Energy-Consistent Multiscale Algorithms for Granular Flows.

摘要

In this final report, we document the achievements made as a result of this Young Investigator Program (YIP) project. We worked on the development of multi scale energy-consistent algorithms to simulate and capture flow phenomena in granular materials. For this, we have made progress in three key areas: (i) the development of unprecedented algorithms at the grain scale to simulate flow of complex (real) granular media; (ii) the development of multiscale and continuum approaches to simulate the continuum behavior of granular media under large flow; (iii) the development of experimental techniques and approaches to model the behavior of granular materials under extreme avalanche flow. In the area of algorithmic development at the grain scale, we have successfully developed a technique, called the granular element method (GEM), and have been able to bypass all the major shortcomings of current discrete element methods (DEM). We have shown that the technique is not only much more efficient algorithmically, but displays a degree of accuracy that is unprecedented for discrete methods. In the area of multiscale and continuum approaches, we have focused our attention on the development of continuum models that are micro-inspired and in this way can bypass much of the phenomenology that plagues current continuum models for granular flow. Finally, in the areas of experimental techniques, we have been able to use clever flow experiments in tandem with our advanced models to decipher the mechanisms controlling the main continuum features observed previously but that have not been able to be captured comprehensively in models. The consequences of these advancements are broad and deep. The GEM method has revolutionized the accuracy of discrete methods: capturing real material behavior for the first time. The multiscale continuum models have been able to explain the source of rate dependence in frictional resistance.