You are on page 1of 10

Engineering Fracture Mechanics 76 (2009) 911–920

Contents lists available at ScienceDirect

Engineering Fracture Mechanics


journal homepage: www.elsevier.com/locate/engfracmech

Determination of stretch zone width using fem


Sanjeev Saxena a,*, N. Ramakrishnan a, B.K. Dutta b
a
Advanced Materials and Processes Research Institute, Hoshangabad Road, Habibganj Naka, Bhopal 462 026, India
b
Reactor Safety Division, Bhabha Atomic Research Centre, Mumbai 400 085, India

a r t i c l e i n f o a b s t r a c t

Article history: The study addresses numerical determination of stretch zone width and its critical value.
Received 25 October 2007 The proposed method is based on the highly deformed stretch zone that is defined as
Received in revised form 18 November 2008 stretch zone width, expected to be better than the half of the crack tip opening distance
Accepted 21 December 2008
which inturn determined using 45° line method. The investigation essentially comprises
Available online 1 January 2009
a number of finite element analyses of compact tension test, using tensile test data. The
proposed methodology also provides insight into the mechanism involved in the creation
Keywords:
of stretch zone.
Ductile fracture
Finite element analysis
Ó 2008 Elsevier Ltd. All rights reserved.
Crack initiation (Ji)
Stretch zone width

1. Introduction

In ductile materials, the pre-existing crack first blunts, subsequently at critical strain, voids are formed ahead of the crack
tip that finally coalesce with the tip leading to crack propagation [1]. A fracture criterion that could accurately predict failure
of cracked bodies would be a useful engineering tool both for the evaluation of structural integrity and the selection of mate-
rials. Complex structures may have stresses in some regions that exceed the elastic limit. This has created need for a fracture
criterion that would also include elastic–plastic to fully plastic behaviour. The path independent J integral, as formulated by
Rice [2], can be viewed as parameter, which is an average measure of the crack tip elastic–plastic field. There is an intimate
relationship between the fracture behaviour of ductile materials and the extent of the plasticity that occurs at the tip of the
crack. There is experimental evidence that this crack-tip plasticity manifests itself as a stretch zone ahead of the fatigue pre-
crack, the extent of which has been related to the fracture toughness.
Several methods have been proposed to understand the process of crack initiation and propagation in ductile materials.
Landes and Begley [3–5] had presented an experimental means to predict the fracture toughness, which is popularly referred
to as load–displacement method. Paranjpe and Banerjee [6], Mills [7], Amouzouvi and Bassim [8], Yin et al. [9], Doig et al.
[10] and Bassim et al. [11] demonstrated experimentally the procedure to evaluate the fracture toughness using SZW mea-
surement. Kobayashi et al. [12] compared the standard procedure of determining the fracture toughness to the one based on
SZW. Bassim [13] and Smith et al. [14] discussed SZW for fracture toughness measurement. Tai [15] employed FEM route to
study the damage ahead of the crack tip. Using CTOD and SZW parameter measurement, Ebrahimi and Seo [16] discussed the
process of crack initiation. Wang and Hwang [17] used crack tip opening angle and J-integral to understand the experimental
test data of CT specimen. Using experimental J–R curves obtained from standard fracture specimens (CT, TPB etc.) several
definitions are available to evaluate crack initiation fracture toughness viz, J0.2, Ji based on SZWc, Ji based on ASTM procedures
and Ji using EGF recommendations [18]. Among these Ji based on critical SZW is getting considered as geometry independent

* Corresponding author. Tel.: +91 755 2457244; fax: +91 755 2457042.
E-mail address: san_bpl@yahoo.com (S. Saxena).

0013-7944/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.engfracmech.2008.12.015
912 S. Saxena et al. / Engineering Fracture Mechanics 76 (2009) 911–920

Nomenclature

B thickness of compact tension specimen


CT compact tension specimen
CTOD crack tip opening displacement
ECrit critical energy density
EFract fracture energy density
FEM finite element method
J fracture toughness
Ji initiation fracture toughness
J1C mode 1 fracture toughness
JQ provisional J1C value
K strength parameter appearing in power law (Eq. (3))
m a material coefficient used in equation (Eq. (1))
n strain hardening exponent
LLD load line displacement
SEM scanning electron microscope
SZH stretch zone height
SZW stretch zone width
SZWc critical stretch zone width
TPB three point bend specimen
W width of compact tension specimen
e elastic–plastic true strain; effective strain
eC critical strain; strain at onset of necking in tensile specimen; Strain hardening exponent (n) value for a specific
case of power law hardening
  variation of true stress–strain material curve
ef fracture strain; ef ¼ ln AA0f ; where A0 is initial cross section area; Af is final cross section area at fracture in ten-
sile specimen
r stress measure

