PSI - Issue 59

Eugene Kondryakov et al. / Procedia Structural Integrity 59 (2024) 50–57 Eugene Kondryakov et al. / Structural Integrity Procedia 00 (2023) 000 – 000

55

6

for each material these parameters need to be calibrated using experimental results. In this work, the maximum principal stress σ c = σ b = 820 MPa Qian et al. (2019) was taken as the crack initiation criterion parameter. To determine the fracture energy, three methods presented in Qian et al. (2019), Yu et al. (2021), Abdellah et al. (2023) were used, resulting in the following values: Gc = 69000 N/m, Gc = 235000 N/m, Gc = 680 N/m. Numerical simulations of crack propagation in the mini - C(T) specimen under eccentric tension were performed for the three G c values. Fig. 7 shows a comparison between the experimental and calculated load-displacement curves. As can be seen, at G c = 235000 N/m the experimental and calculated results are very close, so this value was used in further analyses.

Fig. 6. Opening mode stress distribution ahead of the crack tip in CT-0.5T (a), side-grooved CT-0.5T (b) and miniature CT-0.16T (c) specimens.

Fig. 7. Comparison of experimental and calculated load-displacement P(LLD) curves for the three G c values.

Numerical simulations of crack propagation were performed for the three specimen types using the determined fracture criteria parameter values. Fig. 8 shows the crack propagation kinetics in specimens of the three types. It can be seen that for the 0.5T C(T) and 0.16T C(T) specimens (Fig. 8a,b), the crack front is nonuniform over the whole fracture process, and the crack length in the specimen mid-section is much greater than at its surface. In this case, plane strain conditions are not met. While for the specimen with side grooves (Fig. 8c), the crack front is more uniform. At the beginning of the fracture, the crack length at the surface is even slightly greater than in the middle of the specimen, and when reaching half the distance to its edge, the crack length is almost the same along the whole front. The formation of cross-sectional shrinkage on the specimen surface also reduces the crack propagation velocity V cr . Despite the lower stress level, the maximum crack velocity in specimens with side grooves is 40% higher at a distance L norm = 0.2 from the initial crack tip, compared to specimens without side grooves (Fig. 9). At the same time, the crack velocity in the miniature 0.16T C(T) specimens is higher than in the 0.5T C(T) specimens (Fig. 9) due to the higher stress level at the crack tip (Fig. 6a, c).