TY - CHAP

T1 - Automatic generation and meshing of periodic representative volume elements for matrix-inclusions composites

AU - Schneider, Konrad

AU - Klusemann, Benjamin

AU - Bargmann, Swantje

N1 - Conference code: 1

PY - 2016

Y1 - 2016

N2 - Modern lightweight structures often feature composite materials to meet the requirements of todays engineering applications. The underlying microstructures of these heterogeneous materials have a vital influence on the mechanical properties and deformation behavior. From a computational viewpoint it is necessary to incorporate this information for most exact simulations. To gain deeper insides in the micromechanical behavior state of the art methods arising from the field of computational micromechanicstypically utilize representative volume elements (RVE). The efficiency and trustworthiness regarding numerical homogenization heavily depends on the choice of the RVE and its discretization in terms of quality and representativeness.Concentrating on matrix-inclusion composites (e.g. polymer blends or short fiber reinforced materials) we demonstrate a novel method for the automatic generation of high quality periodic RVEs up to the finite element discretization. Kernel of the generation process is a random sequential adsorption algorithmthat yields the geometry of the microstructure. It incorporates a successive build up of geometric entities employing admissibility checks on the inclusions to meet the non-overlapping constraint. For the purpose of numerical homogenization in the framework of finite element simulations we seek a highquality discretization of the RVE geometry with as few elements as possible while maintaining a high element quality. Special emphasis is put into generating a periodic mesh topology for application of the favourable periodic boundary conditions. The developed meshing algorithm features an individualtreatment of all inclusions and a hierarchical mesh generation. Furthermore, systematic comparisons to standard tools show the benefits of the suggested approach.

AB - Modern lightweight structures often feature composite materials to meet the requirements of todays engineering applications. The underlying microstructures of these heterogeneous materials have a vital influence on the mechanical properties and deformation behavior. From a computational viewpoint it is necessary to incorporate this information for most exact simulations. To gain deeper insides in the micromechanical behavior state of the art methods arising from the field of computational micromechanicstypically utilize representative volume elements (RVE). The efficiency and trustworthiness regarding numerical homogenization heavily depends on the choice of the RVE and its discretization in terms of quality and representativeness.Concentrating on matrix-inclusion composites (e.g. polymer blends or short fiber reinforced materials) we demonstrate a novel method for the automatic generation of high quality periodic RVEs up to the finite element discretization. Kernel of the generation process is a random sequential adsorption algorithmthat yields the geometry of the microstructure. It incorporates a successive build up of geometric entities employing admissibility checks on the inclusions to meet the non-overlapping constraint. For the purpose of numerical homogenization in the framework of finite element simulations we seek a highquality discretization of the RVE geometry with as few elements as possible while maintaining a high element quality. Special emphasis is put into generating a periodic mesh topology for application of the favourable periodic boundary conditions. The developed meshing algorithm features an individualtreatment of all inclusions and a hierarchical mesh generation. Furthermore, systematic comparisons to standard tools show the benefits of the suggested approach.

KW - Engineering

M3 - Published abstract in conference proceedings

SP - 8

BT - Data-driven modeling and numerical simulation of microstructured materials (GAMM AG DATA)

A2 - , GAMM Activity Group

PB - Universität Stuttgart

T2 - GAMM Activity Group: Data-driven modeling and numerical simulation of microstructured materials- GAMM AG Data 2016

Y2 - 26 September 2016 through 27 September 2016

ER -