 6x1/2 (1.99 - x(1-x)(2.15 - 3.93x + 2.7x2)) / (1 + 2x)
specimen is loaded, thus permitting rolling contact. Thus, the (1 - x)3/2
support rollers are allowed limited motion along the plane sur-
faces parallel to the notched side of the specimen, but are ini- and:
tially positively positioned against stops that set the span PQ = load as determined in 7.1.2, kN,
length at 50 mm, and are held in place by low-tension springs B = specimen thickness, cm,
(such as rubber bands). For the bend specimen, the displace-
W = specimen depth (width), cm,
ments will be essentially independent of the gauge length up
to a gauge length of W/2. For bend specimens with S/W = 4, a = crack length, cm
KQ in units of MPa Ë m1/2 is as follows: and
KQ = (PQ/BW 1/2
) ƒ(x) x = a/W.

Page 4 of 8
 IPC-TM-650
Number Subject Date
2.4.52 Fracture Toughness of Resin Systems for Base Materials 07/13

Revision

Tabulated values of ƒ(x) are given in Table 7-1. + 16[8.9 - 33.717x + 79.616x2 -112.952x3 + 84.815x4 -
25.672x5] {[2x(1 - x) + 2x2]/(1 - x)3}
Table 7-1 Calibration Factors SENBA S/W = 4 Values of Φ are given in Table 7-1.
a/W ƒ (x) Φ ψ ηe
0.450 9.14 0.274 45.8 2.00 7.1.6 (Reference ASTM D5045, Section 9.1.3) Check the
0.455 9.27 0.272 46.7 2.00 validity of KQ via the size criteria. Calculate 2.5 (KQ/σy)2 where
0.460 9.41 0.269 47.6 2.01 σy is the yield stress. If this quantity is less than the specimen
thickness, B, the crack length, a, and the ligament (W - a),
0.465 9.55 0.266 48.5 2.01
then KQ is equal to K1c. Otherwise the test is not a valid K1c
0.470 9.70 0.263 49.5 2.02
test.
0.475 9.85 0.260 50.4 2.02
0.480 10.00 0.257 51.4 2.03 NOTE: Use of a specimen with too small a thickness, B, will
result in KQ being higher than the true K1c value while a small
0.485 10.16 0.254 52.5 2.03
(W - a) will result in a KQ value that is lower than the true K1c
0.490 10.32 0.252 53.5 2.03
value. The net effect may be close to the correct K1c but
0.495 10.48 0.249 54.7 2.03 unfortunately in an unpredictable way, since the dependence
0.500 10.65 0.246 55.8 2.03 on B cannot be quantified.
0.505 10.82 0.243 57.0 2.03
0.510 10.99 0.241 58.2 2.04 7.1.7 For the recommended specimen dimensions of W =
0.515 11.17 0.238 59.4 2.04 2B and a/W = 0.5, all the relationships of 7.1.6 are satisfied
simultaneously. In fact, the criterion covers two limitations in
0.520 11.36 0.236 60.7 2.04
that B must be sufficient to ensure plane strain, but (W - a) has
0.525 11.54 0.233 62.1 2.04
to be sufficient to avoid excessive plasticity in the ligament. If
0.530 11.74 0.230 63.5 2.04 (W - a) is too small the test will often violate the linearity crite-
0.535 11.94 0.228 64.9 2.04 ria. If the linearity criterion is violated, a possible option is to
0.540 12.14 0.225 66.4 2.04 increase W for the same a/W and S/W ratios. Values of W/B
0.545 12.35 0.223 67.9 2.04 of up to 4 are permitted.
0.550 12.56 0.220 69.5 2.05
A
Values calculated using A. Bakker, Compatibility Compliance and Stress
7.1.8 If the test result fails to meet the requirements in either
Intensity Expressions for the Standard Three-Point Bend Specimens. Paper 7.1.2 or 7.1.6, or both, it will be necessary to use a larger
submitted for publication in International Journal of Fatigue and Fracture of specimen to determine KQ. The dimensions of the larger
Engineering Materials and Structures (March 1989).
specimen can be estimated on the basis of KQ, but generally
must be increased to 1.5 times those of the specimen that
7.1.5 For the bend specimens calculate GQ [=] kJ/m2 from
failed to produce a valid K1c value.
the corrected energy, U, as follows:

GQ = U/(BWΦ ) or GQ = ηe U/(B(W - a)) 7.2 Displacement Correction for Calculation of GQ (Ref-

erence ASTM D5045, Section 9.2)
Values of ηe are given in Table 7-1. The energy calibration
factor, Φ, is defined as: Make a displacement correction for system compliance,
loading-pin penetration, and specimen compression, then cal-
Φ = C/(dC/d(A/W))
culate G1C from the energy derived from integration of the
and shall be computed from the following: load versus load-point displacement curve.
Φ = (A + 18.64)/(dA/dx)
where: 7.2.1 The procedure for obtaining the corrected displace-
ment, uc (P), at load P from the measured displacement, uQ
A = [16x2/(1 - x)2][8.9 - 33.717x + 79.616x2 - 112.952x3 +
(P), is as follows: Use an un-cracked displacement correction
84.815x4 - 25.672x5],
specimen prepared from the same material as that being
and: tested. Using the same testing parameters as the actual test,
dA/dx = [16x2/(1 - x)2][-33.717 + 159.232x - 338.856x2 + load the specimen to a point at or above the fracture loads
339.26x3 - 128.36x4] observed during actual testing. From the load-displacement

