Finite element modelling of split-Hopkinson pressure bar experiments on sand

Finite element modelling of confined split Hopkinson pressure bar (SHPB) tests on dry quartz sand have been carried out using LS-DYNA to assess the suitability of models *MAT_SOIL_AND_FOAM and *MAT_MOHR_COULOMB for simulating high-pressure transient events in soils. Both models were defined using quasi-static data from high-pressure triaxial and one-dimensional compression tests on the sand in the mac2T test rig. The model SHPB was set up to replicate experimental conditions, and the nu-merical solutions compared to the experimental data. The results show that *MAT_SOIL_AND_FOAM replicates the volumetric response provided as input data, but fails to describe the experimental deviatoric response, behaving as a fluid at high pressures due to dependence on a fixed shear modulus. *MAT_MOHR_COULOMB model failed with negative volume errors at low stresses due to the lack of a compaction response in the model, causing the specimen to behave elastically throughout. Both models are therefore inappropriate for modelling the response of sand in high-pressure tests.