Wieslaw Beres, 1 A s h o k K. Koul, 1 and R.
Thamburaj 2
A Tapered Double-Cantilever-Beam Specimen Designed for
Constant-K Testing at Elevated Temperatures
REFERENCE: Beres, W., Koul, A. K., and Thamburaj, R., "A
Tapered Double-Cantilever-Beam Specimen Designed for Con-
stant-K Testing at Elevated Temperatures," Journal of Testing and
Evaluation, JTEVA, Vol. 25, No. 6, Nov. 1997, pp. 536-542.
ABSTRACT: A compact, constant stress intensity factor, side
grooved, tapered double-cantilever-beam (T-DCB) specimen has been
designed for measuring elevated temperature creep and fatigue crack
growth rates. This specimen is much smaller than the standard DCB
specimen generally used for studying corrosion fatigue in airframe
materials and it can be ea.~ily accommodated in standard furnaces
generally used in elevated-temperature fracture mechanics testing. The
specimen possesses a constant-K region of 30 mm. This constant-K
crack length range was first established analytically through detailed
two- and three-dimensional finite-element analyses and the finite-ele-
ment calculations were further verified experimentally through quanti-
tative fractographic analysis using the striation measurement technique.
KEYWORDS: fracture toughness testing, crack growth rate testing,
double-cantilever-beam specimen
Nomenclature
B Thickness, i.e., distance between sides of specimen, mm
Be Effective thickness of specimen, mm
BE Net thickness of specimen, i.e., distance between roots of
side grooves in side grooved specimen, mm
c Crack length, mm
C Specimen elastic compliance, mm/N
e Distance from wedge tip to line load, mm
E Young's modulus, GPa
F Specimen load, N
Hp Smallest half-height of specimen, mm
Hw Largest half-height of specimen, mm
K Stress intensity factor, MPa.m L/2
KI Mode I stress intensity factor, MPa-m ~/2
W Specimen width, i.e., distance from toad plane to back surface
of specimen, mm
~/ Half of V-shaped side groove included angle
v Poisson's ratio
err Yield strength, MPa
Manuscript received 6/20/96; accepted for publication 3/6/97.
JResearch Officer, Institute for Aerospace Research, National Research
Council, Ottawa, Ontario, Canada, K1A 0R6.
2Director, Advanced Materials and Engineering Systems, Orenda Aero-
space Corp., Mississauga, Ontario, Canada.
536
Copyright by ASTM Int'l (all rights reserved); Thu Oct 1 07:19:29 EDT 2009
Downloaded/printed by
EPFL - Ecole Polytechnique Federale de Lausanne pursuant to License Agreement. No further reproductions authorized.
Abbreviations
DCB Double-cantilever beam
DOF Degree of freedom
FE Finite element
SIF Stress intensity factor
T-DCB Tapered double-cantilever beam
2D Two-dimensional
3D Three-dimensional
Laboratory creep and fatigue crack growth rate (CCGRs and
FCGRs) tests are performed using test specimens of different con-
figuration. In most cases, compact tension (CT) specimens are
used for establishing crack propagation rates of long cracks as a
function of specimen loading conditions where crack advancement
produces a continuous increase in the stress intensity factor K. In
certain situations, however, it is desirable that the K remain constant
during testing. For example, in circumstances where different
microstructures are tested in an aggressive environment it is not
possible to isolate the influence of microstructure and environment
under increasing K conditions. Under such circumstances, one is
not sure whether a particular effect on CCGR or FCGR for a given
material arises from combined contributions of microstructure and
environmental species with changing K or whether the effect is
purely microstructural. The other advantages of keeping K constant
during an experiment are that effects of several different environ-
ments or different temperatures in a single environment can be
determined in a single test. These are typical situations where it
is necessary to maintain the K at a constant value. The paper
describes a new "compact" side-grooved constant-K specimen that
was designed and tested to generate creep and fatigue crack growth
rate data on advanced gas turbine engine materials. This specimen
is considerably smaller than the DCB specimens generally used
for studying corrosion fatigue in airframe materials.
Two methods have been employed to keep the K constant during
testing. The first method relies on manipulating the test specimen
geometry and the second method uses the standard CT specimen
for testing but the specimen load is continuously reduced through
a closed loop control system in a manner such that the K remains
constant when the crack advances. The latter method has been
described, among others, in [1] and [2].
In the first method--the specimen design method--the speci-
mens are contoured in the shape of a tapered double-cantilever-
beam (T-DCB) to maintain a constant-K for some range of crack
length. In addition, side grooves are often provided to avoid,crack
deflection and to maintain the crack path along the midplane of
© 1997 by the American Society for Testing and Materials

