Efficient Multi-Scale Methods for Short Fiber Reinforced Plastics

Material Models for the Development of Digital Twins

The fundamental step of this method is to characterize the mechanical behavior on the microscale (see also AIF failure). Together with our project partners from Bosch, we examined the viscoelastic and fatigue behavior of short glass fiber reinforced thermoplastic samples. Our project partners provided mechanical measurements and CT images to analyze the behavior of samples with specific fiber orientations to us. On this basis, we developed suitable material models to produce digital twins of the composite on the microscale of the sample.
 

A Two-Step Approach Including Full-Field Simulations and Model Order Reductions

We developed a two-step approach to reduce the numerical effort of coupled FEM-FFT multiscale method. In the first step, we used a highly efficient microscale solver, our software tool FeelMath, to perform full-field simulations on representative elementary volumes of the microstructure for sample fiber orientations. Using model order reduction technologies we then obtain effective material models for these fiber orientations and put them them into a database By means of  this database concept we drastically reduce the numerical effort, so that, for the first time, the multiscale method can be applied to industrial problems and thus, can be used profitable by our partner.