material property [19–23]. SZW can be experimentally evaluated on fracture surface of tested standard fracture specimens
[24]. The problem in its experimental evaluation is in identifying the size of stretching zone on a blunted crack front [25].
Works done in the past related to SZW determination were mainly experimental based with a few attempts to evaluate it
numerically. The numerical determination of SZW is usually done considering it as equal to half of the CTOD with the
assumption of semicircular blunting of the crack tip that is not always true. This numerical approach of SZW calculation also
does not account clearly the mechanism involved in the creation of stretched zone.
Current efforts focus on correlating the experimental SZW measurement with numerical simulation. In the current work
deformation behaviour of CT specimens of Armco iron are simulated by FEM under mode I loading. The investigation essen-
tially comprises a number of FEM simulation using different grain sizes of Armco iron material to study & correlate crack tip
blunting with experimental SZW, SZWc and fracture toughness. Using the large deformation FEM analysis and the input of
material tensile test data, a procedure is established, to determined SZW and its critical value and is compared with exper-
imental SZWc values. The numerically predicted SZW and SZWc values are quite comparable with that of experimental re-
sults. The method also explains the mechanism involved in the creation of stretched zone, there by determines the stage to
calculate critical SZW values. Finally, numerically predicted SZWc is correlated with initiation fracture toughness and com-
pared with experimental values. The developed methodology is tested for variety of other materials. The ABAQUS software
has been used in the analysis. The experimentally [7,24,26–29] obtained properties of Armco iron and for other materials are
used as the effective properties of the homogeneous continuum. In the referred experimental data, fracture toughness (J) had
been determined using multiple specimen technique. Calculated J at the onset of fracture was taken as JQ. JQ was considered
as J1C after satisfying the validity criteria. For all the results inferred same procedure had been used for experimental eval-
uation of J1C. The present study essentially pertains to ductile crack initiation and has not concerned with crack growth re-
lated aspects.

2. Fracture toughness based on micro-scale measurement

There is an increasing effort to use the stretch zone ahead of a fatigue pre-crack as a measure of the fracture toughness of
ductile materials, particularly, in cases where alternative means of measuring the toughness are not possible such as high
strain rates. The plastically deformed crack tip appears as a stretched zone on the fracture surface. Critical crack initiation
point can be defined as a point at which the plastic deformation process is stopped. According to this definition, the size
of the stretched zone does not increase during crack propagation and therefore used for assessing the initiation fracture
toughness. The SZW method relies on the microscopic observation of the crack tip blunting and therefore Ji can be
S. Saxena et al. / Engineering Fracture Mechanics 76 (2009) 911–920 913

established with fewer specimens and fewer size restrictions. In the light of the foregoing it seems evident that the principle of
this approach permits geometry independent evaluation of Ji, provided a high degree of precision in measuring the SZW can be
achieved. SZW on the fracture surface is measured using the scanning electron microscope (SEM). Although there have been
enormous efforts to understand the SZW numerically. Conventionally 45° intersection method is used for determining CTOD
and 12 CTOD is numerically taken as SZW. The fracture toughness J can be determined using the following equation [28]:
J ¼ mrCTOD ¼ mrð2SZWÞ ð1Þ
where m turns out to be a material independent parameter 1.25 [28], CTOD is the crack tip opening displacement, r is the
stress measure defined as integral average on true stress–strain material curve, calculated by [28]
R
eC
rde
r ¼ R0 eC ð2Þ
0
de
Here, critical strain (eC) is defined as strain at onset of necking in tensile specimen equals to strain hardening exponent (n) for
a specific case of power law variation of true stress–strain material curve [28].

