In lightweight construction – as well as in the automotive and consumer segments (e. g., drill housings) – fiber reinforced plastics are increasingly replacing metals as the work material. Predicting the strength and damage effects of these components regarding the directional dependency of the mechanical material characteristics is complicated. Therefore, complex multiscale simulations are required for accurate predictions. The computational time (CPU time) and the computer memory requirements are very high for these multiscale simulations.
Compound Research Project MuSiKo
At ITWM, we are working on methods to reduce these high computational efforts on the basis of so-called configurational forces. With these forces, macroscopic indicators can be defined, that will allow to take into account the microstructure precisely only in necessary subdomains of the component. This is the focus of the collaborative research project MuSiKo, funded by the BMBF.
Standard forces describe the external factors that act to deform a fixed object. The configurational forces, in contrast, capture the effect of the microstructural inhomogeneity on the deformation. We use simple and fast homogenization methods for subdomains of the component where the influence of the microstructure is insignificant, whereas, in areas where the microstructure has a greater influence on the strength, more expensive and more accurate methods are required. Depending on the subdomain, these may include: boundary element methods, finite element methods, or the Fast Fourier Transformation of the Lippmann-Schwinger equation. Having this adaptive choice of microstructural solvers, a very high accuracy can be achieved with lower requirements for memory and CPU.
Together with the Chair of Applied Mechanics at TU Kaiserslautern, Fraunhofer ITWM is developing indicators for the selection of a micro solver based on configurational forces and, also, an interface for the scale coupling. ITWM is developing a fast microstructure solver called “Feelmath” for use with areas that require the high resolution of a complex microstructure.