5th International Conference on Civil, Architectural and Hydraulic Engineering (ICCAHE 2016)

Study on calculation of stress intensity factor using finite element

method
a
Lin Li,b Dan YANG
a
Fujian Jiangxia University, Fuzhou, Fujian, P. R. China
b
Huazhong University of Science and Technology, Wuhan, Hubei, P.R. China

KEYWORD: Stress intensity factor; the finite element method; ordinary element; singular element
ABSTRACT: Stress intensity factor can reflects the stability of cracks. it is important to study accu-
rate calculation method of stress intensity factor. in this study, the finite element method with mod-
eling using ordinary element and singular element is applied to calculate stress intensity factor. it is
shown that calculation result from the finite element model using singular element is more accurate
than that from the finite element model using ordinary element.

INTRODUCTION
Cracks in concrete result in stress concentration, which may cause failure of structures with cracks
under low stress level. A lot of studies were carried out on fracture criterion of cracks. Erdogan pro-
posed maximum stress hypothesis, which can predict fracture angle accurately. However, the hypo-
thesis is not suitable to predict fracture toughness(Erdogan & Sih 1963). Stain-energy-density factor
proposed by Sih is widely applied to mixed mode crack problems(Sih 1974). Based on the principle
of energy conservation, an energy-based approach for the finite element modeling of mixed mode
cohesive crack propagation was developed. Compared to the strength-based cohesive crack model-
ing approach, the energy-based approach is both theoretically more fundamental and numerically
more effecient(Xie & Gerstle 1995). An analytical model was developed to determine the effects of
progressive reduction in stiffness as crack formation due to thermal stresses, which was verified by
experimental data(Elbadry & Ghali 1995). The cracking of reinforced concrete beams, with and
without prestressing, resulted from the temperature gradient was studied. it is shown that the partial-
ly prestressed cracked member provided with nonprestressed steel has the advantage of limiting the
width of cracks(Ariyawardena et al. 1997). Stability of cracks is crucial to structural safety. Since
stress intensity factor can reflect the stability of cracks, it is important to study accurate calculation
method of stress intensity factor.

STRESS INTENSITY FACTOR

Fracture mode
There are three basic fracture modes, as shown in Figure 1. Mode I, mode II and mode III are tensile
fracture, in-plane shear fracture and anti-plane shear fracture, respectively. Tensile stress is perpen-
dicular to the fracture surface in mode I. Shear stress is parallel to the fracture surface, and perpen-
dicular to the fracture front in mode ii. Shear stress is parallel to the fracture surface and the fracture
front in mode III. Not all fracture problems can be simplified as one of three basic modes in actual
engineering. There are mixed modes, which are combination of three basic modes.

(a) mode I (b) mode II (c) mode III

Figure 1. Fracture modes

© 2016. The authors - Published by Atlantis Press 592

Analysis of stress intensity factor
Since fracture strength is affected significantly by the deformation in the field near crack tip, it is
necessary to analyze the displacements and stresses in the field.
Since crack size is much smaller than structure size, crack can be supposed to located on an infi-
nite plate. The crack length is 2a and tensile stresses are exerted on the plate in the directions of x
and y axes in mode I, as shown in Figure 2. For convenience, the crack tip is taken as the origin of
coordinates.

Figure 2. diagram of crack

According to the solution to the elastic plane problem, the stress and displacement at a point near
the crack tip, P(r，θ), are as follows
KI θ θ 3θ  
σx = cos 1 − sin sin  
2π r 2 2 2  
KI θ θ 3θ  
σy = cos 1 + sin sin  
2π r 2 2 2  
KI θ θ 3θ 
τ xy = sin cos cos  (1)
2π r 2 2 2 
τ xz = τ yz = 0 