3. Numerical FEM model

To establish the methodology for numerical SZW determination, the experimental results of Armco iron at room temper-
ature were used [26,27]. The tensile and fracture toughness data for different materials used in the analysis are given in Table
1. The investigation was limited to CT specimen analysis subjected to mode I type of loading as per ASTM standard [30]. The
FEM mesh model of standard CT is shown in Fig. 1. The symmetry in this case permits consideration of only one half of the
specimen geometry for computational economy. Various investigators successfully employed dense mesh with conventional
elements and advocate against the use of singular or special crack tip elements unless required specifically [31–33]. Follow-
ing the above guidelines, the mesh was constructed with four-noded quadrilateral elements. A magnified view of the mesh
morphology around the crack tip is shown in Fig. 2a and that of the corresponding deformed mesh in Fig. 2b.
Mesh convergence of FEM results are established considering various FEM meshes with element size near the crack tip
varying from 10 to 50 lm in the radial direction and nearly 5–10 lm along the blunted crack surface. The variation of the
results for different meshes is given in Fig. 3. Elements with size smaller than 10 lm in the radial direction were causing
hour-glassing and mesh distortion problems during the crack tip deformation. Along the blunted crack surface, the element
size lower than 5 lm did not improve the convergence significantly. In Fig. 3 it can be observed that there is not much var-
iation in the blunting shape for different meshes used in the study, thus ensuring the mesh independency of results. Hence in
the further investigations element near the crack tip is stick to 10 lm in the radial direction and 5 lm along the crack sur-
face. Such type of initial mesh configuration near the crack tip region helps in keeping good aspect ratio of the elements near
crack tip, during its blunting. Generally the critical SZW in various materials varies up-to 250 lm. Therefore, dense mesh is
provided up-to 250 lm to arrest the steep parameters variation near the crack tip. The mesh is further coarsening in the re-
gion away from the crack tip. This type of mesh configuration allowed the study of deformation at the microscopic scale in
the vicinity of the crack tip. Generally, for the analysis it is suggested to keep the initial finite element mesh shape and the
minimum element size possible near the crack tip that can maintain good aspect ratio during the blunting process.
In the model of the CT geometry, the nodes falling on the symmetry line were arrested in the Y-direction and the load
point is constrained in the X-direction (Fig. 1). The load line displacement (LLD) was applied in steps and the corresponding
deformation contours were used for the analysis. It is known that mode I type fracture in a CT specimen of highly ductile
materials does not truly conform to plane strain condition. Moreover, the experimental J-integral is determined from the
load-displacement record for the three-dimension situation, which intrinsically provides an average value across the crack
front. In this regard Schmitt and Hollstein’s [34] proposal of averaging the results of the plane strain and plane stress con-
ditions was found to be simple and reliable [28,31,35]. All the results presented in this investigation to evaluate SZW and
SZWc were obtained using this averaging procedure in 2D models results. Analyses have been done considering plane strain
and plane stress conditions in separate 2D FEM models. The material undergoes a large strain and rotation at the crack tip,
which necessitated a constitutive framework based on finite deformation for the numerical simulation. The flow behaviour
of the material is assumed to follow the power law
rf ¼ K en ð3Þ
where rf is the true flow stress, K is the strength coefficient, e is the elastic–plastic true strain and n is the strain hardening
exponent. In the entire simulation power law variation has been used as the input. Where K and n gives the entire material
curve. In the study, material’s K value is derived at a lower strain values to match well the yield strength of the material.
Where as n value is derived using a log–log plot of true stress–strain curve and on an average n value will be the best fit.
On the whole it is expected to hold well for the full range of strain values. Although the study does not model grain or grain
size explicitly the effect of grain size enters the FEM model through the dependence of K and n on the grain size in the case of
Armco iron. The analysis were done using commercial FEM software ABAQUS. Introduction of voids, particles and grain
boundaries in the FEM model such as cases given in [36,37] is deliberately avoided in order to specifically understand the
pure continuum behaviour.
914
S. Saxena et al. / Engineering Fracture Mechanics 76 (2009) 911–920
Table 1
Validation of numerical SZWc with experimental results.

Material Grain Test Elastic Yield stress Tensile n Critical energy Fracture energy Experimental Numerical Experimental Numerical Reference
size temperature modulus (E) (ry) (MPa) strength density (MJ/m3) density (MJ/m3) SZWC (lm) SZWc (lm) J1C (N/mm) J1C (N/mm)
(lm) (°C) (GPa) (MPa)
Armco 420 28 196 151 271 0.255 59 672 110 112 93 90 [27,28]
252 28 196 158 268 0.26 64 768 140 139 120 111
118 28 196 180 296 0.28 81 982 142 140 135 123
78 28 196 189 296 0.29 92 1357 176 190 165 169
Iron 78 113 191 151 338 0.30 105 1396 255 256 230 223 [26,27,29]
78 153 187 150 385 0.41 148 1480 261 265 260 239
78 203 180 148 404 0.48 183 1565 263 260 290 253
78 303 175 146 404 0.38 154 1735 256 258 280 261
Fe–0.5Mo 125 28 196 187 359 0.27 96 1075 108 105 112 94 [26,28]
Nickel 95 28 190 172 341 0.43 147 1230 228 220 234/225 188 [24,28]
Fe–Mn–Al steel 28 200 389 707 0.18 88 1635 170 167 186/187 205 [24,28]
Aluminum 28 72.4 60 200 0.40 57 540 248 244 129/120 87 [24,28]
Incoloy 800 24 195 236 515 0.34 157 2682 279 282 308 326 [7,28]
Inconel 600 24 214 285 640 0.39 255 2225 109 109 209 178 [7,28]
SA387 C steel 316 185 218 476 0.26 88 1060 84 80 76/110 67.4 [7]
S. Saxena et al. / Engineering Fracture Mechanics 76 (2009) 911–920 915

