Download as pdf or txt
Download as pdf or txt
You are on page 1of 9

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/270230243

Determination of Elastic Plastic Fracture Toughness Parameters for a Compact


Tension Specimen Using the Finite Element Method

Article in Materials Testing · June 2012


DOI: 10.3139/120.110345

CITATION READS

1 1,137

2 authors:

Ugur Ozdemir Çinar Yeni

6 PUBLICATIONS 100 CITATIONS


Dokuz Eylul University
33 PUBLICATIONS 483 CITATIONS
SEE PROFILE
SEE PROFILE

All content following this page was uploaded by Çinar Yeni on 05 August 2015.

The user has requested enhancement of the downloaded file.


408 MATERIALS TESTING FOR JOINING TECHNOLOGY

Determination of Elastic Plastic


Fracture Toughness Parameters
for a Compact Tension Specimen
Using the Finite Element Method
Uğur Özdemir and Çınar Yeni, In this study, elastic-plastic fracture toughness of an aluminum alloy is
Bornova-İzmir, Turkey investigated. J-integral was calculated using the finite element software
and formulation. Crack tip opening displacement (CTOD) is determined
by using the plastic hinge model (δph) and direct measurement (δ5). J-Δa
and CTOD-Δa resistance curves are constructed and compared with those
obtained from formulation and those generated from test data, respec-
tively. Good agreement has been obtained between the test results, the
formulation and the results of the finite element analysis. The relation-
ship between δph and δ5 is nonlinear and as CTOD increases δ5 becomes
larger than δph. The relationship between J-integral and CTOD is found
to be linear for both δph and δ5. It has been seen that the proportionality
constant dn considerably depends on the method of calculation of CTOD.

It is expected that the materials used in ar- growth resistance and fracture behaviour imens in the analysis and estimated the
eas requiring advanced technology, such directly affects the useful economical lives magnitude of CTOD using 90°-intercept
as aerospace industry, which are subjected and reliability of the applications. It is pos- method and the plastic hinge model. Their
to various loadings during their service sible to calculate the fracture resistance of results indicated that there existed a vari-
lives and environmental conditions, main- ductile materials by using elastic-plastic ance in estimation of CTOD by 90°-inter-
tain their mechanical and microstructural fracture mechanics. By modeling the homo- cept method and the plastic hinge model.
properties, as well as, being cost efficient geneous materials using methods like fi- The relationship between CTOD and J was
[1]. Fuel efficiency is directly related to the nite element and boundary element meth- linear and the linear proportionality con-
weight of an aircraft. Any increase in the ods, the deformation behaviour and crack stant, dn, was obtained less than 1 in their
weight of the aircraft, causes an increase in growth behavior can be obtained with analysis for CT and SENB specimens with
the amount of fuel consumed and the cost, rapid success [8]. a/W = 0.5. This constant, dn, was found to
and causes more environmental pollution. The elastic-plastic fracture parameters J- be strongly dependent on several parame-
For this reason, such structures must be integral and CTOD (δ) have been developed ters such as CTOD, specimen geometry and
made of lightweight and high strength ma- separately, and they are both widely used a/W ratio [13].
terials [2, 3]. in structural integrity assessments [9]. Nikishkov et al. (1999) investigated the
Al-Mg-Si (6xxx series) alloys are used in Here, J is the J-integral defined by Rice, transformation fracture mechanics para-
a wide range of application areas such as CTOD is the Crack Tip Opening Displace- meters CTOD, δ5, to standard CTOD and J
aerospace industry, aircraft body and wing ment. In elastic-plastic fracture mechanics, for plane sided and side-groved three point
structures, due to their high strength and fracture toughness of a material can be bending (3PB) and C(T) specimens. They
corrosion resistance, improved weldability characterized by J-Δa and CTOD-Δa resist- solved a series of elastic-plastic 3D fracture
performance and high fracture toughness ance curves (called R-curves), where Δa is mechanics problems for 3PB and C(T) spec-
properties [4, 5]. Within these series, new the crack extension [10-12]. imens using the finite element model
generation alloys, such as AA 6013 and Kudari and Kodancha (2008) investi- (FEM). They proposed a simple two-region
6056 are preferred to aircraft lower body gated the relationship between the J-inte- approximation for relations of CTOD, δ5 to
panels [6, 7]. The engineering materials gral and CTOD using the finite element standard CTOD and J [14].
used in these areas are required to have analysis. They used compact tension (C(T)) Shan et al. (1993) presented a numerical
high fracture toughness values, since crack and single edge notched bend (SENB) spec- simulation of the stable crack growth in a