4Oman
, $~mm
4Omm
L3111~m ,
© Thirteen 2D as well as seventeen 3D finite-element models were
w of the side grooves. Since the specimen configuration and loading
FIG. l--Geometries of T-DCBspecimens used in various investigations.
the specimen. Examples of various T-DCB specimens used for
stress corrosion cracking and crack arrest toughness testing, Fig.
1, have been described in [3-12].
However, most of the specimens shown in Fig. 1 are not suitable
for studying the elevated temperature crack growth behavior of
aerospace materials, such as superalloys, because these specimens
are quite large in size, ranging from 140 mm to 254 mm in length.
Specimens of these sizes are expensive to machine, require large
environmental chambers and furnaces, and also pose difficulties
in maintaining a uniform specimen temperature during testing.
Although a large specimen and hence a larger constant-K crack
length will allow one to generate more data with one specimen,
the specimen will be exposed to elevated temperature for a very
long period of time, which can lead to temperature-induced micro-
structural changes in the material and consequently the crack
growth data will not be produced at constant microstructure.
The newly designed specimen eliminates these difficulties. This
paper summarizes the results of this specimen design and the
verification of the design using analytical, numerical, and experi-
mental techniques.
Specimen Design
Since most fracture mechanics data are generated on specimens
machined from 12.7 mm (0.5 in.) thick or thicker plates, the
specimen thickness was fixed at 12.7 mm to satisfy plane strain
constraint requirements. It was further assumed that the total width
of the new specimen should not exceed 82.5 mm so that it would
fit easily into standard tubular furnaces that are often used in
Copyright by ASTM Int'l (all rights reserved); Thu Oct 1 07:19:29 EDT 2009
Downloaded/printed by
EPFL - Ecole Polytechnique Federale de Lausanne pursuant to License Agreement. No further reproductions authorized.
BERES ET AL. ON T-DCB TESTING 537
mechanized test systems. The new specimen configuration is
shown in Fig. 2, while Fig. 3 illustrates the nomenclature used for
specimen geometry analysis. The specimen width (W) is 79.3 mm,
the largest half-height (Hw) is 38.1 mm (0.48W), the smallest half-
height (lip) is 6.42 mm, the thickness outside the crack-plane (B)
is 12.7 mm, and the thickness in the crack plane (BE) is 5.08 mm.
Therefore, the nondimensional parameters of the specimen are Hp/
e = 0.4, WIHp = 12.352, Wle = 4.94, and the thickness reduction
in the crack plane is equal to 60% (BE/B = 0.4). The specimen
V-notch side grooves have included angles 2~ of 45 ° or 90 °. The
edge distance of a specimen, i.e., a distance between the crack
mouth and the load line, is 12.7 mm (0.16W).
The test material used for manufacturing the sPecimens was Ni-
based Inconel Alloy 718 with Young's modulus (E) of 200 GPa,
Poisson's ratio (v) of 0.29 and yield strength (err) in the range of
I000 to 1150 MPa at room temperature and in the range of 855
to 952 MPa at 650°C, depending on the heat treatment conditions
applied [13].
Finite Element Analysis
To estimate the length of the constant-K region and to investigate
the effect of the side groove geometry on the specimen compliance
and K, two- and three-dimensional (2D and 3D) linear elastic
finite-element calculations of non-grooved and side-grooved T-
DCB specimens were performed.
prepared. These FE models differed in terms of crack lengths,
presence or absence of the side grooves, and the included angle
are symmetric with respect to the crack plane (y = 0) and the
specimen mid-plane (z = 0), 2D models represented only one-
half of the specimen, whereas all 3D models represented one-
fourth of the specimen in the FE analyses.
Eight-node isoparametric plane strain elements were used for
all 2D models, while twenty-node isoparametric solid elements
were used to model the specimen in 3D. The 2D models consisted
of 362 nodes and 105 elements with the number of DOFs ranging
between 190 and 209 depending on the crack length. The meshes
in the crack tip area for 2D models consisted of eight triangular
elements with midnodes shifted to a quarter point position to
model the stress field singularity. All 3D models for non-grooved
specimens were generated by translating the mesh of the 2D mod-
els, originally designed in the x-y plane, in the thickness (z) direc-
tion of the specimen. Generally the 3D models were divided into
five layers of equal thickness of 1.27 mm. To find the distribution
of SIF across the specimen's thickness, two 3D models were specif-
ically divided into ten layers of equal thickness of 0.635 mm. The
3D models for side-grooved specimens were obtained from the
appropriate 3D models for non-grooved specimens by tapering
finite-elements in the three outer layers of elements. Each crack
in the 2D as well as the 3D analysis was modeled as a through-
thickness crack. The 3D meshes in the crack front area consisted
of the 40 wedge elements with midnodes shifted to a quarter point
position to model the stress field singularity. The element sizes
around the crack front for both types of analysis were of the order
of 1.00 mm. Seven different crack lengths, i.e., 18.75; 25.00; 31.25;
37.50; 43.75; 50.00 and 56.25 mm were used for numerical
analysis.
The 3D FE mesh with five layers consisted of 2817 nodes and
525 elements, which generated FE models with the number of
DOFs ranging from 7877 to 7941 depending on the crack length.
 538 JOURNALOF TESTING AND EVALUATION

