A kinetic treatment of the brittle fracture materials

Abstract

Brittle fracture is considered as a kinetic process. The experimental data available indicate that the description of fracture from the viewpoint of a critical event which occurs at the moment when a certain limiting stress is reached is impossible. A large number of experimental facts which indicate that brittle fracture kinetics are essentially different from plastic fracture kinetics are discussed. Often it is difficult to obtain the stress dependence of a specimen lifetime (the longevity) for static loading and sometimes the finite lifetime measurement probability (the delayed fracture phenomenon) is extremely low, the temperature and rate dependences of the strength at brittle fracture being anomalous. A physical model of brittle fracture is suggested. Its kinetic peculiarities can be explained in terms of the relaxation which occurs at a lower rate than relaxation at plastic fracture. For this reason it appears impossible to obtain a structure with a constant overstress coefficient at a fracture nucleus during the test period. Quantitative characteristics of brittle fracture are based on three fundamental relationships: the longevity - stress dependence (Zhurkov's law): the logarithmic formula for load stress relaxation due to local inelastic deformation; the principle of linear summation of the damage. The theoretical equations obtained were experimentally checked on different materials (zinc, silicon, mild steel and several polymers) for static and cyclic loadings. Conclusions are drawn about the common physical nature of brittle and plastic fracture, the relation between the brittle fracture characteristics for static and cyclic loadings and the possibility of brittle fracture prediction for long-term tests by using short-term test data. The activation parameters of the process were calculated from lifetime data obtained experimentally.