A smoothed variable horizon peridynamics and its application to the fracture parameters evaluation

Imachi, Michiya and Takei, Takaaki and OZDEMIR, Murat and Tanaka, Satoyuki and Oterkus, Selda and Oterkus, Erkan (2020) A smoothed variable horizon peridynamics and its application to the fracture parameters evaluation. Acta Mechanica. ISSN 1619-6937 (In Press)

[thumbnail of Imachi-etal-AM-2020-A-smoothed-variable-horizon-peridynamics-and-its-application-to-the-fracture-parameters-evaluation]
Preview
Text. Filename: Imachi_etal_AM_2020_A_smoothed_variable_horizon_peridynamics_and_its_application_to_the_fracture_parameters_evaluation.pdf
Accepted Author Manuscript

Download (1MB)| Preview

Abstract

Peridynamics has attractive features for solving several fracture mechanics problems. On the other hand, to ensure the accuracy, a great deal of particles (material points) is required. Introduction of variable horizon is an alternative approach by changing the horizon size over the problem domain. In the present study, we propose a novel variable horizon concept. It is known that undesired “ghost” force arises along the transition region of different horizons. In order to suppress the ghost force, the gradual variation of the horizon size over a certain region, called smoothing length, is introduced between the different scale particle distributions. Efficiency of the smoothed variable horizon peridynamics is demonstrated through several numerical studies employing the ordinary statebased peridynamics. As a basic case, a linear displacement field is considered. It is observed that the proposed approach significantly reduces the ghost forces along the interface of different spatial discretizations. Additionally, the dynamic stress intensity factors of stationary cracks are carefully examined. The path independence of the fracture parameters in the variable horizon peridynamics is ensured. Efficiency of the presented method is then discussed. It is monitored that introduction of the smoothing length concept significantly reduces the computational costs in the peridynamic modeling.