Page 5 of 8
 IPC-TM-650
Number Subject Date
2.4.52 Fracture Toughness of Resin Systems for Base Materials 07/13

Revision

curve, determine ui (P). The corrected displacement is then

calculated using uc (P) = uQ (P) - ui (P). PQ or Pmax

7.2.2 In practice, it is common to obtain a linear displace-

ment correction curve (up to the fracture loads observed dur-
ing actual testing). This simplifies the displacement correction
to be applied to the fracture test. Initial non-linearity due to
penetration of the loading pins into the applied specimen P
should occur during both the calibration test and the actual
Ui
fracture test. Linearization of the near-zero correction data
and the fracture test data can compensate for this initial non-
linearity.
tan-1 C
7.2.3 The displacement correction must be performed for
each material and at each test temperature or rate. Polymers
are generally temperature- and rate-sensitive and the degree u uQ
of loading-pin penetration and sample compression can vary 2-4-52-7-3a.eps

with changes in these variables. Figure 7-3 (a) Method of Correcting for Indentation;
Load - Deflection in Fracture Test
7.2.4 The indentation tests should be performed in such a
way that the loading times are the same as the fracture tests.
Since the indentations are stiffer, this will involve lower rates to
reach the same loads. PQ or Pmax

7.3 Calculation of GQ (Reference ASTM D5045, Section

9.3) In principle, G1C can be obtained from the following:

G1C = (1 - v2) K1C 2

/ E [Ref. 2] P
but for plastics, E must be obtained at the same time and
temperature conditions as the fracture test because of vis- tan-1 C
coelastic effects. Many uncertainties are introduced by this
procedure and it is considered preferable to determine G1C Ui
directly from the energy derived from integration of the load
versus displacement curve up to the same load point as used
for K1C and shown in Figures 7-3 (a and b).

ui u 2-4-52-7-3b.eps
7.3.1 The energy must be corrected for system compliance,
loading-pin penetration, and specimen compression. This is Figure 7-3 (b) Method of Correcting for Indentation;
done by correcting the measured displacement values, as Load - Deflection in Indentation
shown in Figure 7-3 (a and b). Accordingly, if complete linear-
ity is obtained, one form of the integration for energy is as U = 7.3.3 Calculate GQ from U in accordance with the proce-
1/2 PQ (uQ - ui), where PQ is defined in 7.1.2. dure given in 7.1.5.

7.3.2 Alternatively, it is possible to use the integrated areas 7.3.4 A useful cross check on accuracy may be made using
from the measured curve, UQ , of Figure 7-3, a and indenta- the tensile modulus, E, and Poisson’s ratio, v. E/(1 - v2) shall
tion curves, Ui , of Figure 7-3, b in accordance with 7.3.3 and be calculated from the corrected compliance, Cc, using the
following. following:

U = UQ - Ui [Ref.3, SENB]. (E / (1 - v2)) B Cc = 2ƒ2 Φ = ψ [Ref. 4, SENB]

Page 6 of 8
 IPC-TM-650
Number Subject Date
2.4.52 Fracture Toughness of Resin Systems for Base Materials 07/13

