Fatigue-damage prediction with GeoDict

Short-fiber reinforced polymers exhibit a progressive stiffness degradation and subsequent failure under high cycle fatigue (HCF) loading. To enable a reduction of experimental measurements it is imperative to understand the HCF damage process on the microstructural level.

In this talk, we demonstrate an approach to predict HCF damage on the fiber structural level using GeoDict. The matrix material is equipped with a non-localizing HCF damage model to account for the stiffness degradation under cyclic loading. Based on the local degradation in the matrix material, the progressive anisotropic stiffness loss of the composite material is obtained. The HCF damage  model for the matrix material is integrated into the GeoDict-solver FeelMath via a ABAQUS compatible user subroutine (UMAT). A comprehensive short-tutorial will be given to demonstrate the use of UMATs in GeoDict.

Based on the microstructural computations, an effective macroscopic material law can be derived to enable efficient computations on engineering components of industrial size. An overview on this upscaling procedure will be given.