• Title of article

    Reduced basis hybrid computational homogenization based on a mixed incremental formulation

  • Author/Authors

    Fritzen، نويسنده , , Felix and Leuschner، نويسنده , , Matthias، نويسنده ,

  • Issue Information
    روزنامه با شماره پیاپی سال 2013
  • Pages
    12
  • From page
    143
  • To page
    154
  • Abstract
    A new reduced basis method for the computer assisted homogenization of microheterogeneous materials is proposed. Key features of the Nonuniform Transformation Field Analysis (NTFA) [1,2] are taken as a point of departure. A short-coming of the NTFA method is its limitation to simple constitutive models on the microscale. A second disadvantage is the possible loss of accuracy when the reduced basis approximating the plastic strains has increasing dimension. Both issues are related to the simple evolution law for the mode activity coefficients, which are the new macroscopic internal variables of the homogenized material. In the present contribution a generalization of the NTFA is proposed in which the evolution of the new internal variables is derived from a mixed incremental variational formulation. The derivation is based on purely micro-mechanical considerations. The modification allows for arbitrary Generalized Standard Materials [3] on the microscopic scale including, e.g., crystal visco-plasticity. Additionally, the fidelity of the homogenized response is now directly linked to the reduced basis approximating the plastic strain field. Numerical examples for nonlinear viscous materials and single crystal plasticity outline the accuracy of the proposed multi-scale method.
  • Keywords
    Hybridcomputational homogenization , Nonuniform Transformation Field Analysis (NTFA) , Reduced basis model order reduction , Generalized Standard Materials (GSM) , Mixed incremental variational formulation
  • Journal title
    Computer Methods in Applied Mechanics and Engineering
  • Serial Year
    2013
  • Journal title
    Computer Methods in Applied Mechanics and Engineering
  • Record number

    1595996