Crack Paths 2012

the contribution of mode II loading to the crack driving force does not exceed 5%, so

that the test evaluation performed below is based on the modeI stress intensity factor.

Simulation of Crack Growth Paths

As the X F E Mapproach [10] is receiving an increasing interest in fracture mechanics

applications, this numerical algorithm available in A B A Q U[S13] is applied below to

simulate F C Gpaths for the asymmetrically loaded bend specimens (SEB5, SEB6and

SEB7), as well as for the SE(T) configurations denoted by SET1and SET3. The SE(B)

model was meshed using square-shaped elements with the element size of 1 mm.For

the SE(T) models two mesh types – square-shaped and randomly shaped elements

(referred in Figure 4 to as “regular” and “irregular” mesh, respectively) – both having

the element size below 0.5 m mwere employed.

192

SEB5-tesdtata

SEB6-tesdtata

SEB7-tesdtata

SEB- X F E M

m

x , m

6

3

(a) SE(B)

0

0

5

10

15

20

25

30

35

40

y, m m

10

SET1- test data

8

SET1- X F E M ,regular m e s h

6

SET1- X F E M ,irregular m e s h

m

4

y , m

2

0

(b) SE(T), e = 10 m m

-2

0

5

10

15

20

25

30

35

40

x, m m

-4-202468

SET3- test data

SET1- X F E M ,regular m e s h

SET3- X F E M ,irregular m e s h

m

y , m

(c) SE(T), e = 8 m m

0

5

10

15

20

25

30

35

40

x, m m

Figure 4. Calculated vs. measured crack growth paths.

The calculated and experimental crack paths are compared in Figure 4. One can sum

up that a fairly good agreement exists between the numerical and test results for the

SE(B) and SET3 specimens. In the latter case, the X F E Manalysis is able of predicting

crack propagation towards the bore, although it cannot describe the specimen’s fracture

behaviour. At the same time a considerable discrepancy is noticed for the SET1

specimen, suggesting that crack growth mechanisms are not properly captured by

fracture criteria currently implemented in [13].

828