© Carl Hanser Verlag, München Materials Testing 54 (2012) 6


MATERIALS TESTING FOR JOINING TECHNOLOGY 409

rials, for mode I type loading, resistance


curves can be obtained by using K, the
stress intensity factor in linear elastic frac-
ture mechanics (LEFM) approach, whereas
J and CTOD parameters are used in the
elastic-plastic fracture mechanics (EPFM)
approach (ASTM E1820). In this study, the
loading causes severe plastic deformation
ahead of the crack tip, therefore elastic-
plastic fracture parameters CTOD and J are
applied in the analyses. Fracture tough-
ness tests are conducted to determine the
crack resistance curves using C(T) speci-
mens. The dimensions of the C(T) speci-
men are shown schematically in Figure 1.
a) b)
Fatigue pre-cracking was carried out in
Figure 1. a) Dimensions of C(T) 50 specimen used in modeling (a/W = 0.5), b) apparatus for testing order to obtain a sharp crack with a ratio of
compact tension specimens a/W = 0.5 as shown in the figure. CTOD
values are measured directly with a δ5 clip
a) b) gauge during testing. This measuring ap-
paratus is fixed to a distance of 2.5 mm
from the right and left sides of the pre-
crack and measures the CTOD value di-
rectly at the tip of the fatigue pre-crack
(Figure 2a). Because the fixing distance is a
total of 5 mm on both sides of the crack, the
measured CTOD values are called CTOD,
δ5, Figure 2b. Load versus CTOD, δ5 data
are recorded during testing and crack ex-
tension is measured afterwards to con-
struct CTOD, δ5-∆a resistance curves. The
experimental results are employed for cal-
culation of the J-integral and for compari-
son of CTOD values obtained by the finite
Figure 2. a) CTOD, δ5 clip gauge, b) placement of CTOD, δ5 clip gauge onto the crack tip region element analysis.
of C(T) specimen J-Integral. Many of the problems of practi-
cal interest are in the elastic-plastic regime
requiring an estimation scheme to calculate
thick smooth C(T) specimen with two-di- The aim of this study is to determine the the J-integral. The path-independent line in-
mensional elastic-plastic finite element elastic-plastic fracture toughness behav- tegral, which represents the strain energy
analysis. They calculated J directly from iour of an aluminum alloy using the finite release rate for nonlinear elastic materials
the load-displacement curves according to element software. These values will be has been described by Rice as:
the ASTM E813-87 and compared their re- compared with test results and those of for-
⎡ ⎛ ∂u ⎞ ⎤
sults with the numerically calculated J-Δa mulation. Relations between the two CTOD J = ∫ ⎢ wdy − Ti ⎜ i ⎟ ds ⎥ (1)
curves. The J-Δa curve obtained from nu- parameters, namely δph and δ5 and also ⎢ ⎜⎝ ∂x ⎟⎠ ⎥
Γ⎣ ⎦
merical analysis lies in the scatter band of between J-integral and CTOD obtained by
experimental data. They reported that direct measurement as well as by the plas- with w: strain energy density, Γ: arbitrary
there is a satisfactory agreement between tic hinge model will be examined. J-integral closed contour around the crack tip, as
numerical and experimental results [15]. and CTOD fracture toughness parameters shown in Figure 3, Ti: components of the
Shi et al. (1998) investigated the effects are used for comparison and selection of traction vector, ui: displacement vector
of weld strength mismatching and geome- materials, quality assurance and identifica- components, ds: length increment along
try parameters on the relationship between tion of structural defect tolerance pur- the contour Γ.
the J-integral and the CTOD numerically. poses.
The main results indicated that weld The strain energy density is defined as:
strength mismatching has only a weak in- Materials and Procedures
fluence on the relationship between the J- ⎛ εij ⎞
integral and CTOD at low load levels. The In this study, 6013 aluminum alloy (AA w = ⎜ ∫ σ ij dεij ⎟ (2)
⎜ ⎟
results also showed that crack size and 6013), widely used in the aerospace indus- ⎝0 ⎠
weld width have an influence on the rela- try is investigated in terms of its fracture
tionship between the J-integral and CTOD toughness behaviour. To establish the frac- with σij and εij: stress and strain tensors,
at high load levels [16]. ture toughness behaviour of metallic mate- respectively.

