Computational Materials Science 44 (2008) 489–495

Contents lists available at ScienceDirect

Computational Materials Science

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

Fatigue crack growth of a double fillet weld

M. Benachour a, M. Benguediab b,*, A. Hadjoui a, F. Hadjoui a, N. Benachour a
a
Automatic Laboratory, Faculty of Engineering Sciences, University of Tlemcen, Algeria
b
Faculty of Engineering Sciences, University of SidiBel Abbes, Algeria

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

Article history: The welded structures have a broad applicability, steel constructions, car industry, aeronautical, marine,
Received 23 October 2007 pipelines, etc. These structures are generally subjected to cyclic requests. A simple existing defect after
Received in revised form 1 April 2008 welding can generate a catastrophic fracture. This work studies the fatigue crack growth of a double fillet
Accepted 12 April 2008
weld with the existence of a semi-elliptical crack. Two types of aluminum alloys are studied with know-
Available online 5 June 2008
ing the alloy 2024 T351 and the 7075 T6. Crack growth analysis uses linear elastic fracture mechanics and
related crack growth material properties to determine how fast a crack or crack-like defect will grow.
PACS:
Fracture mechanics is based on the concept of stress intensity (K) that describes the magnitude of both
47.11.Fg
62.20.Mk
the stress and strain fields around a crack. It is computed from the stress range (Dr), and crack size (a)
and crack shape (b). The effect on the fatigue life of the geometrical parameters of the crack (a/c ratio), the
Keywords: angle of inclination of the weld bead and the level of loading are studied. In order to predict the fatigue
Fatigue behavior of the welded structure, a constant amplitude loading is applied where the influence of the load
Weld ratio over the fatigue life is presented. A comparative study of fatigue crack growth of the cited aluminum
Semi-elliptical crack alloys are detailed in order to show the effect of several parameters.
Aluminum alloys Ó 2008 Elsevier B.V. All rights reserved.
Load ratio
Aluminum alloy

1. Introduction
In all cases, welding is the primary jointing method and fatigue
is a major design criterion. Welding may be replaced by some
alternative joining method with higher fatigue strength, e.g. adhe-
sive bonding is largely useful [1]. There is growing interest in the
structural use of aluminum alloy, for such applications as aero-
space, aeronautic, automotive, marine applications, etc [2,3]. The
welding of aluminum and its alloys provides a basic understanding
of the metallurgical principles involved in the way that alloy
achieve their strength and how welding can affect their properties
[4].
Two types of cracking will normally cause failure of a fillet
welded joint. They are root cracking and toe cracking [5]. The fail-
ure analysis of weldment suggests that fatigue alone is considered
to account for most of the disruptive failures [6].
Fatigue assessment procedures for welded aluminum structures
presented in reviewed work [7] shown that the cumulative damage
under realistic stress affect the fatigue limit. In discussed work of
Swellam et al. [8], the weld fatigue failure process can be divided
into three stages such as crack initiation and growth as kinked
cracks emanating from the main crack, crack propagation through
* Corresponding author.
E-mail address: benguediab_m@yahoo.fr (M. Benguediab).
the sheet thickness and crack propagation through the width of
specimens.
Fatigue crack growth behaviour of welded joints depends on the
material, loading and in particular, the geometric configurations of
the weld and plate thickness [9,10]. The effect of butt weld geom-
etry parameters (weld toe, flank angle, plate thickness, initial crack
geometry) on the fatigue crack propagation life have been studied
[11] by using Linear Elastic Fracture Mechanics (LEFM), Finite Ele-
ment Analysis (FEA) and superposition approaches. A criterion has
been established to forecast the possible crack initiation region
(toe or root) in the double fillet welded load carrying cruciform
joint [12].
The fatigue behavior of 2024-T3 friction stir welded overlap
joints is characterized by Fersini and Pirondi [13] where in the
welded joint two crack-like unwelded zones are present at overlap
ends. In the Fersini and Pirondi study, the stress intensity factor at
the crack tip and the fatigue crack path have been studied using
the FE code Franc2d and the lifetime has been estimated by inte-
grating the material propagation law with the software AFGROW.
The fatigue crack growth of welded aluminium alloy 2024 T351
was studied under effect of several parameters [14] (residual
stress, weld orientation, etc.).
The aim of this work is the study of fatigue crack growth of a
double fillet weld with the existence of a semi-elliptical default
where two aluminum alloys are considered.

