Contributions in robust optimisation of complex mechanical structures

Abstract: Engineering systems are commonly modelled using high-fidelity computational frameworks, which rely on large models and complex solvers that can predict complex phenomena. These models are often computationally expensive and subject to multiple sources of uncertainty, which can significantly affect their predictions. So, it is crucial to incorporate these uncertainties into the analysis and design of engineering structures. In this context, methods for uncertainty propagation, calibration and robust design are particularly challenging due to the high numerical cost of the solver. These problems can become even more complex when some quantities of interests present complex behaviours (jumps, multi-stability, not always defined, etc.).

After introducing the general framework of uncertainty management for computational models together with current research directions, this talk will focus on two recent works on optimisation:

1)    Robust topology optimisation for metallic additive manufacturing (AM): significant material variability is observed in metallic AM due to the 3D-printing process, which impacts the final mechanical performances of the printed parts. We formulate the design problem as a robust topology optimisation problem under uncertainty, where material variability is explicitly incorporated into the topology optimisation framework, leading to designs that are less sensitive.

2)    Parametric optimisation of bifurcation diagrams: bifurcations occur in the dynamics of many natural and engineered systems. This work investigates how the optimisation of design parameters can be used to control bifurcation locations. To this end, an objective function that encourages the appearance of bifurcation points at targeted locations is proposed. The framework is applied to several mechanical systems.

‼️ Le séminaire peut être suivi en ligne via le lien : https://univ-amu-fr.zoom.us/j/95011527683?pwd=bTeNvaaZY75buaa7bRL0pcucXSASm7.1

Le mardi 26 mai 2026 à 11h00 / Amphithéâtre François Canac, LMA

Voir en ligne : plus d’informations concernant l’orateur