Dual horizon peridynamic approach for studying the effect of porous media on the dynamic crack growth in brittle materials

Abstract

Abstract This study aims to investigate the effect of the stop-hole configurations on dynamic cracks using the Dual Horizon Peridynamic approach (DHPD). The traditional PD formulations require uniform discretization for the solution domain. Therefore, in order to investigate the effect of the stop holes with traditional PD, more densely spaced points are used to represent the holes. If less densely spaced points are used in traditional PD, this may lead to unexpected crack initiations around the stop holes due to strain concentrations originating from the shape of stair steps. If non uniform point spacing is used for representing the outer edge of stop hole accurately in traditional PD, ghost force problem may occur and the forces of the points in the bond may not interact in pairs. Moreover, the use of fine mesh causes high computational efforts. Therefore, the DHPD, which allows non-uniform point spacings unlike traditional PD, is employed for examining the characteristics of dynamic crack propagation in the presence of stop holes. Before stop-hole configurations are examined, the developed DHPD code is validated with the results of a reference solution for the Kalthoff-Winkler test. After the validation process for DHPD code is successfully completed, the crack paths in the stop hole configurations obtained by DHPD are compared with the experiments, and a good agreement between the test and simulation is achieved. Furthermore, the performances of the stop hole configurations are investigated by DHPD in terms of increasing fracture toughness of brittle material. Besides, the effect of crack branching on dynamic crack retardation is also examined. Results of dynamic crack propagation simulated by DHPD are compared with the experimental observations and reference results. The results of DHPD are in good agreement with the experimental observations available in the existing literature.