5J

L__
I T4 i

--~ 12.7~'-

OIM~9ONS ~ [ IN UILUME'}I[RS

FIG. 2--Geometry of the SMPL-IAR T-DCB specimen.

J"1 B
e .L c E
FIG. 3---Geometric parameters of a T-DCB specimen [12,20].
The 3D meshes with ten layers generated FE models with 15 072
DOFs. The specimen pin load (F) was set at 4454 N and its effect
in 2D was modeled using a concentrated load and in 3D using
the line load distributed uniformly through the thickness of the
specimen. Figures 4 and 5 show the FE meshes for the T-DCB
FIG. 4--2D finite-element mesh for the T-DCB spectmen with the 25
mm long crack.
J
specimen, with a 25 mm long crack modeled as a non-grooved
specimen in 2D and as a 45 ° side-grooved specimen in 3D,
respectively.
The NISA II/DISPLAY III FE software package was used to
generate the meshes and to perform the FE calculations. Displace-
ment solutions were calculated using the NISA II module for
all FE models. The SIFs were extracted from the displacement
calculation results using crack opening displacement extrapolation
technique originally proposed in [14] and displacement correlation
technique suggested in [15]. The condition of plane strain was
assumed for all SIF extractions following [16], where it was shown
that the stress state ahead of the crack tip is nearly plane strain,
especially for side-grooved specimens. For 3D calculations, the
SIFs were extracted at all the FE nodes on the crack front, but
the results on the midplane (z = 0) were considered the most
relevant to the specimen design. The elastic compliance, C, calcu-
lated as a ratio of the load line displacement y, (measured at the
specimen midplane (z = 0), to the specimen pin load F, was
calculated for all 2D and 3D models. To establish the maximum
load that can be applied to the specimens, the von Mises stress
contours were also produced from the FE results.
To check the accuracy of the finite-element calculations, one
2D model contained four times more finite elements than the
others: it consisted of 1337 nodes and 420 elements, which resulted
in 2648 DOFs for this model. The displacement results obtained
from this model were compared to the results obtained from corres-
ponding FE model, which contained a smaller number of elements
but represented the same structure, with the same external loading
and the same boundary conditions applied. The maximum relative
difference between these two displacement results did not exceed
0.3%. Therefore, it was assumed that the finite-element models
applied for calculations converge and that the number of finite
elements used for modeling was sufficiently large.