54 (2012) 6
410 MATERIALS TESTING FOR JOINING TECHNOLOGY

Figure 4. Plastic area under load-load line


displacement curve
Figure 3. Closed contour around the crack tip

Figure 5. Estimation of CTOD from a 90°


intercept construction and HRR displacement

The traction vector is a stress vector nor- width, a: distance between the load line with σy: yield strength of the material, m:
a
mal to the contour. That is, if a free body and the crack tip. 2 + dimensionless
⎡ constant
⎛ a ⎞ that depends
⎛ a⎞
2 on ⎛ ⎞ 3
⎛ ⎞ W a
diagram was to be constructed on the mate- For C(T) type specimen the f a =is stress⎢0.886 + 4.64
state and material
⎜ ⎟ − 13.32
properties.
⎜ ⎟ + 14.72 ⎜ ⎟
⎝ W⎠ 3 ⎢
⎝ W ⎠ ⎝ W ⎠ ⎝ W ⎠
rial inside of the contour, Ti would define given by: ⎛ a ⎞ Shih
2 ⎢ has shown that the relationship be-

⎜ 1 − ⎟
the normal stresses acting at the bounda- a ⎝ ⎠ 2 J and δ can 3be obtained theoretically
Wtween
ries. The components of the traction vector 2+ ⎡ by Hutchinson-Rice-Rosengren (HRR) 4⎤
stress
f⎛ a ⎞= W ⎢0.886 + 4.64 a − 13.32 a ⎞ + 14.72 ⎛ a ⎞ − 5.60 ⎛ a ⎞ ⎥
⎛ ⎞ ⎛
are given by: ⎝ W⎠ 3 ⎢ ⎜ ⎟ ⎜ field
⎟ equations ⎜ as:⎟ ⎜ ⎟ ⎥
⎛ ⎝ W⎠ ⎝ W⎠ ⎝ W⎠ ⎝ W⎠ ⎥
a ⎞ 2 ⎢⎣ ⎦
⎜ 1 − ⎟ J
Ti = σijnij (3) ⎝ W⎠ δ = dn (12)
a σy
with nij: ⎛components 2 +of the unit
⎡ vector nor- ⎛ a⎞ ⎛ a⎞
2
⎛ a⎞
3
⎛ a ⎞ ⎤⎥
4

f a ⎞= W ⎢0.886 + 4.64 − 13.32 + 14.72 − 5.60


mal to Γ⎝[17]. W⎠ 3 ⎢ ⎜ ⎟ ⎜ ⎟ ⎜ ⎟ ⎜ ⎟ ⎥ (7) with dn: dimensionless constant which ex-
⎛ ⎞ 2 ⎢ ⎝ W⎠ ⎝ W⎠ ⎝ W⎠ ⎝ W⎠ ⎥
The cracked structure a is ⎣
solved to obtain ⎦ hibits a strong dependence on the strain
⎜1 − ⎟
the displacements. W ⎠ a solution is ob-
⎝ Then, hardening exponent and a mild depencence
tained for a second closed contour by intro- The plastic component of J, Jpl, is given by: on σy/E [13, 18, 19]. From Equations (11)
ducing a small increment to the crack. The and (12) the relation between m and dn is:
ηA pl
J-integral can be determined using differ- J pl = (8)
ent closed contours for increasing crack Bb 1
dn = (13)
lengths. m
J-integral is calculated separately as the with η: dimensionless coefficient and for a
sum of elastic and plastic components by C(T) type specimen it is given by: Theoretically, the crack tip opening dis-
adopting formulations of the results ob- placement is usually estimated using two
tained experimentally. η = 2 + 0.522 ⎛ b ⎞ , b = W – a (9) methods. The first one, δi, is the 90° inter-
⎝ W⎠
cept method, which is defined as the open-
J = Jel + Jpl (4) Apl is the plastic work, calculated from the ing distance between the intercept with the
load versus load line displacement curve as deformed crack profile of two lines drawn
The linear component of J will be equal to shown in Figure 4. back from the crack tip at 45°, Figure 5.
Jel, which is equated to the energy release J-integral as expressed by the sum of its The second one, δph, is determined using
rate, G. elastic and plastic components can be writ- the plastic hinge model by conversion of
ten as: crack mouth opening displacement to
K2I
J el = G = (5) CTOD using a rotation factor. In this case,
K2I ηA pl
E' J= + (10) CTOD δph, is calculated from:
E' Bb
K2I rp (W − a)Vp
with G: energy release rate, E': Young’s δ ph = δ el + δ pl = + (14)
Modulus and E' = E (for plane stress), EPFM considers extensive plastic deforma- mσ y E rp (W − a) + a
'

