Random growth of crack with R-curve: Markov process model

A probabilistic model for the randomness of the progressive crack growth in a quasi-brittle material such as concrete is presented. The model consists of a Markov chain adapted to R-curve behavior. It yields the crack propagation probability in any loading step as well as the probability of failure at any stage of the fracture process. The R-curve is obtained from the given test data on the effect of structure size on the maximum loads. The standard deviation of the peak load is the minimum statistical information required. According to the available test results, this standard deviation is approximately a linear function of the crack propagation distance. The parameter estimation method is formulated and some applications are illustrated.