Robust topology optimization

Additive Layer Manufacturing (a.k.a. 3D printing) allows manufacturing complex, topologically optimized structures. Topology optimization methods provide "bionic" structures, which contain material only where it is required for load bearing.

By tailoring structures to a certain load scenario, classical optimization methods often yield designs, which are quite sensitive to scattering properties such as material parameters or load direction. Robust design optimization techniques provide design solutions, which are lightweight and at the same time robust with respect to uncertainties. The obtained designs often differ significantly from the ones obtained by classical optimization approaches.

Combining topology optimization and robust design optimization is a relatively new field of research with a lot of open questions on how to implement the combination of these approaches.

Probalistic analysis and optimization of composites

When manufacturing fiber composites, the material is made at the same time as the structure. This is one root cause of the relatively large scatter observed in the manufacturing quality. Furthermore, laminated composites are sensitive to impact damages. When sizing fiber composite structures in practice these effects are accounted for by knockdown factors. The concept of knockdown factors implies the assumption that all worst case scenarios occur at the same time (e.g. highest load, worst material quality, worst impact damage, …).

Probabilistic analyses enable predicting the stochastic distribution of a structural response (e.g. failure load) due to the scatter of input variables (e.g. material parameters, load, …). Hence, they allow accounting for the joint probability of occurrence of different sources of uncertainty. In the design of composites structures probabilistic approaches provide huge potential to reduce conservatism and therefore, to design more lightweight.

Fiber composite structures provide a huge potential for optimization, because the layup can be adjusted to the loading situation. Hence, they can also be designed for the best compromise of structural performance and robustness. This is achieved by embedding probabilistic analyses of composites in a robust design optimization framework.