E (for plane strain) tion ahead of the crack tip prior to fracture.
E' =
1 − ν2 The fracture toughness parameters J and with Vp: plastic displacement at the crack
with ν: Poisson’s ratio, KI: stress intensity CTOD in EPFM are required to be inter- mouth, rp: plastic rotation factor, varying
factor given by: changeable with each other. Thus, it is es- with a/W ratio and is between 0.44-0.47 for
sential to examine the relation between J C(T) specimens according to ASTM E1290-
P
KI = f⎛ a ⎞ (6) and δ. The general relation between J and δ 99, b: ligament length and a is the crack
⎝ W⎠ is given by: length of the specimen.
B W
The third determination of CTOD para-
with B: specimen thickness, W: specimen J = mσyδ (11) meter is the one used in this study, which

54 (2012) 6
MATERIALS TESTING FOR JOINING TECHNOLOGY 411

Figure 6. Engineering
stress-strain curve of
6013 alloy

Figure 7. The partitioned geometry of the model

eter and the crack tip element degeneracy


allows different singularity types to be de-
fined. In the present analysis, a value of 0.25
is used for the midside node parameter. This
definition moves the midside nodes on the
element sides adjoining the collapsed edge
Figure 8. Finely meshed
to the 1/4 points of the elements. At the
elements around the
crack tip region crack tip the element sides are collapsed
with single node-type degenerate element
control. These settings combine to create a
1/r singularity in strain [21].
The partitioning of the geometry is de-
fined by the circular lines centered on the
crack tip (Figure 7); this partitioning strat-
egy facilitates the generation of a focused
mesh (Figure 8). 20-node quadratic brick
finite element (C3D20) types were used to
model the specimens. Each specimen con-
tains 17478 elements. The circular parti-
tioned areas are sweep meshed using the
is the CTOD, δ5 directly measured during tion, all other motions of the pin are re- “hex-dominated” element shape and “ad-
testing. strained. Surface-to-surface contact with a vancing front” algorithm. The remaining
In this study, a relationship between δ5 finite-sliding formulation is defined be- portion of the model is meshed using “hex”
and the calculated δph is constructed. The tween the pins and the specimen. Two element shape, “sweep meshing” tech-
elastic part of δph is very small compared to analysis steps are used. In the first step, nique and “medial axis” algorithm.
the plastic part and, hence, it is neglected contact is established between the pins and The center of the Cartesian coordinate
in this work. the specimen by applying a small displace- system was placed on the axis of symmetry
Finite Element Analysis Approach. For ment (1 × 10−5 mm) in the vertical (y) di- of the specimen as seen in Figure 7.
the aluminium alloy 6013, the modulus of rection. In the second step, controlled dis-
elasticity is 68 GPa and Poisson’s ratio is placement loading of the pins is applied. Results and Discussion
0.28. Its yield strength is 360 MPa and the Definition of the Model. The partitioned
engineering stress-strain curve for this ma- geometry of the model is shown in Figure 7. The results obtained from Abaqus for the
terial is plotted as in Figure 6. In the elas- The cracked specimen is highlighted by a J-integral and CTOD, δ5 were compared, re-
tic-plastic finite element analyses of frac- bold black line, which is called seam. The spectively, with the results computed by
ture mechanics specimens, it is often nec- seam is defined in the model as a surface (in using formulation in Matlab software and
essary to use 3D calculations in order to a solid part) that is originally closed but can with the testing results. In order to deter-
obtain realistic curves, although the speci- open during an analysis. The virtual crack mine the plastic component of the J-inte-
men geometries are simple [20]. Therefore, extension direction is specified by the q-vec- gral, the plastic area under the load (P)-
a three-dimensional model is analyzed in tor. In this model, it is defined with the start- load line displacement (VLL) curve, ob-
Abaqus. The loading pins (Figure 1b) are ing point at the crack tip and the end point at tained from fracture toughness test, needs
modeled as 3D analytical rigid shell bodies. the white dot in Figure 7. For a sharp crack to be calculated for each increase in the
The specimen is loaded by applying a dis- the strain field becomes singular at the crack value of displacement. Figure 9 shows the
placement to the pins in the vertical direc- tip. Specification of the midside node param- P-VLL curve used in the calculations.

