Engineering Fracture Mechanics 86 (2012) 48–55

Contents lists available at SciVerse ScienceDirect

Engineering Fracture Mechanics

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

Determination of CTOA in the molten material of spot welds using

the Digital Image Correlation technique
R. Lacroix a,⇑, A. Lens c, G. Kermouche b, J.M. Bergheau b, H. Klöcker a
a
École Nationale Supérieure des Mines de Saint-Étienne, SMS, UMR CNRS 5146, 158, cours Fauriel 42023 SAINT-ÉTIENNE Cedex 2, France
b
Université de Lyon, ENISE, LTDS, UMR 5513 CNRS/ECL/ENISE, 58 rue J. Parot 42023 Saint-Etienne Cedex 2, France
c
ArcelorMittal Maiziéres Research SA Voie Romaine 57280 Maiziéres-lés-Metz, France

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

Article history: In order to compare the strength of the molten material of different spot welds bonding
Received 19 July 2011 steel sheets, a specific wedge test has been developed. It produces stable crack growth at
Received in revised form 6 February 2012 the interface of sectioned spot welds. The deformation of the molten material at the pro-
Accepted 16 February 2012
gressing crack tip is observed in situ during wedge insertion. The Crack Tip Opening Angle
and the extend of the strain localization ahead of the crack tip are measured by Digital
Image Correlation. It is shown that the rather simple measurement of the Crack Tip Open-
Keywords:
ing Angle provides a good indication of the extend of the strain localization, which is
Stable crack growth
Crack Tip Opening Angle
believed to be related to the toughness of the molten material.
Strain localization Ó 2012 Elsevier Ltd. All rights reserved.
Spot weld

1. Introduction

Due to environmental issues and cost reduction, the today’s challenge in the transport industry is weight reduction. Thin
(about 2 mm thickness) steel sheets play a significant role in the automotive industry. These thin sheets are assembled by
spot welding. But, High Strength Steel welds may exhibit full interfacial failure (i.e. crack propagation through the molten
material at the interface between the welded sheets). Each vehicle contains several thousand welds. The safety analysis re-
quires thus the application of crack initiation and crack propagation criteria to structures of thin sheets assembled by
welding.
Several models for the analysis of crack propagation have been developed within the last three decades for homogeneous
materials. The critical Crack Tip Opening Angle (CTOA) was shown to be the most suited for modeling stable crack growth
and instability during the fracture process of homogeneous materials [1–6]. But, in thin sheet material, crack tunneling and
slanting will lead to non constant values of the CTOA [7,8]. Newman et al. [9] showed that the critical CTOA values should be
determined for various specimen configurations, crack lengths, and thicknesses. Unfortunately, spot welding of thin (about
2 mm thick) sheets leads to ‘‘small’’ volumes of the fusion zone, i.e. about 4  8  8 mm. Determining the critical CTOA values
for different specimen configurations and thicknesses is thus almost impossible.
Furthermore, the microstructure in the molten zone of a spot weld of steel is largely heterogeneous, as described by
Zhang and Senkara [10]. Indeed, dendritic growth of austenite occurs during the fast cooling steps of the successive pulses
of the spot welding process, finally resulting in micron-sized lenticular grains of martensite, along with a significant segre-
gation of chemical elements between the periphery and the center of the molten zone. Using classical Linear Elastic Fracture
Mechanics tools to characterize the toughness of such a heterogeneous material and such a non-standard geometry would
raise concerns regarding the validity of the hypothesis associated with these approaches.

⇑ Corresponding author. Tel.: +33 (0)1 48 39 57 62; fax: +33 (0)1 48 39 55 62.
E-mail address: remi.lacroix@mail.com (R. Lacroix).

