13th International Conference on Fracture June 16–21, 2013, Beijing, China -9- temperatures typical of those of in-service conditions. Figure 8. Variation of the fracture toughness with temperature. Experiments and predictions from Beremin-type bimodal model [2]. The calculated values corresponding to failure probabilities of 5%, 63% and 95% are shown. 6. Conclusions 1. Thermal aging of a 18MND5 plate (type A533B) at 450°C for 5000 hours produces a shift of the ductile-to-brittle transition temperature (DBTT) of about 39°C. This shift is not accompanied by a drastic change in fracture modes which remain partly intergranular in the segregated zones and transgranular cleavage in the matrix. The fracture toughness of aged material is much more scattered than that of the as-received material. 2. The slow cooling rate applied during cooling after the stress-relieving heat treatment (~ 30°C/h) generates intergranular phosphorus segregation which is sufficient to provoke intergranular fracture in the segregated zones, even in the as-received conditions. 3. Auger spectrometry measurements showed that during aging at 450°C, phosphorus continues to segregate along the grain boundaries, while carbon concentration at grain boundaries decreases. Both phenomena produce an embrittlement effect of the grain boundaries. 4. A model of intergranular segregation including the interaction between C and P atoms and non-isothermal conditions was identified. The results of this model were used as inputs for the prediction of bimodal fracture toughness of an inhomogeneous material representing the banded structure of the micro-segregated steel. The kinetics is represented using the McLean – Guttmann – Militzer theories. 5. The variations in DBTT and in the scatter of fracture toughness test results induced by aging are correctly predicted using these models of segregation and of fracture toughness. Acknowledgments This work has been made in the frame of the Areva - Ecole des Mines industrial professorship. Support from Areva is acknowledged. The authors would like to acknowledge also D. Loizon, S.
RkJQdWJsaXNoZXIy MjM0NDE=