Second‐order convected particle domain interpolation (CPDI2) with enrichment for weak discontinuities at material interfaces

Abstract

SUMMARYConvected particle domain interpolation (CPDI) is a recently developed extension of the material point method, in which the shape functions on the overlay grid are replaced with alternative shape functions, which (by coupling with the underlying particle topology) facilitate efficient and algorithmically straightforward evaluation of grid node integrals in the weak formulation of the governing equations. In the original CPDI algorithm, herein called CPDI1, particle domains are tracked as parallelograms in 2‐D (or parallelepipeds in 3‐D). In this paper, the CPDI method is enhanced to more accurately track particle domains as quadrilaterals in 2‐D (hexahedra in 3‐D). This enhancement will be referred to as CPDI2. Not only does this minor revision remove overlaps or gaps between particle domains, it also provides flexibility in choosing particle domain shape in the initial configuration and sets a convenient conceptual framework for enrichment of the fields to accurately solve weak discontinuities in the displacement field across a material interface that passes through the interior of a grid cell. The new CPDI2 method is demonstrated, with and without enrichment, using one‐dimensional and two‐dimensional examples. Copyright © 2013 John Wiley & Sons, Ltd.

Evaluation of material point method for use in geotechnics

“…and the convected particle domain interpolation scheme of Sadeghirad et al . (for some more insights on available interpolations see ). In this paper, the GIMP version of the material point method has been adopted, primarily because of the quality and robustness of its implementation in the open source Uintah software (available from Utah University).…”

“…In 2004, Bardenhagen and Kober [10] introduced the generalised interpolation material point method (GIMP), which improved the method greatly. Despite some shortcomings, the GIMP method remains the most popular version of the approach, although some notable new variants are now available, such as the weighted least squares implementation of Wallstedt and Guilkey [11], the versions developed by Ma et al [12] and the convected particle domain interpolation scheme of Sadeghirad et al [13,14] (for some more insights on available interpolations see [15]). In this paper, the GIMP version of the material point method has been adopted, primarily because of the quality and robustness of its implementation in the open source Uintah software (available from Utah University).…”

A convected particle least square interpolation material point method

“…In this section, we examine the accuracy and efficiency of the MLS interpolation. Firstly, a function f (x) = sin(x) is considered on the interval [1,4]. Then, we reconstruct the material point data by setting f p = f (x p ) on the domain with 20 cells and two material point equally distributed in each cell.…”

An efficient binning scheme with application to statistical crack mechanics

“…Just as numerical methods for BVPs ultimately lead to the need to evaluate integrals over spatial domains, this work ultimately requires integrals over reference probability space. The integration scheme developed here is similar in character to the material point method (MPM), which is used to solve large-deformation history-sensitive BVPs [26,27].…”