54 (2012) 6
412 MATERIALS TESTING FOR JOINING TECHNOLOGY

When the crack in front of the nodes is


defined in the finite element modeling,
Abaqus, used for the analysis of J-integral,
can find the closed curve (contour) auto-
matically. Numerical tests suggest that the
estimate from the first ring of elements Figure 9. Load-load line
abutting the crack front does not provide a displacement curve of
high accuracy result, so at least two con- 6013 alloy
tours are recommended [22, 23]. Accord-
ingly, the J-integral for the first contour is
ignored in the analysis, because of numeri-
cal inaccuracies in the stresses and strains
at the crack tip [24]. In this study, 10 con-
tours were used, which is a general prac-
tice for the calculation of the J-integral.
J-∆a resistance curves are created on the
basis of results of the fifth contour taken as
representative contour.
As seen in Figure 10, the results ob-
tained from Abaqus are in very close con-
formance with the results computed using
formulation at Matlab. Figure 10. J-Δa curves
obtained of analysis and
During fracture toughness testing, load- formulation
CTOD, δ5 curves were obtained for various
crack extension values, an example is
shown in Figure 11. The crack tip opening
displacement has been observed to in-
crease in a stable manner and no sharp de-
creases in load values were seen.
In the load-CTOD, δ5 diagram (Figure 11)
the CTOD, δ5 values corresponding to crack
extension, ∆a are applied in the construc-
tion of Figure 12. The CTOD, δ5-Δa resist-
ance curve was obtained by using multi-
specimen method.
On the constructed C(T) model, by iden-
tifying the nodes 2.5 mm away from both Figure 11. Load- CTOD,
sides of the crack tip, which correspond to δ5 curve of 6013 alloy
the locations where the δ5 clip gage would
have been placed in a test, their displace-
ment values in the directon of the applied
load were recorded. The amount of this dis-
placement was compared with the CTOD,
δ5 values measured during the experi-
ments. It is seen from Figure 12 that the
measured CTOD, δ5 values are slightly
higher than those determined by the finite
element method. This might be due to some
stable crack extension which is not taken
into account in the analysis [14].
Either theoretical or experimental, since
all CTOD values are parameters used to Figure 12. CTOD, δ5-Δa
measure fracture toughness properties, resistance curves with
there must exist some relationship between respect to test and analy-
sis results
them. This allows to compare materials
characterized by a critical δ5 value with the
one chracterized by a critical δph value and
vice versa. Figure 13 shows the relationship
between the directly mea-sured δ5 values
against the calculated δph values. For small
loads corresponding to the intiation of the