I Results and Discussion

I ! i~ i ! i i i i i i i i ! ! i i ! ! l
FIG. 5--3D finite-element mesh for the T-DCB specimen with the 45 °
side grooves and the 25 mm long crack.
All Mode I K values obtained through FE calculations were
normalized with respect to the specimen pin load (F), the non-
grooved specimen thickness (B), and the square root of the
specimen width (W) to present them in a dimensionless form,
K~BWt/Z/F. Although several other forms of normalization of K

Copyright by ASTM Int'l (all rights reserved); Thu Oct 1 07:19:29 EDT 2009
Downloaded/printed by
EPFL - Ecole Polytechnique Federale de Lausanne pursuant to License Agreement. No further reproductions authorized.
 BERES El" AL. ON T-DCB TESTING 539

40 1 1 t t I i t •
crack length (c/W) for the non-grooved as well as 45 ° and 90 °
side-grooved specimens obtained from FE calculations. It can be
aS ....... : ..... ! ....... : ..... i ....... i ....... ! ...... : . . . . . seen that all three specimen designs have a relatively large constant-
K range. For the specimen dimensions described in the previous
section this range of constant-K is approximately 39 mm.
• Boundary Collocation It is further observed that the K: value for a given crack length
• 2D FE is much larger in the side-grooved specimens than in the non-
iS . . . . . . . . . . . . . . .
grooved specimen. This is not unexpected because an applied load

=o .... : ...................................... i .... / for a given crack length is acting over a shorter crack front in the
side-grooved specimens. When analyzing these data it may be also
concluded that, relative to the non-grooved specimen, the 45 °
side grooving of the T-DCB specimen increases the KI value by
iS ..................................................
approximately 42%. Upon increasing the included angle of side
grooving to 90 ° the KI value, relative to the non-grooved specimen,
10. ...... }............... i ............... i ....... i ....... ;: ...... increases by approximately 49%. This value is 7% higher than the
45 ° groove results and this increase can be attributed to the
decreased stiffness of the specimen due to its lower mass.
On the basis of these FE analyses the following empirical model
'2 is proposed for the calculation of the Mode I SIF (i.e., K0 for
0 , ' ' ,,, , i , i ,,, ,
0.0 0.1 0.2 0.8 0.4 0.5 0.6 0.7 0.8
non-grooved and grooved T-DCB specimens:
c/W
FIG. 6---Comparison of the SIF obtainedfrom the 2D FE calculations F =A do+ I +d2 + (1)
and from the boundary collocation method reported in [12].
where c is the crack length, W is the specimen width and the
nondimensional coefficients are: do = 0.166, d l = -0.253 and
are possible, the method selected is consistent with the method dE = 0.902.
used in [12,17]. This normalization is easy for practical use because The value of the coefficient A depends on the type of the
the measurements of B and W dimensions are more direct than specimen:
the other geometrical data of the specimen, e.g., Hp (Fig. 3).
2.000, for non-grooved specimen,
Figure 6 plots the normalized Kl versus normalized crack length
A = 2.844, for grooved specimen, 2~/ = 45 °,
(C/W) obtained using the boundary collocation method for the
non-grooved T-DCB specimen [12,17], together with our 2D FE 2.974, for grooved specimen, 2",/ = 90 °.
calculation results obtained for non-grooved specimen. The relative Equation (1) is valid for a normalized crack length range of:
difference between these two solutions does not exceed 2% in the
C
entire range of normalized crack length. 0.2 --< ~ --< 0.7 (2)
Figure 7 shows the normalized Ki values versus normalized
while the region of constant Kl in this equation is established as:
40 ' t t 1 i ' i I C
0.2 --< ~ --< 0.5 (3)
/
k
I
• . . . . . . . . . . . . . . .... /: .......
In this constant K region, the nondimensional values for Kl should
- - - B e Pradlotlon i be taken as:
• ~oo,,o 9o 1 : I11:
ao --- = oto.~4s t ! ..... ! 1I/: .......
KiBv/-~ 1"13.895, Smooth specimen,
Non~ved
1// . .
= ~19.515, Grooved specimen, 2 " / = 45 ° (4)
- ..... ,..", : ...... ! ..... i::" . . . . F [20.495, Grooved specimen, 2~/ = 90°

