AI based fatigue and live cycle analysis of complex materials
Significant Reduction in Composite Simulation Time with STOCHOS
In collaboration with SGL Carbon and DLR, we have successfully demonstrated a more efficient approach to multi-scale composite simulation.
Starting with CT-based digitalization of the microscale structure, we used our probabilistic ML software STOCHOS to create generalized material models. These models are then integrated into macro-level FEM simulations to predict part behavior with high fidelity.
Where conventional FE² methods typically require several months up to a year of computation time, our ML-based solution reduces this to just a few minutes—with no compromise in predictive quality.
This advancement enables more efficient simulation workflows and supports faster, data-driven decision-making in composite part design.