σ z = ν (σ x + σ y ) ( plane strain ) 
σz = 0 ( plane stress ) 
2 (1 + ν ) K I r  θ 3θ  
u= ( 2 k − 1 ) cos − cos 
4E 2π  2 2  
2 (1 + ν ) K I r  θ 3θ  
v= ( 2 k + 1 ) sin − sin 
4E 2π  2 2   (2)

w=0 ( plane strain ) 
ν 
w = − ∫ (σ x + σ y ) dz ( plane stress )
E 
Where r and θ are polar coordinates of a point near crack tip. u, v and w are displacement compo-
nents along x, y and z axes. σ x , σ y , τ xy , σ z , τ xz and τ yz are stress components. E and ν are
respectively elastic modulus and Poisson’s ratio. k and KI are modulus of dilatation and stress in-
tensity factor, respectively.
3 − 4ν , ( plane strain )

k =  3 −ν (3)
 , ( plane stres s )
 1 +ν
593
 KI = σ π a (4)
Similarly, the stress, displacement and stress intensity factor of mode II are as follows
− K II θ 3θ 3θ  
σx = sin  2 + cos cos  
2π r 2 2 2 
K II θ θ 3θ 
σy = sin cos cos 
2π r 2 2 2 
K II θ θ 3θ  
τ xy = cos 1 − sin sin   (5)
2π r 2  2 2  
τ xz = τ yz = 0 

σ z = ν (σ x + σ y ) ( plane strain ) 
σz = 0 ( plane stress ) 
2 (1 + ν ) K II r  θ 3θ  
u= ( 2 k + 3 ) sin + sin 
4E 2π  2 2  
2 (1 + ν ) K II r  θ 3θ  
v= ( 2 k − 2 ) cos + co s 
4E 2π  2 2   (6)

w=0 ( plane strain ) 
ν 
w = − ∫ (σ x + σ y ) dz ( plane stress ) 
E 
K II = τ π a (7)
Corresponding components of mode III has following expressions

σ x = σ y = σ z = τ xy = 0 

K III θ 
τ xz = sin  (8)
2π r 2 
K III θ 
τ yz = cos 
2π r 2 
u=v=0 

2 (1 +ν ) K III 2r θ (9)
w= sin 
E π 2
K III = τ π a (10)
From above stress expressions of mode I, II and III, it can be found that stresses can not be acted
as strength condition. However, stress intensity factor is finite, which represents strength of stress
field.It is appropriate for stress intensity factor to establish fracture criterion.

CALCUALTION OF STRESS INTENSITY FACTOR USING FININTE ELEMENT ME-

THOD
Calculation method of stress intensity factor
Calculation method of stress intensity factor can be classified into three categories: analytical me-
thod, the finite element method and experimental method. Analytical method is an exact one, which
is suitable for solving problems with simple geometrical shapes. Stress intensity factors can be con-
sulted from references for simple structures. However, it is difficult for complex structures to calcu-
late stress intensity factors using analytical method. It is common for most structures to calculate
stress intensity factors using the finite element method. in some special cases, stress intensity factors
594
need to be verified by experiment. Following study focuses on the finite element method for cal-
culating stress intensity factors.
There are two types of elements to simulate crack, ordinary elements and singular elements.
When ordinary elements are selected, stress intensity factor is derived from stresses and displace-
ments in the field near crack tip. When singular elements are selected, stress intensity factor is cal-
culated directly.
Ordinary element
The displacement, vi , and coordinate, ri , at an integration point of ordinary element near crack tip
can be obtained from finite element program. Substituting θ = π into Equation (2) yields

vi =
( k + 1)(1 +ν ) K ri (11)
2π
Ii
E
Substituting Equation (3) into Equation (11) yields
E 2π
 vi ( plane stress )
 4 ri
K Ii =  (12)
 E 2 π
v
 4 (1 −ν ) 2 i r ( plane strain )
 i

Suppose K I is a linear function of r . Based on the least square method and data derived from
Equation (12), following equation can be obtained
K I = Ar + B (13)
When r = 0 , K I = B .
Similarly, stress intensity factor can also be derived from the extrapolation method based on
stress, where data for curve fitting are obtained from Equation (1). For mode II and mode III frac-
tures, stress intensity factors can be derived from Equation (6) and Equation (9)，respectively.
it should be noted that the stress intensity factors derived from the extrapolation method vary ac-
cording to mesh density.
Singular element
Singularity can be reflected by singular element, which has little related to mesh density. Singular
elements include triangular element with singular strain, isoparametric element with singular strain.
A singular element in ANSYS program is shown in Figure 3.

