Effect of the Support Domain Size in SPH Fracture Simulations

článek v časopise ve Scopus, Jsc

angličtina

The simulation of numerical methods is accompanied by problems caused by false dependencies arising from the mathematical basis of such methods. The use of the Smoothed Particle Hydrodynamics (SPH) method removes the problems caused by the presence of a physical mesh that occur, e.g. when the Finite Element Method (FEM) is used. Although meshfree methods are generally less likely to produce false numerical dependencies, in some cases certain measures must be taken in order to avoid obtaining unexpected results. In the case of the SPH this necessarily involved the regular distribution of particles in discretized domains. This contribution describes a fracture mechanics experiment in which L-shaped concrete specimens undergo dynamic fracture. The experiment is simulated via the SPH method, during which clusters of particles are artificially created so that the resulting distribution in the discretized domain (or a zone within it) is irregular. The consequences of this irregularity and its effect on the form of failure are studied along with possible ways in which false (dependent) behaviour can be prevented. The results from the SPH method are also compared to FEM results.

Concrete, dynamic fracture, nonlinear constitutive model, smoothed particle hydrodynamics, support domai

HUŠEK, M.; KALA, J.; KRÁL, P.; HOKEŠ, F.

1. 10. 2016

1998-4448

International Journal of Mechanics

10

1

Spojené státy americké

396

402

7

http://naun.org/cms.action?id=12113