Revision

The factors f, Φ and ψ are given in Table 7-1 and Table 7-2 Table 7-3 Testing Summary
for each geometry. This value of E/(1 - v2) shall be compared Fracture Test Parameters
with that obtained from K1c 2 /G1c. The former value should Testing Laboratory
be the larger, but the difference should be <15 %. The cor- Materials/orientation
rected compliance, Cc, is obtained from the measured com- Specimen geometry
pliance in the fracture test, CQ, and the compliance from the Test temperature, °C
indentation test, Ci, in accordance with the following: Loading rate, m/s
Cc = CQ - Ci [Ref. 5, SENB] Notching method
Specimen number
Table 7-2 Calibration Factors Compact TensionA Width (W), mm
a/W ƒ (x) Φ ψ ηe Crack length from 7.2.2, mm
Pmax, N
0.450 8.34 0.208 28.9 2.64
Pmax loading rate, s
0.455 8.45 0.207 29.6 2.63
PQ loading time, s
0.460 8.57 0.207 30.4 2.61 Stable or unstable growth
0.465 8.70 0.206 31.1 2.60 KQ, MPa - m1/2
0.470 8.83 0.205 31.9 2.58 Uncorrected energy, J
Corrected energy, J
0.475 8.96 0.204 32.7 2.57
GIc, kJ/m2
0.480 9.09 0.203 33.5 2.56
Tensile Test Parameters
0.485 9.23 0.202 34.4 2.54
σy, MPa
0.490 9.36 0.201 35.3 2.53
σy loading time, s
0.495 9.51 0.200 35.3 2.53
Validity Checks
0.500 9.65 0.199 37.1 2.51
Pmax/PQ
0.505 9.81 0.198 38.0 2.50 2.5 (KQ/sy)2
0.510 9.96 0.197 39.0 2.49 E/(1 - ν2) via C, MPa
0.515 10.12 0.196 40.0 2.48 E/(1 - ν2) via KQ2/Gc, MPa
0.520 10.28 0.194 41.1 2.47
0.525 10.45 0.193 42.1 2.46 Table 7-4 Precision Statistics from Round-Robin
Study in Accordance with Practice ASTM E691
0.530 10.62 0.192 43.3 2.45
MaterialA Average Sx Sr SR r R
0.535 10.80 0.190 44.4 2.44
A 4.34 0.652 0.235 0.679 0.658 1.90
0.540 10.98 0.189 45.6 2.43
B 5.70 1.420 0.618 1.510 1.730 4.23
0.545 11.17 0.188 46.8 2.42
C 3.60 0.692 0.343 0.747 0.960 2.09
0.550 11.36 0.186 48.1 2.41
A
D 5.90 1.950 0.944 2.100 2.640 7.39
Values calculated using J. A. Knapp, G. S. Leger and B. Gross, Fracture
A
Mechanics Sixteenth Symposium, ASTM, STP 868, 19, pp. 27 - 44. Material A is values of KIc for nylon. Material B is values of GIc for nylon.
Material C is values of KIc for polycarbonate. Material D is values of GIc for
polycarbonate. Units for all columns are as follows: KIc [=] MPa Ë m1/2 & GIc
7.4 Report List the information required to perform the test [=] kJ/m2.
and the results obtained in the form of a table. The form to
use is provided in Table 7-3.
Each laboratory obtained one test result for each material. The
7.4.1 Table 7-4 is based on a round robin conducted in following explanations of r and R are only intended to present
1988 in accordance with E-691, involving four materials a meaningful way of considering the approximate precision of
tested by nine laboratories. For each material, all the samples this test method. The data in Table 7-4 should not be rigor-
were prepared at one source, but the individual specimens ously applied to acceptance or rejection of material, as those
were prepared at the laboratories which tested them. Each data are specific to the round robin and may not be represen-
test result was the average of three individual determinations. tative of other lots, conditions, materials, or laboratories.

Page 7 of 8
 IPC-TM-650
Number Subject Date
2.4.52 Fracture Toughness of Resin Systems for Base Materials 07/13

Revision

Users of this test method should apply the principles outlined 7.4.4 Keywords (Reference ASTM D 5045, Section 12)
in E-691 to generate the data specific to their laboratory and
• Critical-strain energy release rate
materials, or between specific laboratories. The principles
would then be valid for such data. • Energy-to-break
• Fracture toughness
7.4.2 Concept of r and R (Reference ASTM D5045, Section • Plane-strain fracture toughness
11.2). If Sr and SR have been calculated from a large enough
body of data, and for test results that were averages from References
testing three specimens, the following information applies. (1) Brown, W. F., Jr., and Srawley, J. E., “Plane Strain Crack
Toughness Testing of High Strength Metallic Materials,” ASTM
7.4.2.1 Repeatability, r (comparing two test results for the STP 410, ASTM, 1966, p.1.
same material, obtained by the same operator using the same
(2) “Fracture Toughness Testing and Its Applications,” ASTM
equipment on the same day). The two test results should be
STP 381, ASTM, April 1965, p.30.
judged not equivalent if they differ by more than the r value for
that material. (3) Srawley, J. E., “Wide Range Stress Intensity Factor
Expressions for ASTM E399 Standard Fracture Toughness
7.4.2.2 Reproducibility, R (comparing two test results for the Specimens,” International Journal of Fracture Mechanics, Vol.
same material, obtained by different operators using different 12, June 1976, p.475.
equipment on the same day). The two test results should be (4) Newman, J. C., “Stress Analysis of Compact Specimens
judged not equivalent if they differ by more than the R value Including the Effects of Pin Loading,” ASTM STP 560, ASTM,
for that material. 1974, p.105.

7.4.2.3 Any judgement in accordance with the above would (5) Williams, J. G., “Fracture Mechanics of Polymers,” Ellis
have an approximate 95 % (0.95) probability of being correct. Horwood/Wiley, 1985.

(6) Towers, O. L., “Stress Intensity Factors, Compliances and

7.4.3 Bias There are no recognized standards by which to Elastic ηe Factors for Six Test Geometries,” The Welding Insti-
estimate bias of these test methods. tute, March 1981.

Page 8 of 8