Figure 3. diagram of PLANE82

NUMERICAL EXAMPLE
A plate with mode I fracture was simulated using ANSYS program. The length, width and thickness
of the plate is 200mm, 400mm and 1mm, respectively. The length of crack is 40mm. Modulus of
elasticity and Poisson’s ratio is 2×105 MPa and 0.25, respectively. σ is 30 MPa, as shown in Figure
4.

595
 Figure 4. diagram of a plate with mode I fracture

The analytical expression of stress intensity factor has the following form
πa  a 
2 4
a
K I = σ π a sec 1 − 0.025   + 0.06    (14)
2W   w  w  
Based on the Equation (14), the analytical solution of stress intensity factor is obtained,
K I = 243.6 Mpa ⋅ mm .
The crack was first simulated using ordinary element, and the mesh near the crack was refined. A
half of the plate was analyzed according to symmetry. Total element number is 9640, as shown in
Figure 5. The deformation of the field around the crack is shown in Figure 6.

Figure 5. diagram of model meshed with ordinary element

Figure 6. deformation of the field around the crack

Based on the displacements near the crack tip, stress intensity factor was obtained,
K I = 243.9 Mpa ⋅ mm .

596
 in order to analyze the effect of element size, the sizes of all elements were doubled. There are
2410 elements in this case, and K I = 238.2 Mpa ⋅ mm .
Finally, the crack was modeled using singular element, as shown in Figure 7. Stress intensity fac-
tor was calculated, K I = 243.7 Mpa ⋅ mm . All solutions are listed in
Conclusion
Stress intensity factor can be obtained by finite element analysis with modeling using ordinary ele-
ment or singular element. Error in the model using ordinary element is larger than that in the model
using singular element. Moreover, calculation results from the finite element model using ordinary
element vary from element size to element size. Therefore, it is more appropriate to calculate stress
intensity factor by singular element.
Table 1.

Figure 7. diagram of model meshed with singular element

CONCLUSION
Stress intensity factor can be obtained by finite element analysis with modeling using ordinary ele-
ment or singular element. Error in the model using ordinary element is larger than that in the model
using singular element. Moreover, calculation results from the finite element model using ordinary
element vary from element size to element size. Therefore, it is more appropriate to calculate stress
intensity factor by singular element.
Table 1. Solutions to stress intensity factor
Ana- Ordi- Ordi- Singu-
lytical nary nary lar ele-
solu- ele- ele- ele-
tion ment ment ment
with
large
size
Ki
243.6 243.9 238.2 243.7
Mpa ⋅ mm
Error
0 0.13 2.21 0.03
%

ACKNOWLEDGEMENTS
The authors acknowledge the financial support by the National Natural Science Foundation of Chi-
na (Grant No. 51278214).

REFERENCES
[1] Ariyawardena, N., Ghali, A. & Elbadry, M. 1997. Experimental study on thermal cracking in
reinforced concrete members. ACi Structural Journal 94(4): 432-441.
[2] Elbadry, M. & Ghali, A. 1995. Control of thermal cracking of concrete structures. ACi Structur-
al Journal 92(4): 435-450.
[3] Erdogan, F. & Sih, G. C. 1963. On the crack extension in plates under plane loading and trans-
verse shear. Journal of Basic Engineering 85(4): 519-525.
[4] Sih, G. C. 1974. Strain-energy-density factor applied to mixed mode crack problems. interna-

597
 tional Journal of Fracture 10(3): 305-321.
[5] Xie, M. & Gerstle, W. 1995. Energy-Based cohesive crack propagation modeling. Journal of
Engineering Mechanics 121(12): 1349-1358.

598