54 (2012) 6
MATERIALS TESTING FOR JOINING TECHNOLOGY 413

crack extension regime, there is almost a This might arise by the fact that since δ5 is Since both, the J-integral and the CTOD,
linear reIationship between δ5 and δph, how- calculated at the tip of the fatigue pre-crack, can be used as fracture parameters in
ever, as the crack tip extends further this for long crack extension it might result in EPFM analysis, J-Δa and CTOD-Δa curves
relationship becomes nonlinear and as seen, greater values than those computed using must show a similar trend, which is con-
δph values become smaller than δ5 values. the plastic hinge model. firmed as seen in Figure 14. In this figure,
J-Δa and CTOD, δ5-Δa resistance curves are
obtained from the finite element analysis.
Figure 15 presents δ5 and δph against J/σy
relations. For small values of δ, the meas-
ured δ5 and the plastic hinge model δph give
approximate results, but the discrepancy
grows more as δ increases. dn parameters,
as determined from the slopes of δ-J/σy rela-
tions are also shown in the figure. The value
Figure 13. δph versus of dn calculated for J/σy values obtained
δ5 curve from formulation is 0.46, while for those ob-
tained from analysis it is 0.47, when δph is
considered. These two values are very simi-
lar, which confirms that results obtained
from Abaqus are in good agreement with
those computed using formulation. The dn
value for the directly measured δ5 concern-
ing J/σy is obtained as 1.09 by formulation
and 1.12 by analysis. The discrepancy be-
tween δ5 and δph may be attributed to the
measurement method of CTOD associated
with them. It is mentioned in ASTM E1290-
99 [25] that the plastic rotation factor rp is
not constant, rather it is a complex function
of specimen configuration and size, applied
loading and material [13]. These difficulties
Figure 14. Similarity may possibly alter the results of δ obtained
of R-curves by both methods. The variation of both
forms of δ against J/σy is linear. This nature
of variation is in good agreement with the
results of Kudari et al.[13] and Panontin et
al. [26]. Several other studies have men-
tioned that the parameter dn is between 0
and 1. On the other hand, for C(T) speci-
mens with a/W < 0.5 magnitudes of dn are
higher, and is found to be > 1 if the direct
measurement is used. This result shows
that relation between J and δph shows agree-
ment with literature, whereas no such
agreement can be found for J and δ5.
The value of dn as calculated from δph us-
ing J values obtained from analysis is 2.17 %
higher than the value of dn for J values ob-
tained by formulation, and the value of dn as
calculated from δ5 using J values obtained
Figure 15. δph , δ5 versus by analysis is 2.7 % higher than for J values
J/σy relation obtained by formulation. The main reason
of this discrepancy might be that δ5 is
measured away from the crack tip for large
crack extension.

