On modelling quasi-static uniaxial tension and compression tests on rock with explicit time stepping

Modelling rock failure in engineering applications, such as blasting and percussive drilling, involve stress wave propagation, high stress rates, and substantial fracturing/fragmentation during extremely short time spans. Such circumstances dictate using explicit time stepping in solving the global system of problem governing equations.
The material model development requires validation under dynamic loadings, especially in uniaxial tension and compression tests. However, the failure model must also be able to predict the quasi-static uniaxial tensile and compressive strengths as well as the failure modes of the rock type involved. Unfortunately, the explicit time stepping is only conditionally stable and, thus, modelling quasi-static tests of a laboratory sample size numerical rock sample becomes a computationally laborious task. It is, therefore, tempting to increase the loading rate as much as possible when performing these validation simulations. Notwithstanding, using too high strain rates leads to strain rate hardening effects and affect the failure mode, triggering even a transition from a single (few) macro-failure plane(s) to multiple fracture/fragmentation beyond certain loading rate depending on the loading type and
sample size. However, there seems to be no guiding lines in the literature as to how high a loading rate can be used in uniaxial tension and compression tests, to save the CPU time, so that the simulation results can still be considered valid. The present study addresses this gap of knowledge by performing a series of numerical tests
on brittle rock under uniaxial tests using an explicit (in time) finite element code. The rock failure is described in the continuum sense based on a damage-viscoplasticity model. The 2D simulations demonstrate that with a sample of size 25*50 (mm), strain rates up to 1 s-1 can be used in both tension and compression without significant deviations from the quasi-static case.