A discrete element study of the relationship of fabric to wave propagational behaviours in granular materials

Wave propagation in granular materials is numerically studied through discrete element simulation. Two-dimensional (2-D) model material systems composed of large numbers of circular particles were numerically generated. The particles in these model materials were randomly distributed with a biasing algorithm to produce fabric anisotropy so as to create preferred directions within the material. Wave motion is introduced through dynamic loadings to appropriate boundary particles to produce horizontal and vertical plane wave propagation within each model material Discrete element simulation with a non-linear hysteretic interparticle contact law is used to model the dynamic behaviour of the model granular systems, and this yields information on the wave speed and amplitude attenuation. Through the investigation of several model systems, relationships are established between wave propagational characteristics and granular microstructure or fabric. Specific fabric measures which were used included branch vectors, path microstructures and void characteristics. Distributions of these fabric descriptors were determined, and comparisons and correlations were made with the discrete element wave propagation results. Conclusions of this study indicated that while all three fabric measures provided some degree of correlation with the wave motion behaviours, the void fabric descriptor produced the best correlation for the assemblies under investigation. © 1997 by John Wiley & Sons, Ltd