Issue 70

D. Kosov et alii, Frattura ed Integrità Strutturale, 70 (2024) 133-156; DOI: 10.3221/IGF-ESIS.70.08

(a) (c) Figure 15: Compact tension shear specimen (a,b) and test setup (c). (b)

(a) (b) Figure 16: S-shaped grips (a) and finite element mesh with CTS specimen (b).

The situation of initial pure shear is a subject of ongoing discussion in the research community. In continuation of this discussion, this paper presents an experimental substantiation of our numerical results. In the literature, the main question is which damage mechanism is dominant – branching or kinked. Following the formulation for the total potential energy functional (subsection “ The total potential energy functional for ductile fracture problem ”), damage can be reads by decomposition of the strain energy density as volumetric-deviatoric split (Eq.11) or spectral decomposition (Eq.12). Fig. 17 shows the experimental trajectories of crack growth obtained by Shlyannikov and Fedotova in the CTS sample for the conditions of pure mode I (Fig.17a) and the initial mode II of pure shear (Fig.17c, f) [25]. These results relate to samples made of high-strength steel 34X. It can be noted that for the same loading conditions and elastic properties of the material, branching and kinking of the initial pure shear crack are possible (Fig.17c,e). The results of modeling phase fields for the corresponding loading conditions are illustrated in Fig. 17(b,d,f). The phase fields in Fig. 17e correspond to the use of Eq.9, whereas in Fig.17f the phase fields refer to Eq.12. Recall that the differences in these equations lie in the decomposition of the strain energy density. Thus, due to various forms of writing the components of the governing equations of the energy balance, it is possible to obtain an adequate description of the experimental data after the fact, however, the question of predicting the crack growth trajectory remains open. In the context of the following elastic calculations, it is essential to establish the most physically meaningful relationship between G c , and l , considering that the length scale dictates the size of the fracture process zone and therefore cannot be regarded merely as a numerical parameter. A similar conclusion about the significance of the choice of constitutive equations was also expressed by Tsakmakis and Vormwald [6] based on a comparison of isotropic and kinematic hardening models in the analysis of phase fields during nonlinear cyclic deformation.

148