Fig. 1. Finite element mesh for compact tension specimen. Geometry dimension: W = 100 mm, B = 50 mm, a0/W = 0.55.

Fig. 2. Magnified view of the FEM mesh at the crack tip (a) initial configuration; (b) deformed configuration. CTOD = crack tip opening displacement;
SZH = stretch zone height; SZW = stretch zone width.

4. Numerical determination

4.1. Stretch zone width

In the present study, using tensile test data, numerical investigation was performed to establish a procedure for numerical
determination of SZW, its critical value (SZWc) and an insight into the possible mechanism behind the formation of the
stretched zone. In this SZW evaluation, no assumption is made about the shape of the blunted crack tip or material, in trying
916 S. Saxena et al. / Engineering Fracture Mechanics 76 (2009) 911–920

0.25
A B
0.20 A
Enlarged View
B

C D
0.15

Y (mm)
C D
0.10 Mesh 1 = 10 µm RD ; 5 µm BCS
Mesh 2 = 15 µm RD ; 5 µm BCS

0.05 Mesh 3 = 25 µm RD ; 5 µm BCS


Mesh 4 = 50 µm RD ; 5 µm BCS
Mesh 5 = 50 µm RD ; 10 µm BCS
0.00
0.0 0.1 0.2 0.3 0.4 0.5
X (mm)

Fig. 3. Convergence test for the finalization of the mesh: Variation of the blunting shape for a given LLD for different mesh morphologies. RD = element size
along radial direction; BCS = element size along blunted crack surface.

to achieve a physically tangible methodology using probable mechanism involved in the creation of the stretched zone. This
was missing in its earlier method of numerical determination which assumes it as half the CTOD defined by 45° line method
[38,39]. Several parameter variations were considered in the study to define the methodology to assess the SZW numerically.
Parameters considered in the study include equivalent stress, equivalent strain and plastic energy density variations. The
past experience [28,31,35] shows plastic energy density parameter to be more suitable for the problem description. Exper-
imentally the value of SZW initially increases with LLD but saturates once the SZW reaches a critical value (SZWC). To sim-
ulate the experimental behaviour in numerical experimentation, the variation of the plastic energy density with increase in
LLD is plotted in Fig. 4 for different nodes near the crack tip. The plot of variation of plastic energy density showed that al-
most the entire energy density dissipates near the crack tip as compared to the negligible amount of energy density dissi-
pated in the remaining region. The numerical results although roughly define the region of maximum energy density

a 1200
1100
Plastic energy density (MJ/m )

1000
3

900
800
700
600
500
node-4
400
node-3 node-5
300 node-6
node-2
200 node-1
node- 8 & 9
100
node-7
0
0 1 2 3 4 5
Load line displacement(LLD)(mm)

92 9
340 8
7
216 6
5
4

2 Node number

1 BLUNTED CRACK TIP

Fig. 4. Results of Armco Iron (Grain size = 78 lm): (a) Variation in the plastic energy density with increase in LLD under plane strain condition. (b) Snap shot
variation at LLD = 4.5 mm.
S. Saxena et al. / Engineering Fracture Mechanics 76 (2009) 911–920 917

dissipation but to quantify this region a critical energy density still needs to be defined. The critical delineating energy den-
sity can corresponds to onset of necking at the critical strain (eC) in tensile test. Thus SZW is related to large plastic defor-
mation, which starts on the blunted free crack surface at the onset of necking and is equivalent to the onset of necking at the
critical strain (eC), in tensile test. This stretched region defined by critical delineating energy density will keep on increasing
with increase in LLD. And, it is difficult to define the critical LLD to measure the critical SZW without defining a limiting en-
ergy density corresponding to crack initiation in fracture specimen.
For a specific case like power law variation of true stress–strain material curve, the critical strain (eC) is equal to the strain
hardening exponent (n). For Armco iron the hardening exponent (n) varies between 0.25 and 0.3 [27]. The same definition
has been used to delineate the intensely deformed region while defining the ‘‘characteristic distance” of the material [31,35].
Following this, the critical energy density used to delineate the highly stretched region is defined on the true stress–strain
material curve as the energy density integral up-to the critical strain (eC) equal to the strain hardening exponent (n) for the
case of power law variation and is equivalent to the onset of necking in tensile test [31]
Z eC
ECrit ¼ rde ð4Þ
0