0013-7944/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved.
doi:10.1016/j.engfracmech.2012.02.012
 R. Lacroix et al. / Engineering Fracture Mechanics 86 (2012) 48–55 49

Nomenclature

Variables
a initial crack length
dlow vertical displacement of the lower sheet measured at a distance l behind the current crack tip
dup displacement of the upper sheet measured at a distance l behind the current crack tip
F load applied to the wedge
L apparent length of the welded joint
l distance behind the current crack tip at which the CTOD is measured
u opening displacement
Deyy incremental vertical strain
DG difference in grey level at each pixel between two successive frames
DUy incremental vertical displacement

Acronyms
CTOA Crack Tip Opening Angle
CTOD Crack Tip Opening Displacement

The present paper thus aims at highlighting the major interest of the CTOA in characterizing the fracture resistance of the
molten zone of spot welds of thin sheets, with no particular hypothesis regarding their behavior. Although that methodology
does not provide a straight-forward fracture parameter, the authors propose it as a robust basis for comparison between var-
ious steel grades, and for further investigations aiming at quantifying fracture parameters when coupled other measurements.
In this paper the relation between the strain localization and the CTOA during stable crack growth through the Fusion
Zone material of spot welds is analyzed. In order to allow stable and reproducible crack propagation along the weld interface,
a specific wedge test has been developed (Fig. 1, [11]). Appropriate interpretations of the wedge test output are thus inves-
tigated to identify a reliable characterization of the resistance of the molten material to crack growth. In that purpose, this
paper intends to quantify the relevance of two measurements extracted from the in situ observation: the extent of the strain
localization, and the CTOA. The variation of the CTOA with the cracked surface will be related to the extent strain localization
at the crack front.
In a first part, the experimental set up of the wedge test is presented along with the procedure of sample preparation and
an example of observation. The second part shows a brief summary of existing approaches characterizing the resistance to
crack propagation from in situ observations. A third part is dedicated to the algorithm used to measure the in-plane displace-
ments by Digital Image Correlation. The set up of the CTOA measurement is then detailed in the fourth part, while the fifth
part presents the determination of the strain localization. A relation between the strain localization and the CTOA is finally
highlighted in the sixth part, presenting two examples of measurements results. Finally, a discussion of the relevance of
these measurements to characterize the resistance to crack growth is proposed.

2. The wedge test

Various wedge tests has been proposed in the literature to characterize the resistance to interfacial crack growth, for
example in adhesives [12], or in steel brazed joints [13]. The load is applied by inserting a wedge at the interface of an assem-
bly, implying an opening displacement where the wedge is in contact with the sample, eventually leading to a progressive
failure of the interface. It stands out as a very stiff system, thus favoring stable rather than unstable regimes of crack
growth. This part details the preparation of the samples from spot welded sheets, the experimental set up of the wedge test
developed for this geometry of samples, and an example of observation.

Fig. 1. Schematic set up of the wedge test and Crack Tip Opening Angle, defined as the opening angle required for crack growth, measured at a given
distance l behind the current crack tip.
50 R. Lacroix et al. / Engineering Fracture Mechanics 86 (2012) 48–55

2.1. Sample preparation

An advanced high strength steel have been investigated in this study. 2 mm thick sheets of Dual Phase 780 steel have
been assembled by resistance spot welding at 10 kA intensity with 8 mm electrodes, forming welds of about 9 mm diameter.
Micro-hardness of the base metal, respectively the molten zone, are measured at 270 ± 30HV0.05, respectively
450 ± 30HV0.05. The microstructure of the fusion zone was characterized in previous work by SEM and EBSD observations
[14] The method presented in present work applies to strongly heterogeneous materials.
An assembly of two stacked coupons welded in their center, as shown in Fig. 2, is used to form two samples for the wedge
test. One sample is obtained by sectioning the welded coupons as illustrated by the Fig. 2. This operation implies that the
periphery of the molten zone will only be considered. The area of the remaining interface between the sheets is about
one fifth of the initial spot weld’s interface. Sectioning aims at reducing the dimensions of the interface relatively to those
of the sheets, in order to form a relatively weak bond between the sheets of the sample. Although the entire interface can
obviously not be characterized by such a procedure, this happens to be a required compromise in order to allow reproducible
interfacial failures, i.e. through the remaining molten material.
An asymmetric sectioning is performed, leaving two different arm length as apparent on the observed surface. The length
of the long arm is about 25 mm, while the short one is about 25 mm long, and defines the initial crack length a. The apparent
length of the welded joint L is about 7 mm (Fig. 3a).
The sectioned surface thus exhibits two notches delimiting the bond between the sheets. This surface will be observed
during the progressive failure of the interface (Fig. 3b). In order to quantify the evolution of the in-plane displacements
on this surface, a speckle-like pattern with an average particle diameter of 25 lm is generated by sand blasting. This method
has been preferred to paint-based methods for its simplicity and the reproducibility of the generated pattern.

