Numerical study of scale effect using a dynamic random lattice model

The scale effect on rock properties has been discussed and studied since the introduction of the weakest-link idea by Weibull (1939). In this study, the capabilities of a dynamic lattice network model to simulate failure of rock under uniaxial tension is utilized. A number of randomly distributed microcracks that contribute to the development of uneven stressfields are introduced. In addition, each unit element in the model is assigned a random value representing its strength properties. Each application creates a randomly created specimen that is subjected to uniaxial tension. The specimens are grouped into three sets according to their size. The results of the study show that the scale effect is present in a statistical sense, i.e., the observed statistical distributions of strength for each specimen group is distinctly separate, and they follow the expected inverse relationship between strength and size. The observed probability distribution for each group of specimens give rise to the estimation of a different set of parameters. In this paper, the potential relationships among these parameters as a function of specimen size is investigated.