The variation of plastic energy density near the blunted crack tip with respect to deformed X-co-ordinates with increase
in LLD is shown in Fig. 5a. The region having energy density greater than the critical energy density delineates the stretched
region (along A–A line with Ecrit. = 92 MJ/m3 in Fig. 5a), denoted as numerically determined SZW.
In Fig. 5b, the proposed methodology based SZW results are compared with half CTOD (SZH) results measured conven-
tionally (45° line method). The SZH is also determined using the proposed method of SZW considering as Y-deformation of
the region where the plastic energy density exceeds the critical energy density defined by Eq. (4). At smaller LLD both ways
are same. At higher LLD, the proposed method SZW is lower than the SZH given by the two methods. It can be appreciated
that in a particular case blunting shape angle works out to be greater than 45°. In conventional 45-degree line method semi-
circular blunting shape is always assumed whereas no such assumption is made in the proposed method of SZW determi-
nation. The validation of proposed methodology of SZW determination is also done using the experimental results of SZW
and SZWc values obtained using multiple specimens technique [29]. Fig. 6 showed the averaged FEM (plane strain and plane
stress condition) results of J-integral and the variation of experimental fracture toughness results with SZW. The figure also

a 1100
1000
Plastic energy density (MJ/m 3)

Plane Strain Condition


900
A-A : Critical deformation energy
800 LLD- 4.5
density for SZW [ 26,27]
LLD- 4.0
700
LLD- 3.5 SZW increases along the A-A line at
600 3
E Crit. = 92 MJ/m
500 LLD- 3.0

400 LLD- 2.5

LLD- 2.0
300
LLD- 1.5
200
LLD- 1.0
100 A A
LLD- 0.5
0
0.0 0.1 0.2 0.3 0.4
X (mm)

b 0.3
Plane strain condition
0.25 Proposed method- SZW
Stretch zone size (mm)

Proposed method- SZH


0.2 45 degree cutoff = 1/2 CTOD

0.15

0.1

0.05

0
0 1 2 3 4 5
Load-line-displacment (mm)

Fig. 5. Results of Armco iron (grain size = 78 lm): (a) plastic energy density variation with X coordinates (deformed) from the crack tip under plane strain
condition; (b) variation of stretch zone size with LLD using 45° line method and proposed method; n = 0.29 [26,27]; ECrit = 92 MJ/m3.
918 S. Saxena et al. / Engineering Fracture Mechanics 76 (2009) 911–920

200

175 Material:

Increasing J integral (kJ/m )


2
Armco iron
150 (78 µ m grain size)
125

100

75 Average of plane strain and plane stress

Experimental [29]
50
Eqn. 1 [28] ; SZW by proposed method
25 J-integral FEM

0
0 0.05 0.1 0.15 0.2 0.25
SZW (mm)

Fig. 6. Comparison of experimental, FEM and Eq. (1) based on SZW calculated numerically.

0.3 1500
Average of plane strain and plane stress

Plastic energy density at crack tip


1200

0.2
SZW (mm)

900

(MJ/m 3)
600
0.1

300
LLD-SZW
LLD-Plastic energy densit
0 0
0 1 2 3 4 5
Load-line-displacement (mm)

Fig. 7. Determination of critical SZW in Armco iron (Grain size = 78 lm); ef = 2.33 [26,27]; EFract. = 1357 MJ/m3.

showed the fracture toughness calculated by Eq. (1) using averaged FEM results of SZW determined using the proposed
method. The experimental blunting slope matches very well with the two FEM results, thus again validating the proposed
SZW determination methodology.