From experiments performed on larger T-DCB specimens, used

20 .... :, .:-:- ..... . .... -' . ........ ..... . --
for stress corrosion cracking investigations, it is known that the
depth and included angle of side grooves of the specimens affect
15 . . . . . . . . . . . . . . . . . . . . the stability of crack propagation. For compact (CT) and three-
point bend specimens ASTM recommends a 45 ° side groove with
10 25% thickness reduction along the crack plane (BN/B = 0.75) as
:
described in ASTM Standard Test Method for J~, a Measure of
s .................... ....... i .................. i .........
Fracture Toughness (E 813-89). The equation for the effective
thickness (B~) for K calculations for either the CT specimen with
25% side grooves (ASTM E 813) or the T-DCB specimen with
, I i , I
side grooving as suggested in [10] is:
0.0 0.1 0.2 0.8 0.4 0.5 0.6 0.7 O.S

e/W
-BN (5)
FIG. 7iNormaJized SIF versus normalized crack length for non- To check the validity of using this ASTM recommended proce-
~.rooved and 45° and 90" side-grooved specimens. dure for grooved CT specimen for grooved T-DCB specimens,

Copyright by ASTM Int'l (all rights reserved); Thu Oct 1 07:19:29 EDT 2009
Downloaded/printed by
EPFL - Ecole Polytechnique Federale de Lausanne pursuant to License Agreement. No further reproductions authorized.
 540 JOURNAL OF TESTING AND EVALUATION

160 ~ ~ ~ ~ '"""i ,
prediction of stress intensity factors based on the non-grooved
! .
specimen results incorporating the thickness modification i

according to Eq 5, drawn in dashed lines, are compared with the 140 ........ Be Prediction ..... : Y : jt / '-
Groove g0 ,' ./- /
FE results in Fig. 7. It can be seen that the prediction overestimates • Groove 4~ / //' / "
KI by approximately 11.5% when compared with the FE results 120 • Non-,rooved / ./// ......

o,°°.o,o.o
for the 45 ° side-grooved specimen and by 6.5% when compared A//"
' " s : "
to the K~ for the 90 ° grooved specimens. It is therefore not recom-
mended to use the standard ASTM procedure for CT specimen
for the calculation of the effective thickness of the side-grooved
compact T-DCB specimens as stated in [10], but application of
Eq 4 is suggested instead.
All three types of specimens investigated show some KI varia-
tions along the crack front, though the KI is more uniform in the
non-grooved specimens. Distributions of Ki along the crack tront
for the non-grooved and the 45 ° side-grooved T-DCB specimens
are shown in Fig. 8. Both K~ curves were normalized with respect
to their values at z = 0, i.e., the specimen midplane. The magnitude 20 ......... .............. !. . . . . . . . .
of Ki for the non-grooved specimen (solid line) is the largest in : !

