Learn Material Point Method (MPM): A hands-on training and coding session

About the Webinar

This training aims to provide participants with a comprehensive introduction to the Material Point Method (MPM) and its applications in large-deformation modeling. Participants will gain hands-on experience by coding a 1D MPM using Python and utilizing the CB-Geo MPM to simulate a column collapse problem on DesignSafe. We will use ParaView to visualize the results.

Pre-requisites: Basic Python (numpy)

The Material Point Method (MPM) is a computational technique that has gained significant traction in large-deformation modeling. One of the standout features of MPM is its ability to handle the complexities of large deformations without the mesh entanglement issues commonly encountered in traditional Finite Element Methods (FEM). MPM employs a dual representation of the material: particles (Lagrangian) carry material information, while a background mesh (Eulerian) is used for computations, making it adept at handling problems with substantial deformations, such as landslides, tsunamis, and snow avalanches.

Training Outline:

1. Introduction to MPM
2. Coding a 1D MPM in Python
3. Running CB-Geo MPM on DesignSafe
4. Visualization with ParaView

By the end of this training, participants will have a solid understanding of the Material Point Method, its advantages in large-deformation modeling, and practical experience in implementing and visualizing MPM simulations.

Presenter

Krishna Kumar is an Assistant Professor at the Civil, Architecture, and Environmental Engineering at the University of Texas at Austin. Krishna completed his Ph.D. from the University of Cambridge in January 2015 on multi-scale multiphase modeling of granular flows. Krishna’s work involves developing exascale micro and macro-scale numerical methods: Material Point Method, Lattice Boltzmann - Discrete Element coupling, Finite Element Method, and Lattice Element method. His work in high-performance computing in geomechanics provides insights into the mechanics of natural hazards such as landslides. Krishna uses Machine Learning (ML) to model multi-scale problems in geomechanics. Krishna is a Software Sustainability Institute Fellow, UK, and has developed many open-source research codes. Krishna also builds large-scale graph networks and agent-based models for simulating the resilience of city-scale infrastructure systems.