2.2. Wedge test set up and mechanical loading

Stable crack growth through the molten material is achieved by imposing a progressive opening displacement u on one
side of the sample, while the opposite side is clamped. The Fig. 3a shows the experimental set up of one sample in the

Fig. 2. Examples of spot welds samples. The sectioning of the welded coupons is illustrated by the solid line.

(a) (b)
Fig. 3. (a) Experimental set up of the wedge test. (b) Upper face of the specimen and region of in situ observation.
 R. Lacroix et al. / Engineering Fracture Mechanics 86 (2012) 48–55 51

(a) (b)
Fig. 4. Observation of the sectioned surface. (a) Example of recorded frame. (b) Illustration of the grey level difference between this frame and the
successive one.

machine. The opening displacement is applied by inserting a 60°-wedge in the sample’s interfacial plane, at a speed of
0.02 mm/s. The clamping of the sample is concentrated on the edges opposite to the wedge, thus approaching a hinge-type
link with the frame of the machine, and allowing the opening of the sample without dissipation of energy.
The load F applied to the wedge are recorded by a strain gage based load cell, made in-lab, of capacity of 10 kN and max-
imum error to linearity of 30 N, and the wedge displacement is measured by a LVDT sensor of capacity 25 mm and maximum
error to linearity of ±0.2%.

2.3. Optical tracking of the crack tip

The experimental set up presented above allows the observation of the sectioned surface of the sample during loading. A
CCD camera of 1280  1024 pixels records the observation of the region neighboring the notch on the side of the wedge at a
rate of 2 images per second at a 8-bit depth. A 12:5 optical lens is used to reduce the observation on a region of 5  4 mm2, as
shown in Fig. 4a. A typical test record presents about 200 frames between the start of the wedge load and the final unstable
failure.
Manual processing of each couple of successive recorded frames is carried out to locate the crack tip during the entire test.
In that purpose the difference in grey level at each pixel between two successive frames, or DG, is considered to determine
the current position of the crack tip. The Fig. 4b illustrate the image for which the local grey level I results from the following
relation: I = 255 10  DG. The current crack tip is defined as the point at the extreme left of the ‘V-shaped’ region, formed
by the pixels exhibiting a significant difference in grey level between two successive frames, as illustrated in Fig. 4b. This
definition visually matches the tip of the dark zone identified as the separated region behind the crack tip, as illustrated
in Fig. 4a. Moreover, it has been observed that the crack tip does not necessarily follow one line of this ‘V-shaped’ zone,
but rather progresses in between these two lines. Accuracy of this manual processing is estimated at about ±5 pixels, or
±20 lm.

3. In situ characterization of the resistance to crack growth

The major interest to characterize the resistance to crack growth from optical in situ observations is that additional data
can be extracted from a fracture test, possibly providing different insight to the material resistance than the classical load
and displacement records.
Such approaches are based on fields measurements in the region of the crack tip. Most of the studies in the literature are
limited to surface measurements, thus limiting their validity in the case of minor variations between surface and in-bulk
crack growth, as small scale yielding for example.
Various authors have proposed methods to characterize the resistance to crack growth based on optical observations. The
measurement of the Crack Tip Opening Angle (CTOA) as proposed by Sutton [7], Newman [9] or Heerens [8] has proven to be
a simple and reproducible characterization under various kind of loading. Hild and Réthoré [15] have proposed to extract
stress intensity factors from the measured displacement fields in the crack tip region. Strain localization in the region of
the crack tip is the main factor limiting the application of these latter approaches in cases where small scale yielding is
not verified, and thus exhibiting strain fields different from the theoretical K-based fields.
In the case of complex geometries, such as interfacial failures of spot welds, interpretation of measurement of the CTOA
can prove to be difficult. It is proposed in this paper to investigate the extend of the strain localization during crack growth as
an indicator of the encountered resistance to crack growth. Moreover, this latter measurement and the CTOA are compared
in order to determine whether or not an relation appears between these two measurements during crack growth through the
molten zones of spot welds.
The method used in the works presented in this paper is the Digital Image Correlation investigating in-plane displace-
ments on the observed surface. Details of the implemented method are provided in the first part. The set up of the CTOA
52 R. Lacroix et al. / Engineering Fracture Mechanics 86 (2012) 48–55

measurement is then presented in the second part. An estimation of the incremental strain field is proposed in the third part,
and the extend of the strain localization is defined in this study by a threshold applied on the vertical incremental strain field.
Results of these measurements for two different samples are then presented and discussed.

