Abstract
This paper demonstrates the importance of a proper contact algorithm selection when a constitutive model is correlated and validated, especially in the case of brittle materials. A parametric study is carried out to study the influence of contact parameters on the outcomes of the numerical simulations of a dynamic compression test. The split Hopkinson pressure bar (SHPB) model is developed, and sandstone rock is considered as a representative material having considerably different properties compared to SHPB bars. The finite element method (FEM) and smoothed particle hydrodynamics (SPH) were used to simulate specimen behaviour using a LS-Dyna solver. Two contact types based on the penalty method are analysed: nodes to surface (FEM and SPH) and surface to surface (FEM only). Furthermore, three approaches of contact stiffness calculation are used for each contact type. The waveform data and failure patterns are then compared among all simulated cases and the corresponding experimental outcomes. It is found that the soft constraint stiffness (SOFT = 1) provides the best outcomes, especially in the case of one-way contact, and is nearly insensitive to stiffness scaling parameters. By contrast, standard (SOFT = 0) and segment-based (SOFT =2) approaches require a substantial effort in adjusting the stiffness scaling parameters to obtain satisfactory results. This paper provides valuable guidance for correlating and validating parameters of constitutive models for rock and other brittle materials in the SHPB test.