the center and gradually drops by 26% near the specimen side
surfaces. However, the opposite holds true for the 45 ° grooved 0,0 0,1 0.2 0,3 0.4 0.5 0,5 0.7 0.8
specimen. The Ki is smallest in the middle plane of the specimen
c/W
(z - 0), and increases sharply towards the free edges, that is
FIG. 9--Normalized compliance curves for non-grooved and 45° and
towards the roots of the side grooves. This increase is of the order 90° side-grooved specimens.
of 32%. The same qualitative results for the CT specimen were
also reported in [18,19], although the variations of the KI for their
25% grooved specimen were much smaller. This difference can thickness modified according to the equation recommended by
be attributed to the difference in the depth of side grooves (BNI ASTM E 813 Standard for CT specimens:
B) which is equal to 0.75 for the CT specimen but is as high as
0.4 for the T-DCB specimen discussed here. Be = B - (B - BN)2/B (6)
Calculated specimen compliance values were normalized with
is also shown in this figure (dashed lines). It can be seen that the
respect to the specimen thickness (B) and Young's modulus of the
compliance varies almost linearly with the crack length over the
specimen material (E), to present them in a dimensionless form,
nondimensional crack length range of 0.2 < c / W < 0.55. It is also
ECB. Figure 9 shows the normalized compliance curves versus
evident that when compared to the non-grooved specimen, the 45 °
the normalized crack length for the three types of specimens inves-
side grooving increases the compliance by 21% for a relative crack
tigated. For comparison, prediction of specimen compliance varia-
length of 0.236 and by 13% for a relative crack length of 0.552.
tions based on the non-grooved specimen results but with the
The 90 ° side grooving further increases the compliance for these
two crack lengths by 7%.
The following empirical models are proposed for the calculation
2.0 I l ] 1
I I I ]

i
of T-DCB specimen compliance over a normalized crack length
1.8
range of 0.2 < c / W < 0.55:
• : ! i
1.6 E . C - B = eo + el (7)

1.4 • . . . . . . ...... .......... ! ...... ...... . . . . . il ........

where c is the crack length, and the nondimensional coefficients
e0 and el have the following values which change with the speci-
A
1.2
o
II
men type:
N • s

1.0

Specimen e0 e~
0.8

Non-grooved - 19.25 204.21

0.6 Grooved 23' = 45 ° - 17.64 223.19
- - - e - - - Side Grooved 45 |
Non-grooved Grooved 23' = 90° - 18.69 238.47
0.4 J
0.2 . . . . . . . . : . . . . • ...... '. Finite-element calculations performed on the T-DCB specimens
also revealed the conditions necessary for yielding in the specimen
0.0 .............{ i i i I neck region, which is the most critical region of the specimen. It
-0.5 -0.4 -0.3 -0.2 -0.1 0.0 0.1 0.2 0.3 0.4 0.5 was found that for the 90 ° side-grooved specimen machined out
Z/B of Inconel 718 the maximum pin load of 14 kN causes yielding
FIG. 8--Distribution of SIF across the specimen thickness for the non- in the neck of the specimen if the yield strength crr for Inconel
grooved and 45 ° side-grooved specimens. 718 at room temperature is assumed to be 1000 MPa. At elevated

Copyright by ASTM Int'l (all rights reserved); Thu Oct 1 07:19:29 EDT 2009
Downloaded/printed by
EPFL - Ecole Polytechnique Federale de Lausanne pursuant to License Agreement. No further reproductions authorized.

