Issue 50

J. Papuga et alii, Frattura ed Integrità Strutturale, 50 (2019) 163-183; DOI: 10.3221/IGF-ESIS.50.15

dominant axial loading leads to the greatest differences. These cases could serve for proving whether the phase shift effect of these criteria is adequate.

Figure 8 : Results of the sensitivity analysis on two versions of the Dang Van method.

Figure 9 : Results of the sensitivity analysis on two integral criteria.

The integral methods shown in Fig. 9 show behavior similar with critical plane criteria of the MSSR type, i.e. with the non zero phase shift worsening the fatigue response for brittle materials. This effect can be observed above all for the PI method, while it is less prominent for the Liu & Zenner method. The phase shift effect is also pronounced for extra-ductile materials, where it improves the fatigue response. The phase shift effect approaches minimum values for load cases with dominant torsion. The phase shift effect of the PI method for a ductile material approaches zero – it would obviously be hard to detect in experiments. There are some load combinations by the Liu & Zenner method, where the phase shift effect for ductile materials exceeds 5% difference. The behavior of the Crossland method in Fig. 10 quite well resembles Fig. 8 for the original Dang Van solution. The phase shift effect of all materials for a stress ratio of 1.73 would be easily detectable in any experiment. The Liu & Mahadevan method, also shown in Fig. 10, draws attention to another interesting consideration – the functions show a slightly irregular character. This is probably caused by the two-level search for the final critical plane. Unlike all other methods, the method by Papadopoulos and the MMP method would show only a horizontal line in graphs similar to Figs. 6-10 at the position FI OOP / FI IP =1.0. These two methods could not be distinguished from each other by the trend of the phase-shift effect alone.

177