We propose a variational h-adaption strategy in which the evolution of the mesh is driven directly by the governing minimum principle. This minimum principle is the principle of minimum potential energy in the case of elastostatics; and a minimum principle for the incremental static problem of elasto-viscoplasticity. In particular, the mesh is refined locally when the resulting energy or incremental pseudo-energy released exceeds a certain threshold value. In order to avoid global recomputes, we estimate the local energy released by mesh refinement by means of a lower bound obtained by relaxing a local patch of elements. This bound can be computed locally, which reduces the complexity of the refinement algorithm to O(N). We also demonstrate how variational h-refinement can be combined with variational r-refinement to obtain a variational hr-refinement algorithm. Because of the strict variational nature of the h-refinement algorithm, the resulting meshes are anisotropic and outperform other refinement strategies based on aspect ratio or other purely geometrical measures of mesh quality. The versatility and rate of convergence of the resulting approach are illustrated by means of selected numerical examples. Copyright © 2007 John Wiley & Sons, Ltd.
CITATION STYLE
Mosler, J., & Ortiz, M. (2007). Variational h-adaption in finite deformation elasticity and plasticity. International Journal for Numerical Methods in Engineering, 72(5), 505–523. https://doi.org/10.1002/nme.2011
Mendeley helps you to discover research relevant for your work.