Numerical simulations
At our groupwe develop theoretical and numerical models. Our researchers cooperatively develop DEM, DEM-LB (Combination of LatticeBoltzmann and DEM) and continuous models to accurately describe particle systems on different spatio-temporal scales. On the other hand, recently graphics cards have become a promising alternative for parallel computing on clusters or supercomputers.. Our grouphas gained experience on developmentof hybrid GPU-CPU codes on the NVIDIA-CUDA architecture. In some examples, we have obtained very significant gains in the efficiency of our algorithms, compared to over-CPU codes.
In the past, we have explored some applied scenarios such as granular media, crowd dynamics and complex fluids. We have conducted parametric programs of study, examining the influence of particle shape, contactmodels and other specific details. Our numerical descriptions have helped to identify the most relevant parameters that facilitate calibrations, using experimental dataand analytical predictions. In addition, we have used averaging approaches that link the behaviour of these systems, at the local scale (contacts between constituent elements) with their macroscopic response. Numerical-experimental synergy has enabled the applicability of our methods and findings. Our final scientific goalis to obtain continuous descriptions of particle flows, based on the micromechanical details. This will shed light on the knowledgeof how local micromechanical details affect the constitutive response and jamming of complex granular flows.
More information at