A New Update Criteria of Verlet List for Geotechnical Dense Granular Materials Under Periodic Loading

Discrete element method (DEM) based on graphic processing unit (GPU) is widely employed for investigating the responses of geotechnical dense granular materials under periodic or traffic loading. In most conventional GPU-based DEM, the maximum displacement of particles in the global coordinate system is considered as the criteria for updating Verlet list. Once the displacement exceeds the specified threshold, the Verlet list is updated. Although the particles experience considerable quasi-periodic displacement under periodic loading, the potential contact targets for most particles will not change during quite a few loading cycles. As a result, data between central processing unit (CPU) and GPU is transferred with an over-demanded high frequency and thus restrict the computational efficiency. In this study, we propose a criteria to improve the updates of Verlet list, in which the maximum displacement of particles in local particle coordinates system is used as the update criteria. The improved update criteria is plugged in the open source DEM software MUSEN. Then by simulating a previous laboratory full-scale half-sleeper model test, the performance of the proposed criteria is testified. The results show that the application of the proposed update criteria has no significant influence on the multi-scale responses of the ballast bed. Meanwhile, unnecessary updates of Verlet list resulting from only overall displacement of particles are effectively decreased. This indicates a potential way to improve the computational efficiency in GPU-based DEM for geotechnical dense granular materials under periodic loading.