Computational & Technology Resources
an online resource for computational,
engineering & technology publications |
|
Civil-Comp Proceedings
ISSN 1759-3433 CCP: 94
PROCEEDINGS OF THE SEVENTH INTERNATIONAL CONFERENCE ON ENGINEERING COMPUTATIONAL TECHNOLOGY Edited by:
Paper 83
Homogenization of Perfusion in a Large-Deforming Medium using the Updated Lagrangian Formulation E. Rohan and V. Lukeš
Department of Mechanics, Faculty of Applied Sciences, New Technologies Research Centre, University of West Bohemia, Pilsen, Czech Republic , "Homogenization of Perfusion in a Large-Deforming Medium using the Updated Lagrangian Formulation", in , (Editors), "Proceedings of the Seventh International Conference on Engineering Computational Technology", Civil-Comp Press, Stirlingshire, UK, Paper 83, 2010. doi:10.4203/ccp.94.83
Keywords: porous medium, perfusion, large deformation, Biot model, homogenization, multiscale modeling.
Summary
In this paper we explain our homogenization-based approach to the multiscale modeling of large deforming fluid saturated materials. Such problem arises in many engineering applications involving the mechanics of soils or tissue biomechanics, namely in the tissue blood perfusion problem.
We consider a fluid-saturated poroelastic medium with large heterogeneities in the elastic and permeability coefficients. In the linear-elastic case, analysis of such medium was treated by homogenization so that the effective material parameters involved in the macroscopic model were obtained in advance for given specific microstructures using solutions of auxiliary microscopic problems [1]. In the nonlinear case the microstructure changes with progressing deformation, so that it must be updated in time, as also discussed in [2,3]. We propose to combine the linear homogenization approach and the time discretization with the local reference microstructures updated from their previous deformation history. The new macroscopic response of the time increment is obtained using a homogenized linear subproblem. The computational algorithm can be characterized as the cycle of the following steps:
This multiscale analysis leads to the split of the computational procedure into two levels: at the macroscopic level information from all local configurations is collected and the computation of the macroscopic response is performed, at the microscopic level the auxiliary problems are solved independently for each LC. We provide a numerical illustration of this approach. References
purchase the full-text of this paper (price £20)
go to the previous paper |
|