Numerical modelling is an effective alternative for studying the mechanical behaviour of rock masses. Very often, in solving the same problem with the same code, different models yield remarkably different results. There are many reasons. Amongst them, the boundary conditions are often overlooked but it has significant influence on the stress wave propagation in jointed rock masses. Usually, in solving dynamic problems, the incident boundaries are described either by stress-history input (SHI) or velocity-history input (VHI). This paper presents the study on the effect of these two approaches using discrete element method (DEM) on wave propagation and wave attenuation in jointed rock masses. Two cases were designed to illustrate the points. The first one shows one-dimensional P-wave propagation along a rock bar having an internal cross joint. The second one shows the blasting wave propagating through jointed rock masses. The blasting wave is induced by an underground explosion. The rock bar/masses are discretized into elements and the analyses adopt the DEM, which is available in the code ‘UDEC’. The results reveal that the VHI approach can yield results in accord with field-test data, while the SHI approach may lead to unreasonable results. The two approaches may yield the same results for homogeneous rock, but definitely not the same for jointed rock masses. In addition to the numerical illustrations, this paper also presents some insights of the reason behind it.
Read full abstract