PSI - Issue 2_B

7

Chretien Gaëlle et al. / Procedia Structural Integrity 2 (2016) 950–957 Gaëlle Chretien et al. / Structural Integrity Procedia 00 (2016) 000–000

956

a

b

Fig. 9. Kitagawa diagram for Ti-6Al-4V at R=0.1 at: (a) 20°C; (b) 400°C.

With a 0 the transition crack size developed in part 4.1 (Equation 3). This model is plotted on figures 9 a and b by dashed lines and fits relatively well the evolution of the threshold with crack length at both 20°C and 400°C. a 0 decrease when the temperature increases. However this length gets no physical meaning. 4.3. Other criteria The bend in the Kitagawa diagram must be explicated by the evolution of crack closure. McEvily and Murakami (2003) developed an equation to fit the curve. It decomposes the threshold in two parts: the effective threshold and a part which expresses the crack opening. The crack propagation threshold of the short cracks is expressed as equation 6 with k a material parameter dependent of loading and temperature conditions and which takes into account the development of crack closure. No more physical meaning for this expression. � e 1 �.� K K � K � a � K a. k eff th LC , nom th eff ht th          (6) Chapetti (2005) modified this equation by introducing microstructural parameters, especially the size of the strongest microstructural barrier d. (Equation 7). � e 1 �.� K K � K � a � K � d a .� k d th LC , nom th d th th           (7) With the microstructural threshold: d. . . Y K th d th     and d th FL th d th K K K . d4 1 k      . Despite the very interesting physical meaning of this proposition, the model overestimates the crack length without propagation for a given stress range (Figure 9 a and b, dashed and pointed lines). Recently Maierhofer et al. (2014) adopted a new approach taking into account a decomposition of closure by each known mechanism of closure. It bases on the equation 6 developed by McEvily and Murakami (2003). To apply this approach, the precise involved crack closure mechanisms would be known. Thus roughness and oxide thickness must be measured on all specimens at 20°C and 400°C. 5. Conclusions This study on the contribution of the effect of crack closure on the propagation of 2D physical short cracks in comparison to long cracks and on the determination of a non-propagation criterion leads to the following conclusions:  The effect of temperature on crack growth propagation of long crack is negligible except in the plateau domain.