A simple time stepping algorithm for Material Point Method

Wojciech T. Sołowski*, Daichao Sheng

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference article in proceedingsScientificpeer-review

Abstract

The Material Point Method (MPM) is an emerging numerical method that provides a natural solution to dynamic problems involving large deformation and contact/impact. However, the method can be prohibitively time consuming because extremely small time steps often have to be used to obtain a reliable solution. The paper presents an adaptive time stepping algorithm with error control for the material point method. The proposed approach is based on an error estimate derived from the global energy. This error estimate is then used to control the time step size. The proposed algorithm is validated for a 1D dynamic problem involving impact of two elastic bars. The proposed time stepping algorithm seems to be able to recover the correct solution with arbitrary step sizes specified by the user. As such, it is hoped that the proposed algorithm may offer a way to overcome some of the shortcomings of the MPM. A conceptually similar algorithm may be used for 2D & 3D problems with more advanced boundary conditions and material models.

Original languageEnglish
Title of host publicationNumerical Methods in Geotechnical Engineering - Proceedings of the 7th European Conference on Numerical Methods in Geotechnical Engineering
Pages157-162
Number of pages6
Publication statusPublished - 2010
MoE publication typeA4 Conference publication
EventEuropean Conference on Numerical Methods in Geotechnical Engineering - Trondheim, Norway
Duration: 2 Jun 20104 Jun 2010
Conference number: 7

Conference

ConferenceEuropean Conference on Numerical Methods in Geotechnical Engineering
Abbreviated titleNUMGE
Country/TerritoryNorway
CityTrondheim
Period02/06/201004/06/2010

Fingerprint

Dive into the research topics of 'A simple time stepping algorithm for Material Point Method'. Together they form a unique fingerprint.

Cite this