0927-0256/$ - see front matter Ó 2008 Elsevier B.V. All rights reserved.
doi:10.1016/j.commatsci.2008.04.015
490 M. Benachour et al. / Computational Materials Science 44 (2008) 489–495

2. Fatigue crack growth of double fillet weld Semi-elliptical surface cracks occur frequently at the weld toes
of welded joints [15]. Fatigue crack growth rate is usually analyzed
The evaluation of the fatigue life of welded structures is compli- in terms of fracture mechanics by using relations involving stress
cated by large variations in weld geometry, welding defects, resid- intensity factor.
ual stress, etc. The crack propagation is the dominant part of the The fatigue process under constant amplitude loading can be
fatigue life. Different regions of fatigue crack growth are shown described by the following equation:
in Fig. 1.
da
¼ f ðDK; RÞ ð1Þ
dN
where DK is stress intensity factor and R is stress ratio.
The most frequently used model to describe the fatigue crack
propagation is Paris’ Law [16]

da
¼ C DK m ð2Þ
dN
where C and m are material constants.
Others models are proposed to describe the total curve of fati-
gue crack growth and effects of different parameters [17–19].
In order to analyze the fatigue behaviour in welded joints using
fracture mechanics techniques, it is necessary to calculate the va-
lue of the stress intensity factor of the fatigue crack.
Fig. 1. Schematisation of fatigue crack growth. It is known that the stress intensity factor (K) for crack opening
mode (mode I) can be expressed in the following form:
pffiffiffiffiffiffiffiffiffiffi
K ¼ br pa ð3Þ
A correctional factor Mk was introduced for considering the ef-
fect of welding geometry for the cracks which propagate in the re-
gion of stress concentrations produced by the geometry of welded
joints, i.e. cracks at weld toes, a further correction factor (Mk) is
introduced and known as geometry magnification factor [20]. Then
Eq. (3) can be written as follows:
pffiffiffiffiffiffiffiffiffiffi
K ¼ b0  Mk  r  pa ð4Þ
The geometry of a double fillet weld is shown in Fig. 2.
To predict the effect of several parameters on fatigue crack
growth, a fatigue calculator code is used. Crack growth analysis re-
quires a crack growth curve for the material. The linear portion of
the curve represents stable crack growth and is characterized by an
Fig. 2. Semi-elliptical crack in double fillet weld. intercept (C) and slope (m).
For the considered materials [21,22], 2024 T351 and 7075 T6
the characteristic materials and coefficient C and m are presented
Table 1
Characteristic of materials in Table 1.
pffiffiffiffiffi The number of cycles required to propagate a crack from an ini-
Materials E (GPa) DK th ðMPa mÞ C m
tial crack size a0 = 0.1 mm, to a final crack af = 10 mm can be calcu-
2024 T351 71.02 2.0 4.34  1011 3.45 lated by using the following equation when numerical integration
7075 T6 70.90 2.3 1.10  1010 3.45
is applied:

1.2E-02
Crack length a (m)

1.0E-02

8.0E-03

6.0E-03

4.0E-03

2.0E-03

0.0E+00
0.0E+00 2.0E+04 4.0E+04 6.0E+04 8.0E+04 1.0E+05 1.2E+05
Number of cycle NR

Fig. 3. Fatigue crack growth for h = 30°.

 M. Benachour et al. / Computational Materials Science 44 (2008) 489–495 491

1.2E-02

Crack length a (m)

1.0E-02

8.0E-03

6.0E-03

4.0E-03

2.0E-03

0.0E+00
0.0E+00 1.0E+04 2.0E+04 3.0E+04 4.0E+04 5.0E+04

Number of Cycle NR

Fig. 4. Fatigue crack growth for h = 45°.

1.2E-02
Crack length a (m)

1.0E-02

8.0E-03

6.0E-03

4.0E-03

2.0E-03

0.0E+00
0.0E+00 5.0E+03 1.0E+04 1.5E+04 2.0E+04 2.5E+04 3.0E+04
Number of cycle NR

Fig. 5. Fatigue crack growth for h = 60°.

Z ai
da tio a/c of the crack over the crack growth and the total life. For h
Ntot ¼ ð5Þ
af C  DK m equal 30°, there is intersection of the curves for weak ratios a/c.
Then for angle h equal 45° and 60°, one notices bearings where ad-
vanced crack is delayed.
2.1. Fatigue crack growth for 2024 T351 The effect of angle h is clearly shown in Fig. 6 when a/c = 0.1.
The increasing of this angle decreases the fatigue life. For different
For fatigue crack growth the effect of the ratio a/c of crack is ratios a/c, the effect of the angle h on the fatigue life is shown in
examined. Figs. 3–5 show the effect of the increase of the initial ra- Figs. 3–5.