4.2. Critical stretch zone width

With increasing LLD the energy density accumulated near the crack tip increases and reaches a value associated with frac-
ture. It is defined on the true stress–strain material curve as the energy density integral up-to fracture strain (ef), obtained in
tensile test specimen when fracture occurs, given as
Z ef
EFract: ¼ rde ð5Þ
0

For a grain size of Armco iron, the variation of plastic energy density at the crack tip and SZW (deformed X-co-ordinate from
the crack tip) with increase in LLD is shown in Fig. 7. The above defined fracture energy density is used to determine the crit-
ical SZW at critical LLD, where the crack initiation occurs in standard fracture specimen. The procedure to determine critical
SZW is shown in Fig. 7. The terms ECrit. and EFract. (Eqs. (4) and (5)) are determined from the total strains of the tensile test.
However, considering that the elastic contribution is negligibly small compared to the plastic part and also as ABAQUS soft-
ware provides only the plastic contribution, the two have been compared in Figs. 5 and 7. The numerically determined crit-
ical SZW results for different grain sizes of Armco iron are given in Table 1. In Armco iron, at room temperature the
experimental SZWc results showed decrease in SZWc values with increase in grain size. The proposed method of SZW deter-
mination accurately predicts the trend as well as the magnitude of critical SZW with increase in grain size. This methodology
also explains the mechanism involved in the creation of stretched zone and defines the stage where to calculate critical SZW.
Primarily, the methodology for determining SZW as well as its critical value is validated using the experimental results
published in reference [26,27] pertaining to Armco iron of different grain sizes. Armco iron being a clean single phase
S. Saxena et al. / Engineering Fracture Mechanics 76 (2009) 911–920 919

material, it allowed numerical experiments to be conducted without a major concern regarding heterogeneous (un-clean)
microstructures. However, in order to verify the validity of the procedure in a wider domain the experimental results pub-
lished pertaining to Fe–0.5Mo [26], Nickel [24], Fe–Mn–Al steel [28], Aluminum [24], Incoloy 800, Inconel 600 and SA387 C
Steel [7] were also numerically simulated as per the above mentioned procedure. The details of experimental results and
numerically predicted results are given in Table 1. In all the materials, the proposed method predicts the trend as well as
the magnitude of the critical SZW in CT specimen reasonably well, satisfactorily confirming the wider applicability of the
suggested procedure.
The procedure to numerically predict SZW and SZWc can be summarized as follows:

(1) Plot the variation of plastic energy density with deformed X-co-ordinates w.r.t. crack tip, for every increase in load line
displacement (LLD) considering plane stress and plane strain conditions in separate models.
(2) Define the critical and fracture energy densities for the material. Critical energy density corresponds to the energy density
at onset of necking in tensile specimen and fracture energy density corresponds to the energy density required to frac-
ture the tensile specimen. Both these energy densities are defined on a true stress–strain material variation, as same
material definition is used in FEM analysis.
(3) Measure the region (deformed X-co-ordinates) exceeding the material’s critical energy density defined by Eq. (4) in the
figures drawn in step 1. The average deformed X-co-ordinate w.r.t. crack tip (average of plane stress and plane strain
condition results) is SZW and it will be a region on blunted crack surface, where the energy density exceeds the critical
energy density.
(4) The SZW measured in step 3 will keep on increasing with LLD until the average energy density at the crack tip (average
of plane stress and plane strain condition results) reaches fracture energy density. Critical SZW is then measured at a
critical LLD when the average energy density at the crack tip exceeds the fracture energy density (Fig. 7). SZWc will now
be a region on blunted crack surface, where the energy density varies from critical at one end and fracture energy den-
sity on the other end (at the crack tip).

Once the critical stretch zone width is predicted numerically accurately, the fracture toughness can easily be calculated
using Eq. (1).

5. Numerical J1C determination

The yield strength as a stress measure does not relate the behaviour of the material during strain hardening subsequent to
yielding. The flow stress (average of yield to ultimate) although taking hardening into account, implicitly assumes the var-
iation to be linear. To account for the non-linearity, Suresh et al. [28] considered the stress measure as given by Eq. (2) that is
based on an integral average on the true stress–strain material curve. For a specific case like power law variation, it was as-
sumed the critical strain to be the strain hardening exponent (n) [28]. The fracture toughness variation results obtained by
Eq. (1) using FEM results of SZW are shown in Fig. 6. The averaged plane strain and plane stress J-integral value and corre-
sponding SZW results determined by FEM analyses are also shown in this. In Fig. 6, it can be appreciated that the numerically
predicted fracture toughness values using Eq. (1) is comparable with experimental results. Similar behaviour can also be
seen in Table 1, where the experimental and numerical (using Eq. (1) and numerically predicted critical SZW) results of dif-
ferent grain sizes of Armco iron tested at different temperature and for different other materials are tabulated. Thus it shows
that using critical material flow values, the present method can reasonably predict the experimental initiation fracture
toughness results in different ductile materials.