Conclusions

1. The J-integral values obtained by formu-


lation and finite element analysis are
consistent. Moreover, experimental δ5

54 (2012) 6
414 MATERIALS TESTING FOR JOINING TECHNOLOGY

values are in close agreement with anal- Abstract


ysis results.
2. J-Δa and CTOD-Δa curves are in good Bestimmung der elastisch-plastischen Bruchmechanikparameter für
agreement, which allows them to be in-
eine Kompaktzugprobe mittels der Finite Elemente Methode. In der die-
terchangeable.
3. For small loads corresponding to the in- sem Beitrag zugrunde liegenden Studie wurde das bruchmechanische
tiation of the crack extension regime, Verhalten einer Aluminiumlegierung untersucht. Hierzu wurde das J-Inte-
the reIation between δ5 and δph is al- gral unter Verwendung einer Finite Elemente Software mit der entspre-
most linear, however, as the crack tip chenden Formulierung bestimmt. Der CTOD-Wert wurde mittels des plas-
extends further this relationship be-
comes nonlinear. This might be due to
tischen Hinge-Modells (δph) sowie anhand direkter Messung (δ5) ermittelt.
the fact that since δ5 is calculated at the Die J-Δa und CTOD-Δa Widerstandkurven wurden entwickelt sowie je-
tip of the fatigue pre-crack, for long weils mit den formulierten und mit denen aus Testdaten generierten ver-
crack extension it might result in glichen. Es konnte eine gute Übereinstimmung zwischen den Testergeb-
greater values than those computed us-
nissen, der Formulierung und den Ergebnissen der Finite Elemente Ana-
ing the plastic hinge model.
4. The relationship between J and δph, δ5 lysen erzielt werden. Das Verhältnis zwischen δph und δ5 ist nichtlinear
are linear and the linear proportionality und wenn der CTOD-Wert ansteigt, wird δ5 größer als δph. Das Verhältnis
constant, dn, obtained in this study for zwischen dem J-Integral und dem CTOD-Wert war sowohl für δph als auch
C(T) specimen with a/W = 0.5, consider-
für δ5 linear. Es konnte festgestellt werden, dass die Proportionalitätskon-
ing δph is found to be less than 1. This
results coincide with literature. stante dn bemerkenswert von dem Kalkulationsverfahren des CTOD-Wer-
5. It is seen from this study that the resist- tes abhängt.
ance curves for EPFM can successfully
support the results obtained from frac-
ture toughness testing. 6 J. Faleskog, X. Gao, C. F. Shih: Cell model for 16 Y. Shi, S. Sun, H. Murakawa, Y. Ueda: Finite
nonlinear fracture mechanics-I micromechanics element analysis on relationships between the
Acknowledgement calibration, Int. J. Fract. 89 (1998), pp. 335-373 J-integral and CTOD for stationary cracks in
7 R. Braun: Nd:YAG laser butt welding of welded tensile specimens, Int. J. Pres. Ves.
The experimental part of this study was AA6013 using silicon and magnesium contain- Pip. 75 (1998), pp. 197-202
ing filler powders, Mat. Sci. Eng. A 426 (2006), 17 T. L. Anderson: Fracture Mechanics: Funda-
carried out at Helmholtz-Zentrum
pp. 250-262 mentals and Applications, 2nd Edition, CRC
Geesthacht (formerly named as GKSS Re- 8 P. Nègre, D. Steglich, W. Brocks: Crack exten- Press, Texas (1991)
search Center), Germany. The authors sion at an interface: prediction of fracture 18 B. Omidvar, M. P. Wnuk, M. Choroszynski:
would like to thank the Institute of Materi- toughness and simulation of crack path devia- Relationship between the CTOD and the J-inte-
als Research for giving them the chance to tion, Int. J. Fracture 134 (2005), pp. 209-229 gral for stationary and growing cracks, closed-
carry out the experiments and especially to 9 Y. Huang, J. Chen, G. Liu, Yang H: A new form solutions, Int. J. Fracture 87 (1997), pp.
Dr. Mustafa Koçak, Head of the Department method of crack tip opening displacement 331-343
determined based on maximum crack opening 19 R. Knockaert, I. Doghri, Y. Marchal,
of Joining and Assessment.
displacement, Eng. Fract. Mech. 78 (2011), T. Pardoen, F. Delannay: Experimental and
pp. 1441-1451 numerical investigation of fracture in double-
References 10 U. Mühlich, J. R. Donoso, J. D. Landes: edge notched steel plates, Int. J. Fract. 81
A J-integral estimation method for C(T) speci- (1996), pp. 383-399
1 A. L. Lafly, D. Allehaux, F. Marie, C. Dalle,
mens using the common format equation, 20 G. X. Shan, O. Kolednik, H. P. Stüwe, F. D. Fischer:
H. Döker: Impact of FSW Techniques, Proc of
Int. J. Fracture 133 (2005), pp. 371-388 A substitution method for 3D elastic-plastic FE
the 58th Annual Assembly Conference of the 11 N. Gubeljak, M. D. Chapetti, J. Predan, analyses of fracture mechanics specimens,
International Institute of Welding (IIW), J. D. Landes: CTOD-R curve construction from Eng. Fract. Mech. 41 (1992), pp. 625-633
Prague (2005) surface displacement measurements, Eng. 21 Abaqus Technology Brief, TB-04-FMCAE-1,
2 A. Asserin-Lebert, J. Besson, A. F. Geurgues: Fract. Mech. 78 (2011), pp. 2286-2297 Fracture mechanics study of a compact ten-
Fracture of 6056 aluminum sheet materials: 12 Z. Taş, A. Polat: Determination of the fracture sion specimen using Abaqus/CAE (2007)
effect of specimen thickness and hardening toughness pipe steels using J-integral method, 22 C. Yang, W. Sun, J. S. Tomblin, S. S. Smeltzer:
behaivour on strain localization and tough- J. of CB Soma Vocational School of Tech. Sci. 1 A semi-analytical method for determining the
ness, Mat. Sci. Eng. A 395 (2005), pp. 186-194 (2007), pp. 7 strain energy release rate of cracks in adhe-
3 U. Çevik, U. Özdemir, Ç. Yeni, M. Koçak: 13 S. K. Kudari; K. G. Kodancha: On the relation- sively-bonded single-lap composite joints,
Investigation of laser beam welded 6056 ship between J-Integral and CTOD for CT and J. Compos. Mater. 41 (2007), pp. 1579-1602
aluminium alloys mechanical properties and SENB specimens, Frattura ed Integrita Strut- 23 C. Yang, A. Chadegani, J. S. Tomblin: Strain
fracture behaviour, Proc. of the XVI. National turale 6, (2008), pp. 3-6 energy release rate determination of pre-
Congress of Mechanics, Turkey (2009), 14 G. P. Nikishkov; J. Heerens; K.-H. Schwalbe: scribed cracks in adhesively-bonded single-lap
pp. 433-441 Transformation of CTOD δ5 to CTOD δ5BS and composite joints with thick bondlines,
4 P. Nègre, D. Steglich, W. Brocks, M. Koçak: J-integral for 3PB- and CT-specimens, Eng. Compos. Part B-Eng. 39 (2008), pp. 863-873
Numerical simulation of crack extension in Fract. Mech. 63 (1999), pp. 573-589 24 W. Brocks, I. Scheider: Numerical aspects
aluminium welds, Comp. Mater. Sci. 28 15 G. X. Shan, O. Kolednik, F. D. Fischer, of the path-dependence of the J-Integral in
(2003), pp. 723-731 H. P. Stüwe: A 2D model for numerical investi- incremental plasticity-how to calculate reliable
5 J. Ehrström, T. Warner: Metallurgical design gations of stable crack growth in thick smooth J-values in FE analyses, GKSS Forschungs-
of alloys for aerospace structures, Mater. Sci. fracture mechanics specimens, 45 (1993), zentrum Internal Report-GKSS/WMS/01/08
Forum 5 (2000), pp. 331-337 pp. 99-106 (2001).