1.2E-02

1.0E-02
Crack length a (m)

8.0E-03

6.0E-03

4.0E-03

2.0E-03

0.0E+00
0.0E+00 5.0E+03 1.0E+04 1.5E+04 2.0E+04 2.5E+04
Number of cycle NR

Fig. 6. Effect of angle h for a/c = 0.1 on fatigue crack growth for 2024 T351.
492 M. Benachour et al. / Computational Materials Science 44 (2008) 489–495

The effect of stress ratio on the fatigue behavior of material
weldment is investigated [23]. For two aluminum alloys 2024
T851 and 7475 T7351, the effect of stress ratio on fatigue crack
growth is examined and the shift of the propagation curves is
shown [24].
The fatigue crack growth equation should contain in principle,
at least two loading parameters, the stress amplitude (Dr) and
the stress ratio (R) (see Eq. (1), so that the fatigue loading is prop-
erly defined. The effect of the stress ratio on fatigue life is shown in
Figs. 7 and 8. As the stress ratio increases, the fatigue life increases.

1.2E-02

1.0E-02
Crack length a (m)

8.0E-03

6.0E-03

4.0E-03

2.0E-03

0.0E+00
0.0E+00 5.0E+04 1.0E+05 1.5E+05 2.0E+05 2.5E+05
Number of cycle NR

Fig. 7. Fatigue crack growth curves 2024 T351 at rmin = 20 MPa.

1.2E-02
Crack length a (m)

1.0E-02

8.0E-03

6.0E-03

4.0E-03

2.0E-03

0.0E+00
0.0E+00 2.0E+04 4.0E+04 6.0E+04 8.0E+04 1.0E+05
Number of cycle NR

Fig. 8. Fatigue crack growth curves for 2024 T351 at rmax = 160 MPa.

1.2E-02
Crack length a (m)

1.0E-02

8.0E-03

6.0E-03

4.0E-03

2.0E-03

0.0E+00
0.0E+00 5.0E+03 1.0E+04 1.5E+04 2.0E+04
Number of cycle NR

Fig. 9. Fatigue crack growth for h = 45°.

 M. Benachour et al. / Computational Materials Science 44 (2008) 489–495 493

1.2E-02

1.0E-02

Crakc length a (m)

8.0E-03

6.0E-03

4.0E-03

2.0E-03

0.0E+00
0.0E+00 2.0E+03 4.0E+03 6.0E+03 8.0E+03 1.0E+04
Number of cycle NR

Fig. 10. Effect of angle h for a/c = 0.1 on fatigue crack growth for 7050 T6.

1.2E-02

1.0E-02
Crack length a (m)

8.0E-03

6.0E-03

4.0E-03

2.0E-03

0.0E+00
0.0E+00 5.0E+03 1.0E+04 1.5E+04 2.0E+04 2.5E+04
Number of cycle NR

Fig. 11. Fatigue crack growth curves for a/c = 0.1 and h = 30°.

1.2E-02

1.0E-02
Crack length a (m)

8.0E-03

6.0E-03

4.0E-03

2.0E-03

0.0E+00
0.0E+00 2.0E+03 4.0E+03 6.0E+03 8.0E+03 1.0E+04
Number of cycle NR

Fig. 12. Fatigue crack growth curves for a/c = 0.1 and h = 45°.
494 M. Benachour et al. / Computational Materials Science 44 (2008) 489–495

1.2E-02

1.0E-02

Crack len gth a (m)

8.0E-03

6.0E-03

4.0E-03

2.0E-03

0.0E+00
0.0E+00 1.0E+03 2.0E+03 3.0E+03 4.0E+03 5.0E+03 6.0E+03

Number of cycle NR
Fig. 13. Fatigue crack growth curves for a/c = 0.1 and h = 60°.

1.2E-02

1.0E-02
Crack length a (m)

8.0E-03

6.0E-03

4.0E-03

2.0E-03

0.0E+00
0.0E+00 5.0E+03 1.0E+04 1.5E+04 2.0E+04 2.5E+04 3.0E+04
Number of cycles NR

Fig. 14. Fatigue crack growth curves for a/c = 0.4 and h = 45°.