6. Discussion and conclusion

The present study attempts at achieving an insight into defining the procedure for numerical determination of stretch
zone width (SZW) and its critical value (SZWc) using a large deformation FEM and to observe SZWc correlation with fracture
parameters. The investigation essentially comprised of number of FEM simulations to study and correlate various ductile
fracture parameters with experimental results. To establish the procedure of numerical determination of SZW, a detailed
analysis of the zone of intense plastic deformation was carried out. FEM simulations were performed to find the correspon-
dence, if any, between important test results obtained in tensile specimen test and the different types of stress, strain and
energy density distributions in the vicinity of the crack tip during blunting phenomenon. An attempt has been made to
establish a procedure to determine SZW, its critical value and is finally compared with experimental values. The region of
high deformation at the crack tip is delineated using the critical and fracture energy densities obtained on the true stress–
strain material curve using tensile test data. The proposed method of SZW determination accurately predicts the trend as
well as the magnitude of SZW, critical SZW and compares well with experimental results. This methodology also explains
the mechanism involved in the creation of stretch zone and defines the stage where to calculate critical SZW. Its dependence
on grain size matched the experimental findings. Finally it is correlated with initiation fracture toughness. It is concluded
that the proposed method of SZW determination can reasonably simulate the process of blunting of the crack tip and can
predict the material’s initiation fracture toughness (JIC) via SZW and its critical value.
920 S. Saxena et al. / Engineering Fracture Mechanics 76 (2009) 911–920

