The purpose of the research activities is to develop mathematical models, numerical algorithms and computer software in order to numerically simulate the deformation, damage and failure of engineering materials. Accurate material modeling is essential in order to predict the performance of actual parts and structures made of those materials.
The research focuses on homogenization methods and multiscale models, which enable to predict the influence of the micro-structure on the macroscopic or engineering properties. The research activities concern several classes of composite materials (with continuous, long or short fibers, with particles or platelets), porous materials, multiscaling-based ICME (integrated computational materials engineering), damage and failure models, high cycle fatigue, model reduction based on machine learning methods.
The homogenization and multiscaling approach achieves several targets of scientific and industrial importance.
The approach allows insight into materials and to systematically understand mechanisms that dominate the macroscopic material properties arising from the microscopic composition.
Based on this understanding, it is possible to further identify promising candidates for new composite materials, thereby reducing the amount of experimental effort needed. This helps to save money and reduce the time required to develop new materials.
In industrial applications the approach supports the design of composite structures by providing high quality micromechanical material models that can be used in integrative simulations to describe the performance of the part in the best way possible.
The approach also enables to identify unknown in-situ per-phase properties following a reverse engineering strategy. Thus, the unknown constituents’ parameters are varied in a characteristic physical range to match the global performance of the material as observed in anisotropic measurements.