3.1. Digital Image Correlation

Digital Image Correlation has been used to quantify both the Crack Tip Opening Angle and the extend of the strain local-
ization. A subset-based approach has been implemented in-lab, computing the correlation product between subsets by mean
of Fast Fourier Transformation (FFT). Subpixel accuracy is obtained by bi-cubic interpolation of the peak of the correlation
product, and by FFT subpixel shift of the subsets [16]. Squared rigid subset of 64  64 pixels are used, and the measurement
accuracy has been estimated at 0.015 pixel using the experimental images recorded from spot weld samples.
A discretization of the experimental displacement field between two frames is thus obtained by computing the rigid dis-
placements of a grid of overlapping subsets, regularly spaced with a period of 5 pixels, forming a grid of 243  192 nodes.
This operation is performed between each successive frames, and the Fig. 5b illustrates the resulting discretization of the
incremental vertical displacement field DUy.

3.2. Crack Tip Opening Angle Measurements

The current position of the crack tip is manually determined as presented in part Section 2.3 for each frame recorded dur-
ing crack growth. The Crack Tip Opening Displacement (CTOD) is determined from two points located at a distance l = 2 mm
behind the crack tip, and symmetrically located on the upper and lower side of the crack, at a distance l from each other (red
crosses on the Fig. 6a). The displacement of these points is then tracked until the beginning of the frame sequence (blue cir-
cles on the Fig. 6a), thus defining the CTOD as the sum of the vertical displacements of these two points CTOD = dup + dlow.
The size of the image subset used to track the upper and lower points is displayed as a black square surrounding the red
crosses on the Fig. 6a.
The current CTOA is given by CTOA = tan 1(CTOD/l), and can be determined during crack growth, as illustrated by the
Fig. 6b.

3.3. Strain localization at the crack tip

Vertical displacements fields are measured as described in part Section 3.1 between each pair of successive frames, thus
quantifying incremental displacements. A linear approximation of the vertical incremental strain field is given by the follow-
ing relation between each node of measurement of the incremental displacement: Deyy = @ DUy/@y.
Smoothing is performed on the resulting data, by associating the incremental strain value Deyy on one node to the average
of the 7  7 neighboring nodes. The resulting incremental vertical strain field is illustrated in Fig. 7b, for values ranging be-
tween 0.002 and 0.008.
It appears that the resulting incremental strain field presents an overall similar geometry than the zone of significant
change in grey-level, as illustrates the comparison between Fig. 7a and b. The main difference between these two zones
is their thickness, which is very thin in the case of the modified grey-level zone. Indeed, the method of determination of
the strain field considers the displacement field as continuous, and averages the resulting strain field, thus spreading the ac-
tual discontinuity in a rather thicker zone. However, it is proposed that this simple measurement provides relevant data to
estimate the extend of strain localization at the crack tip.
The extend of the strain localization associated to one pair of successive frames is defined by a threshold operation on the
Deyy field. In the current parameters of measurement, it has been arbitrarily chosen in this study that the extremity of the

(a) (b)
Fig. 5. (a) Recorded frame. (b) Measured incremental vertical displacement field DUy in pixel.
 R. Lacroix et al. / Engineering Fracture Mechanics 86 (2012) 48–55 53

14
Initial
Current 12
10

CTOA (deg)
8
6
4
2
1mm 0
0 0.5 1 1.5 2 2.5 3 3.5
Crack advance (mm)
(a) (b)
Fig. 6. (a) Determination of the current Crack Tip Opening Displacement (CTOD), the current crack tip is located by the star, the CTOD is measured at the
points located by the red crosses, using image subsets as displayed by the black squares. The resulting positions of these points on the initial frame are
illustrated by the blue circles. (b) Example of evolution of CTOA as a function of crack advance. (For interpretation of the references to color in this figure
legend, the reader is referred to the web version of this article.)