temperatures, 650°C, the yield strength of Inconel 718 drops to
950 MPa, which decreases the maximum allowable pin load to
13.3 kN.
Experimental Verification
Further verification of the constant K range for the specimen
was carried out experimentally through fractography using the
striation-spacing measurement technique. A type 304 stainless steel
T-DCB specimen with the geometry shown in Fig. 2 was used for
these experiments. The specimen was cycled to fracture at room
temperature, at a AK of 45 MPa.m lz2, over a crack length of
expected constant AK of 30 mm, at a load ratio R = 0.05 and a
frequency of 1.0 Hz. Subsequently, the fracture surface was studied
by scanning electron microscopy and striation spacings were mea-
sured along the length of the fracture surface [20,21]. The results
are shown in Fig. 10, where it is observed that the striation spacings
tend to be nearly constant over the crack length interval of 12.5
to 44.5 mm from the load line of the specimen. It was thus con-
firmed that the designed T-DCB specimen provides a useful con-
stant-K region of nearly 32 mm.
Summary and Conclusions
A new "compact" side-grooved T-DCB specimen has been
designed for elevated temperature fatigue and creep crack growth
rate testing under constant-Kconditions using 2D and 3D elastic FE
models. This specimen is much smaller than the DCB specimens
generally used for studying corrosion fatigue in airframe materials.
The specimen width is set at 79.3 mm, the largest half-height is
38.1 mm, the smallest half-height is 6.42 mm, the thickness outside
the crack plane is 12.7 mm, and the thickness in the crack plane
is 5.08 mm. The specimen V-notch side grooves may have an
.2.2
- 2.0
-J.4 ~
1.2
0.6
._.a • 4103, CINDAS/Purdue University, West Lafayette, IN.
0.6
! I 04
' z; ' 37.
' 5 ' 510 67.
CRACK LENGTH, 3{mm)
FIG. lO---Results of the striation spacing measurements performed on
the T-DCB specimen [20].
Copyright by ASTM Int'l (all rights reserved); Thu Oct 1 07:19:29 EDT 2009
Downloaded/printed by
EPFL - Ecole Polytechnique Federale de Lausanne pursuant to License Agreement. No further reproductions authorized.
BERES ET AL. ON T-DCB TESTING 541
included angle 2~ of 45 ° or 90 °. The specimen has a constant-K
range of 32 mm and this was verified experimentally using the
striation-spacing measurement technique. Empirical models Eqs
(1) and (7) are proposed for calculating the normalized KI and
compliance values for the non-grooved as well as side-grooved
specimens.
To study the effect of the specimen side grooving in detail a
comparison of the K~ and the compliance values computed for the
non-grooved and the 45 ° and 90 ° side-grooved specimens was
carried out. It was found that, contrary to previous workers' sugges-
tions, the correction for the effective thickness for the K~calculation
recommended in ASTM E 813 for grooved CT specimens overesti-
mates the K~ and is not valid for the T-DCB specimens. To take
into account the effect of side grooves, application of Eq 4 is
recommended.
The maximum applied pin load that can create yielding in the
designed T-DCB specimen neck region depends on the testing
temperature and the material being tested. It was established that
for the Inconel 718 the yielding occurred at a load of 14 kN at a
room temperature and 13.3 kN at 650°C.
References
[1] Pulos, G. C. and Knauss, W. G., "Constant Stress Intensity Factors
Through Closed-Loop Control," International Journal of Fracture,
VoL 63, 1991, pp. 101-112.
[2] Zuidema, J. and Mannesse, M., "A Model for Predicting Slant Fatigue
Crack Growth in AI 2024," Engineering Fracture Mechanics, Vol.
34, 1989, pp. 445-456.
[3] Jakubowski, M., "Geometry Factors in Corrosion Fatigue Crack Prop-
agation," Fatigue and Fracture of Engineering Materials and Struc-