54 (2012) 6
MATERIALS TESTING FOR JOINING TECHNOLOGY 415

25 ASTM E1290-99: Standard test method for σy yield stress of the material
crack tip opening displacement (CTOD) J J-integral
fractures toughness measurement a crack length
26 T. L. Panontin, A. Makino, J. F. Williams: W specimen with
Crack tip opening displacement estimation B specimen thickness
formulae for C(T) specimens, Eng. Fract. P load
Mech. 67 (2000), pp. 293-301 R radius of pin hole
Jel elastic component of J-integral
Nomenclature Jpl plastic component of J-integral
G energy release rate
δ, CTOD crack tip opening displacement C(T) compact specimen
Δa crack extension SENB single edge notch bend
a0 initial crack length 3PB three point bend specimen
δ5 CTOD measured across the fatigue crack tip at VLL load-line displacement
a gauge of 5 mm FEM finite element model
δph CTOD estimated by plastic hinge model EPFM elastic-plastic fracture mechanics
δi CTOD estimated by 90° intercept method LEFM linear elastic fracture mechanics
ν Poisson’s ratio
w strain energy density The Authors of This Contribution
Γ arbitrary closed contour around the crack tip
Ti components of the traction vector Uğur Özdemir studied Mechanical Engineering at
ui displacement vector components the Mustafa Kemal University, Turkey, in 2001.
ds length increment along the contour Γ He has carried out the experimental work of him
εij strain tensor M. Sc. thesis at GKSS Research Center, Germany,
σij stress tensor in diffusion bonding of TiAl alloys and earned his
nij unit vector normal M. Sc. from the Mustafa Kemal University, in
η dimensionless coefficient and for a C(T) type 2003. He is doing Ph.D in elastic-plastic fracture
specimen mechanics of laser welded materials at Dokuz
K stress intensity factor Eylül University, Turkey.
E' Young’s modulus E' = E (for plane stress), Çınar Yeni was born in 1967. She has carried
E out the experimental work of her M. Sc. and Ph. D.
E' = theses at Helmholtz-Zentrum Geesthacht (formerly
1 − ν2 named as GKSS Research Center), Germany, in
(for plane strain) fatigue and elastic-plastic fracture mechanics sub-
rp plastic rotation factor jects and also participated in projects on noval
Vp plastic displacement at the crack mouth welding methods during her post-doc studies. She
Apl plastic work is currently working as an associate professor at
Kl stress intensity factor Dokuz Eylül University, Turkey. Her main areas of
m constant in relation between J and δ interest are fatigue and fracture of welded materi-
dn constant in relation between J and δ als, namely laser and friction stir welding.

You will find the article and additional material by entering the document number MP110345
on our website at www.materialstesting.de

54 (2012) 6

View publication stats

You might also like