Fig. 7. (a) Image resulting from the processing (255 10  DG), with DG being the grey-level difference between two successive frames, highlighting the
zone of significant grey-level changes. (b) Strain localization Deyy between these frames.

strain localization zone is defined by the furthest node from the crack tip exhibiting a value Deyy = 0.002. Therefore, the ex-
tend of the strain localization is given by the distance between the crack tip and the latter node (Fig. 7b).

4. Results

4.1. Stable interfacial failure during wedge load and displacement

An example of wedge test on a spot weld of DP780 steel is presented in this part. The recorded wedge load and displace-
ment on the wedge during stable crack growth are illustrated by the Fig. 8a. The fractured portion of the interfacial plane is
revealed on interrupted tests by oxidation of two different samples A and B, respectively interrupted for a level of wedge
displacement equivalent to those represented on the load–displacement curve. The Fig. 8b shows the interfacial plane of
the two antagonist sheets after complete fracture of these two samples. The oxidized zone appears in light blue, and high-
lights the portion of the interface fractured at the test interruption.
The current crack advance as defined on the previous part can be illustrated as a function of the wedge displacement, as
shown in Fig. 8a.

4.2. CTOA and strain localization

The measurements of CTOA and the extend of the strain localization have been performed on two different samples for
one spot weld of DP780 steel. Different positions of the sectioning result in different sizes of the molten zone taken from that
spot weld for each of these two samples. The Fig. 9a shows the simultaneous evolutions of the CTOA and the extend of the
strain localization during crack growth for the smaller sample, while Fig. 9b illustrates these evolutions for the larger sample.
It appears first that these two samples exhibit different behaviors during crack growth. However, during each of these
test, the evolutions of CTOA and the extend of the strain localization show an apparent consistency. Indeed, while the CTOA
54 R. Lacroix et al. / Engineering Fracture Mechanics 86 (2012) 48–55

(a) (b)
Fig. 8. (a) Example of recorded load, displacement and crack advance. (b) Associated spot weld interface for two equivalent samples illustrating two
instants of the test. The solid line shows the current crack front.

(a) (b)
Fig. 9. Evolution of the Crack Tip Opening Angle and of the extend of the strain localization as a function of the crack advance. (a) Smaller sample. (b) Larger
sample.

increases continuously for the smaller sample, the extend of the strain localization grows as well almost continuously. In the
case of the larger sample, the CTOA presents a plateau during crack advance, and the extend of the measured strain local-
ization shows a constant, although scattered, evolution.

5. Discussions

The observed consistency between the two measurements, illustrated in Fig. 9a and b, yields the following proposition:
the CTOA measurement is able to indicate the evolution of the amount of strain localization ahead of the crack tip.
Based on the incremental description of the resistance to crack growth by the energy dissipation rate [17,18], it is more-
over assumed that the amount of incremental strain localization ahead of the crack tip provides a relevant characterization
of the resistance to crack propagation.
The apparent relation between the CTOA and the extend of the strain localization is therefore consistent with the relation
proposed by Li and Turner [19], quantifying the amount of incremental volumetric energy dissipation ahead the tip of a
propagating crack by the CTOA multiplied by the local yield stress.
This proposed relation could thus be used to explain the proposed limitations of the CTOA measurements for C(T) sam-
ples. Heerens et al. [8] specify that CTOA measurement must not be considered when the crack length exceeds amax, a given
portion of the uncracked ligament. The observed increase of CTOA beyond this limit can thus be associated with the fact that
the plastic zone ahead of the crack tip starts to encompass the whole remaining ligament, and thus to increase as well.

6. Conclusion

The advances proposed in this paper can be summarized as follow:

1. Stable crack growth through the molten material of spot welds have been achieved, providing a new insight on their frac-
ture behavior. However, the complex geometry required by this experiment and the observed strain localization ahead of
the crack tip invalidate the use of the classical linear elastic fracture mechanics to interpret the results.
 R. Lacroix et al. / Engineering Fracture Mechanics 86 (2012) 48–55 55

2. An experimental evidence of the relevance of the Crack Tip Opening Angle in the characterization of the extend of the
strain localization is presented, in the case of a complex geometry, heterogeneous material and large scale yielding.
3. The extend of the strain localization is interpreted as an indicator of the resistance to crack propagation, and thus of the
toughness of the molten material.

