P. Mackenzie-Helnwein, P. Arduino
University of Washington/US
Many geological and geotechnical problems, e.g., slope stability or debris flows, are treated either as a solid-dominant or a flow-like problems, respectively. However, events like earthquakes or detonations can transform a quasi-static, solid-like behavior into a fluid-like dynamic flow problem (liquefaction, etc.), and may transform back to solid-like behavior after the event passes. Similar problems arise in, e.g., processing and manufacturing of powder materials or as part of normal operation of silos.
All of these processes are characterized by very large deformations and significant non-linear behavior. The mathematical representation of such processes changes from an elliptic problem to a hyperbolic one and back. Together, this poses significant problems on numerical methods such as the FEM (mesh distortion, near incompressibility) or the Control Volume Method ("dry states" due to moving boundaries). Particle methods such as, e.g., the Material Point Method (MPM), Particle FEM, Smooth Particle Hydro-dynamics (SPH) and others were developed to aleviate shortcommings of the more tradiational methods by introducing a hybrid Eulerian-Lagrangian approach.
However, while allowing for dynamic simulations at very large inelastic deformations, these methods and their variants still show, among other problem specific issues, difficulties with near incompressible motions, lacks shock-capturing capabilities, and a need for improved stress recovery (especially with the MPM) or treatment of material memory (SPH).
This mini-symposium shall be dedicated to the advancement of various forms of particle based methods toward improved modeling capabilities for modeling of solid and fluid-like behavior with special emphasis on but not restricted to the MPM and its variants.