13th International Conference on Fracture June 16–21, 2013, Beijing, China -6- precipitously, then decreases gradually and tends to be a constant as the increase of thermal shock temperature difference ∆T. Also the Fig. 3 shows that the critical thermal shock temperature difference ∆Tc with temperature-dependent material properties is higher than the one which does not take the effect of temperature on material properties into account. And, if the ∆T is less than a certain value the thermal shock residual strength with temperature-dependent material properties is higher than the one without the consideration of temperature-dependent material properties. However, the results are reversed as the ∆T becomes larger than the certain value. Therefore, if the temperature dependence of material is ignored, the critical thermal shock temperature difference ∆Tc will be underestimated, and the thermal shock residual strength will be wrong determined. Thus, consideration of temperature-dependent material properties is essential for correct evaluation of thermal shock resistance and thermal shock residual strength of a material, especially for UHTCs suffered high temperature thermal shock. 0.00 0.05 0.10 0.15 0.20 0.25 0 3 6 9 12 15 T0=600 oC temperature-dependent temperature-independent t (s) KI (MPa⋅m1/2) h=50 kW/(m2⋅K) c0/a=0.1 T0=1000 oC (a) 0.00 0.05 0.10 0.15 0.20 0.25 0 3 6 9 12 15 18 T0=600 oC temperature-dependent temperature-independent t (s) KI (MPa⋅m1/2) h=80 kW/(m2⋅K) c0/a=0.1 T0=1000 oC (b) Figure 4. Thermal stress intensity factor as a function of time ( (a), h=50kW/(m2·K), (b), h=80kW/(m2·K) ) In Fig. 4, setting the crack to be the initial length during the thermal shock proceeding, the TSIF KI is plotted as a function of thermal shock time for thermal shock initial temperature T0=600oC and 1000oC with surface heat transfer coefficient h=50kW/(m2·K) in Fig. 4 (a) and h=80kW/(m2·K) in Fig. 4 (b), respectively. The results show that the peak values TSIF KI for 600 oC with temperature-dependent material properties are lower than the one without temperature-dependent material properties, on the contrary, when thermal shock initial temperature is 1000oC, the peak values with temperature-dependent material properties is higher than the one which does not take the effect of temperature on material properties into account. Because the coefficient of thermal diffusion of high temperature without temperature-dependent material properties is larger than the one with temperature-dependent material properties, the serious temperature gradient will moderate sooner, which can result in relaxing of thermal stress induced by thermal shock. Therefore, the phenomenon, that if the ∆T is less than a certain value the thermal shock residual strength with temperature-dependent material properties is higher than the one without the consideration of temperature-dependent material properties but the results are reversed as the ∆T increase shown in Fig. 3, is explained. And it is obvious that the thermal shock residual strength is strongly affected by the dependence of temperature of material properties. Thus, when determining the thermal residual strength of UHTCs for high temperature used, the result of high temperature thermal shock, that thermal residual strength with the consideration of temperature-dependent material properties is higher than the one without temperature-dependent material properties, is not always right. It must fully consider the effect of temperature on the material properties. 5. Conclusions
RkJQdWJsaXNoZXIy MjM0NDE=