EFFICIENCY ANALYSIS FOR THE CONTACT DETECTION ALGORITHMS IN THE DISCRETE ELEMENT METHOD

The proposed article provides a brief overview of publications devoted to solving problems of granular media mechanics using the discrete element method. These publications discuss the conditions for the successful application of algorithms and software developed for the effective functioning of computing systems implementing the discrete element method. It is noted that the most labor-intensive aspect of modeling deformation processes using the discrete element method are calculations associated with determining contact forces. The key problem is the evolutionary nature of contacts: during deformation of the medium under study, discrete elements can come into contact, which can subsequently be interrupted depending on the loading conditions. In this regard, it is necessary to carefully study the organization and management of memory associated with computer modeling of contacts. The main attention in the article is paid to algorithms for searching for discrete elements coming into contact. These algorithms should be robust (i.e. reliable), easy to implement; effective in terms of speed and use of RAM. Two variants of the algorithm based on the use of auxiliary (screening) grids are compared: the screening contact detection algorithm and the Munjiza-NBS contact detection algorithm. A brief description of these algorithms and their software implementation is given. The results of verification of the algorithms on test problems and a comparative analysis of their performance are presented. It is shown that the selected algorithms provide reliable information about discrete elements contacting during motion and can be successfully used in the software implementation of the discrete element method as applied to problems of granular media dynamics.