tures, Vol. 16, 1993, pp. 495-507.
[4] Ottens, H. H. and Lof, C. J., "Finite Element Calculations of the
Compliance of a Tapered Double Cantilever Beam Specimen for
Different Crack Configurations," National Aerospace Laboratory of
the Netherlands Report, NLR-TR-72083-U, 1972.
[5] Schra, L., Boerema, P. J., and van Leeuwen, H. P., "Experimental
Determination of the Dependence of Compliance on Crack Tip Con-
figuration of a Tapered Double Cantilever Beam Specimen," National
Aerospace Laboratory of the Netherlands Report, NLR-TR-73025
U, 1973.
[6] Radon, J. C., "Corrosion Fatigue of Aluminium Alloy RR58," The
Influence of Environment on Fatigue, I Mech E Conference Publica-
tions 1977-4 Paper No. Cl08/77, 1977, pp. 85-92.
[7] Gallagher, J. P., "Experimentally Determined Stress Intensity Factors
for Several Contoured Double Cantilever Beam Specimens," Engi-
neering Fracture Mechanics, Vol. 3, 1971, pp. 27-43.
[8] Crosley, P. B. and Ripling, E. J., "Significance of Crack Arrest
Toughness (Kla) Testing," ASTM STP-71I, 1980, pp. 321-337.
[9] Marcus, H. L. and Sih, C., "A Crackline-Loaded Edge-Crack Stress
E
Corrosion Specimen," Engineering Fracture Mechanics, Vol. 3, 1971,
::L pp. 453--461.
[10] Mostovoy, S., Crosley, P. B., and Ripling, E. J., "Use of Crack-Line-
Loaded Specimens for Measuring Plane-Strain Fracture Toughness,"
z
Journal of Materials, Vol. 2, 1967, pp. 661-687.
[11] Schijve, J. and de Koening, A. U., "A Specimen for a Constant Stress
Intensity Factor," Engineering Fracture Mechanics, Vol. 9, 1977,
z
o pp. 331-340.
[12] Srawley, J. E. and Gross, B., "Stress Intensity Factors for Crackline-
Loaded Edge-Crack Specimens," Materials Research and Standards,
Vol. 7, 1967, pp. 155-162.
[13] Aerospace Structural Metals Handbook, 1992 edition, Vol. 4, CODE
[14] Chan, S. K., Tuba, I. S., and Wilson, W. K., "On the Finite Element
Method in Linear Fracture Mechanics," Engineering Fracture
Mechanics, Vol. 2, 1970, pp. 1-17.
[15] Shih, C. E, deLorenzi, H. G., and German, M. D., "Crack Extension
Modelling with Singular Quadratic lsoparametric Elements," Interna-
tional Journal of Fracture, Vol. 12, 1976, pp. 647-651.
[16] Newman, Jr., J. C., Booth, B. C., and Shivakumar, K. N., "An Elastic-
Plastic Finite Element Analysis of the ./-Resistance Curve Using a
 542 JOURNALOF TESTING AND EVALUATION
CTOD Criterion," Fracture Mechanics: 18th Symposium, ASTM STP
945, D. T. Reed and D. Reed, Eds., ASTM, Philadelphia, 1988,
pp. 665-685.
[17] Gross, B. and Srawley, J. E., "Stress-Intensity Factors by Boundary
Collocation for Single-Edge-Notch Specimens Subject to Splitting
Forces," NASA Technical Note, TN D-3295, 1966, pp. 1-10.
[18] Shivakumar, K. N. and Newman, Jr., J. C., "Numerical Fracture
Simulation of Bend Specimens Using a CTOD Criterion," Engi-
neering Fracture Mechanics, Vol. 32, 1989, pp. 203-210.
[19] Shivakumar, K. N. and Newman, Jr., J. C., "Verification of Effective
Thickness for Side-Grooved Compact Specimens," Engineering
Fracture Mechanics, Vol. 43, 1992, pp. 269-275.
Copyright by ASTM Int'l (all rights reserved); Thu Oct 1 07:19:29 EDT 2009
Downloaded/printed by
EPFL - Ecole Polytechnique Federale de Lausanne pursuant to License Agreement. No further reproductions authorized.
[20] Thamburaj, R., Koul, A. K., Wallace, W., and de Malherbe, M. C.,
"Effect of Microstructure and Environment on the Elevated Tempera-
ture Creep and Low Cycle Fatigue Crack Growth Rates in Inconel
718," Laboratory Technical Report LTR-ST- 1586, National Research
Council, National Aeronautical Establishment, Ottawa, 1986, pp.
1-138.
[21] Dainty, R. V., "Application of Optical and Electron Microscopic
Techniques in the Fractographic Determination of Fatigue Crack
Growth Rates," Fatigue Crack Topography, AGARD Conference Pro-
ceedings, No. 376, 1984, pp. 15-1/15-26.