As a conclusion, the measurement of the Crack Tip Opening Angle appears to be a simple and flexible method to quantify
the resistance to crack growth, even in the case of the complex geometry of spot welds’ molten material.

References

[1] de Koning AU. A contribution to the analysis of slow stable crack growth. National Aerospace Laboratory Report NLR MP75035U; 1975.
[2] Shih CF, G de Lorenzi H, Andrews WR. Studies on crack initiation and stable crack growth. ASTM STP 1979;668:65–120.
[3] Kanninen MF, Rybicki EF, Stonesifer RB, Broek David, Rosenfield AR, Nalin GT. Elastic–plastic fracture mechanics for two dimensional stable crack
growth and instability problems. ASTM STP 1979;668:121–50.
[4] Newman JC. An elastic–plastic finite element analysis of crack initiation, stable crack growth, and instability. ASTM STP 1984;833:93–117.
[5] Brocks W, Yuan H. Numerical studies on stable crack growth. Defect assessment in components fundamentals and applications, vol. 9. ESIS Publication;
1991. p. 19–33.
[6] Newman JC, Shivakumar KN, McCabe DE. Finite element fracture simulation of A533B steel sheet specimens. Defect assessment in components
fundamentals and applications, vol. 9. ESIS Publication Edition; 1991. p. 117–26.
[7] Sutton Michael A, Helm Jeffrey D, Boone Michael L. Experimental study of crack growth in thin sheet 2024-T3 aluminum under tension–torsion
loading. Int J Fract 2001;109(3):285–301.
[8] Heerens J, Schödel M. Characterization of stable crack extension in aluminium sheet material using the crack tip opening angle determined optically
and by the [delta]5 clip gauge technique. Eng Fract Mech 2009;76(1):101–13.
[9] Newman JC, James MA, Zerbst U. A review of the CTOA/CTOD fracture criterion. Eng Fract Mech 2003;70(3–4):371–85.
[10] Zhang H, Senkara J. Resistance welding: fundamentals and applications. CRC Press; 2006. IBSN 9780849323461.
[11] Lacroix R, Monatte J, Lens A, Kermouche G, Bergheau JM, Klöcker H. Spot weld strength determination using the wedge test, in situ observations and
coupled simulations. Appl Mech Mater 2010;24–25:299–304.
[12] Cognard J. Use of the wedge test to estimate the lifetime of an adhesive joint in an aggressive environment. Int J Adhes Adhes 1986;6(4):215–20.
[13] Philips NR, He MY, Evans AG. A wedge fracture toughness test for intermediate toughness materials: application to brazed joints. Acta Mater
2008;56:4593. 4600.
[14] Dancette M, Dupuy T, Lens A, Nait Oultit B, Massardier V, Fabrègue D, et al. An investigation of failure types in high-strengh steel resistance spot. Weld
World 2010;54(3–4):1–14.
[15] Réthoré Julien, Roux Stéphane, Hild François. Noise-robust stress intensity factor determination from kinematic field measurements. Eng Fract Mech
2008;75(13):3763–81.
[16] Grédiac Michel, Hild François. Mesures de champs et identification en mécanique des solides. Hermes Science Publications; 2011.
[17] Turner CE. A re-assessment of ductile tearing resistance, part: I. The geometry dependence of J–R curves in fully plastic bending. Part II: Energy
dissipation rate and associate r curves on normalized axes. In: Firraro D, editor. Fracture behaviour and design of materials and structures, vol. 2.
Warley; 1990.
[18] Siegmund T, Brocks W. A numerical study on the correlation between the work of separation and the dissipation rate in ductile fracture. Eng Fract
Mech 2000;67(2):139. 154.
[19] Li ZF, Turner CE. Crack-opening angle and dissipation-rate analysis of r-curves for side-grooved pieces of HY130 steel in bending. J Mater Sci
1993;28(21):5922–30.