References

[1] Anderson TL. Fracture mechanics fundamental and applications. 2nd ed. CRC press; 1994.
[2] Rice JR. A path independent integral and the approximate analysis of strain concentration by notches and cracks. Tran ASME, J Appl Mech
1968;35:379–86.
[3] Landes JD, Begley JA. The effect of specimen geometry on J1C, fracture toughness. In: ASTM STP 514. American Society for Testing and Materials; 1971.
p. 24–39.
[4] Begley JA, Landes JD. The J integral as a fracture criterion. In: ASTM STP 514. American Society for Testing and Materials; 1972. p. 1–20.
[5] Landes JD, Begley JA. Test results from J-integral studies: an attempt to establish a J1C testing procedure, fracture analysis. In: ASTM STP 560. American
Society for Testing and Materials; 1974. p. 170–86.
[6] Paranjpe SA, Banerjee S. Interrelation of crack opening displacement and J-integral. Engng Fract Mech 1979;11:43–53.
[7] Mills WJ. On the relationship between stretch zone formation and the J integral for high strain-hardening materials. J Test Eval 1981;9(1):56–62.
[8] Amouzouvi KF, Bassim MN. Determination of fracture toughness from stretch zone width measurement in predeformed AISI Type 4340 steel. Mat Sci
Engng 1982;55:257–62.
[9] Yin SW, Gerbrands RA, Hartevelt M. An investigation of the blunting line. Engng Fract Mech 1983;18(5):1025–36.
[10] Doig P, Smith RF, Flewitt PEJ. The use of stretch zone width measurements in the determination of fracture toughness of low strength steels. Engng
Fract Mech 1984;19(4):653–64.
[11] Bassim MN, Mathews JR, Hyatt CV. Evaluation of fracture toughness of HSLA80 steel at high loading rates using stretch zone measurements. Engng
Fract Mech 1992;43(2):297–303.
[12] Kobayashi Hideo, Nakamura Haruo, Nakazawa Hajime. Comparison of J1C test methods recommended by ASTM and JSME, elastic–plastic fracture test
methods: user’s experience. In: Wessel ET, Loss FJ, editors. ASTM STP 856. American society for Testing and Materials; 1985. p. 3–22.
[13] Bassim MN. Use of the stretch zone for the characterization of ductile fracture. J Mater Process Tech 1995;54:109–13.
[14] Smith J, Bassim MN, Liu CD. Effect of fatigue pre-cracking on stretch zone formation. Engng Fract Mech 1995;52(3):401–8.
[15] Tai WH. Computer simulation of plastic damage evolution in cracked body of low-carbon steel. Engng Fract Mech 1996;54(1):141–6.
[16] Ebrahimi F, Seo HK. Ductile crack initiation in steels. Acta Mater 1996;44(2):831–43.
[17] Wang Bo, Hwang KC. Structural integrity assessment in elastic–plastic fracture. Engng Fract Mech 1996;54(4):471–8.
[18] Singh PK, Chattopadhyay J, Kushwaha HS, Tarafder S, Ranganath VR. Tensile and fracture properties evaluation of PHT system piping material of PHWR.
Int J Press Ves Pip 1998;75:271–80.
[19] Kim YJ, Kim JS, Cho SM, Kim YJ. 3-D constraint effects on J testing and crack tip constraint in M(T), SE(B), SE(T) and C(T) specimens: numerical study.
Engng Fract Mech 2004;71:1203–18.
[20] Chattopadhyay J, Dutta BK, Kushwaha HS. Experimental and analytical study of three point bend specimen and throughwall circumferentially cracked
straight pipe. Int J Pres Vess Pip 2000;77:445–71.
[21] Saxena S, Ramakrishanan N. Characterizing crack initiation load in circumferentially through-wall cracked elbows under bending load. Int J Press Vess
Pip 2007;84:493–501.
[22] Eisele U, Herter KH, Schuler X. Influence of the multiaxility of stress state on the ductile fracture behaviour of degraded piping components. In:
Schwalbe KH, Berger C, editors. ECF 10: structural integrity: experiments models and applications, vol. 1. Berlin; 1994; p. 249–54.
[23] Roos E, Eisele U, Silcher U. Effect of stress state on the ductile fracture behavior of large-scale specimens. In: Hackett EM, Schwalbe KH, Dodds Jr RH,
editors. Constraint effects in fracture, ASTM STP 1171. Philadelphia: American Society for Testing and Materials; 1993. p. 41–64.
[24] Srinivas M, Kamat SV, Rama Rao P. A fractographic technique for the estimation of initiation fracture toughness J1c for ductile materials. ASTM J Test
Eval 1994:302–8.
[25] Tarafder M, Dey S, Sash B, Sivaprasad S, Tarafder S. Initiation fracture toughness of HSLA steel through automatic measurement of stretch zone. Mater
Sci Tech 2004;20:1531–7.
[26] Srinivas M, Sundararajan G, Malakondaiah G, Rama Rao P. An analysis of ductile fracture initiation toughness in iron, its binary alloys and nickel. Proc
Roy Soc Lond 1994;447(A):237–51.
[27] Srinivas M, Malakondaiah G, Armstrong RW, Rama Rao P. Ductile fracture toughness of polycrystalline Armco Iron of varying grain size. Acta Mater
1991;39(5):807–16.
[28] Suresh RKV, Ramakrishnan N, Srinivas M, Rama Rao P. On the determination of JIC using the stretch zone width method. J Test Eval 1999;27(3):211–7.
[29] Srinivas M, Malakondaiah G, Rama Rao P. Fracture toughness of FCC Nickel and BCC iron in the temperature range 77–773 K. Acta Metall Mater
1993;41(4):1301–12.
[30] ASTM standard E813-81. Standard test method for J1C, a measure of fracture toughness. Annual book of ASTM standards Philadelphia; 1986. p. 768–86.
[31] Ramakrishnan N, Rama Rao P. An FEM study on crack tip blunting in ductile fracture initiation. Comput Mater Cont 2005;2(3):163–76.
[32] Liebowitz H, Moyer ET. Finite element methods in fracture mechanics. Comput Struct 1989;31:1–9.
[33] Tsamaphyros G, Glannakopoulos AE. The optimum finite element grids around crack singularities in bilinear elasto-plastic materials. Engng Fract Mech
1989;32:515–22.
[34] Schmitt, Hollstein T. Numerical evaluation of crack tip opening displacements: 2D and 3D applications. In: Proceedings of the Workshop on the CTOD
Methodology; 1985. p. 3–20.
[35] Saxena S, Ramakrishnan NA. Comparison of micro, meso and macroscale FEM analysis of ductile fracture in a CT specimen (mode I). Comput Mater Sci
2007;39(1):1–7.
[36] Gao X, Faleskog J, Shih CF, Dodds Jr RH. Ductile tearing in part-through cracks: experiments and cell-model predictions. Engng Fract Mech
1998;59(6):761–77.
[37] Gullerud Arne S, Gao Xiaosheng, Dodds Jr RH, Haj-Ali R. Simulation of ductile crack growth using computational cell: numerical aspects. Engng Fract
Mech 2000;66:65–92.
[38] Shih CF. Relationships between the J-integral and the crack opening displacement for stationary and extending cracks. J Mech Phys Solid
1981;29:305–81.
[39] Tracy DM. Finite element solutions for crack-tip behaviour in small-scale yielding. Trans ASME. J Engng Mater Tech 1976:p–151.

You might also like