Theses results are in agreement with the results of Srivastava and
Garg [25] but disagreement with other work [26].
For constant minimal stress (Fig. 7), h = 45° and a/c = 0.2, the
increasing of maximum stress (rmax), increase and shift the fatigue
crack curve.
Of same for constant maximal stress (Fig. 8.) the decreasing of
minimal stress (rmin) increases the fatigue life.
2.2. Fatigue crack growth for 7075 T6
For this material, the ratio a/c of crack is examined. The initial
ratio a/c of the crack affects the total life of fatigue crack growth
(Fig. 9). The increasing of the angle h (Fig. 10) decreases the total
fatigue life when the shift is to the right-hand side towards the left.
2.3. Fatigue crack growth comparison for 2024 T351 and 7075 T6
The comparison of fatigue crack growth of the considered mate-
rials is presented (Figs. 11–14) for a/c = 0.1 and variable angle. For
the material 2024 T351 the total fatigue life is higher compared to
the material 7075 T6 under the same loading. The increasing of an-
gle h increase the fatigue crack growth (Figs. 11–13).
Of the same, for other ratio a/c (a/c = 0.4 and h = 45°), the total
life for two materials is affected (Fig. 14). One always notices an in-
crease in the total fatigue life for two materials comparatively to
the change of the initial state of crack and the angle of the weld
joint.
3. Conclusion
The paper presented the fatigue crack growth of a double fillet
weld when a semi-elliptical crack exists. Two aluminum alloys are
studied such as the alloy 2024 T351 and the 7075 T6.
The initial ratio (a/c) of semi-elliptical crack affects the total fa-
tigue life. The increasing of this ratio, increase the fatigue life. The
angle orientation of the weld joint effect the fatigue crack growth.
Crack growth data show the influence of R. For constant amplitude
loading (e.g rmax or rmin), the increasing of the stress ratio R, in-
crease the total fatigue life.
The fatigue crack growth for aluminum alloy 2024 T351 and
7075 T6 is affected by the variation of loading parameter, weld
geometry and initial default.
References
[1] M.L. Sharp, G.E. Nordmark, G.C. Mensemer, Fatigue Design of Aluminum
Components and Structures, McGraw Hill, New York, 1996.
[2] P. Simoncini, Behaviour of welded aluminium alloys in relationship with
seawater Corrosion. 2nd International Symposium on Aluminum Surface
 M. Benachour et al. / Computational Materials Science 44 (2008) 489–495 495

Science and Technology (ASST 2000). 21–25, May 2000, UMIST, Manchester, [13] D. Fersini, A. Pirondi, Eng. Fract. Mech. 74 (2007) 468–480.
England, UK. [14] G. Bussu, P.E. Irving, Int. J. Fatigue 25 (2003) 77–88.
[3] W. Fricke, Mar. Struct. 16 (2003) 185–200. [15] P.S. Song, B.C. Sheu, Y.L. Shieh, Int. J. Pres. Ves. Pip. 79 (2002) 273–278.
[4] X. Cheng, J.W. Ficher, H.J. Prask, T. Gnaupel-Herold, B.T. Yen, S. Roy, Int. J. [16] P. Paris, F. Erdogan, J. Basic Eng. 85 (1963) 528–534.
Fatigue 25 (2003) 1259–1269. [17] R.G. Forman, R.M. Engel, J. Basic Eng. 89 (1967) 459–463.
[5] K.H. Frank, J. Struct. Div. 105 (1979) 1727–1739. [18] W. Elber, ASTM STP 486 (1971) 230–242.
[6] C.L. Tsai, J. Struct. Eng. 112 (8) (1986) 1761–1780. [19] NASA, Fatigue Crack Growth Computer Program NASGRO Version 3.0–
[7] S.J. Maddox, Int. J. Fatigue 25 (8) (2003) 1359–1378. Reference Manual. JSC-22267B, NASA, Lyndon B. Johnson Space Center,
[8] M.H. Swellam, P. Kurath, F.V. Lawrence, Electric-potential-drop studies of Texas, 2000.
fatigue crack development in tensile-shear spot welds, in: M.R. Mitchell, R.W. [20] S.J. Maddox, Fatigue Strength of Welded Structures, Abington, 1991.
Landgrap (Eds.), Advances In Fatigue Lifetime Predictive Techniques, ASTM, [21] M.S. Miller, J.P. Gallagher, In ASTM STP 738, 1981, pp 205-251.
STP 1122, American Society for testing and materials, Philadelphia, 1992, pp. [22] D. Taylor, A Compendium of Fatigue Thresholds and Growth Rates, EMAS Pub,
383–401. 1985.
[9] J.A.M. Ferreira, C.M. Branco, Theor. Appl. Fract. Mech. 9 (1988) 23–32. [23] Y.B. Lee, C.S. Chung, Y.K. Park, H.K. Kim, Int. J. Fatigue 20 (8) (1998) 565–573.
[10] R. Yee, D.J. Burns, Thickness effect and fatigue crack development in welded [24] W. Guo, C.H. Wang, L.R. Frose, Fatigue Fract. Eng. Mater. Struct. 22 (1999) 437–
joints, Proc. 7th Inter. Conf. on Offshore Mechanics and Arctic Engineering, 444.
Houston, Texas, 1988, pp. 447–457. [25] Y.P. Srivastava, B.L. Garg, Eng. Fract. Mech. 22 (6) (1985) 915–926.
[11] T.N. Nguyen, M.A. Wahab, Fract. Eng. Mater. Struct. 51 (1) (1995) 1–18. [26] S. Dubey, A.B.O. Soboyejo, W.O. Soboyejo, Acta Mater. 45 (7) (1997) 2777–
[12] V. Balasubramanian, B. Guha, Eng. Fail. Anal. 11 (2004) 967–974. 2787.