Multi-fidelity optimization of metal sheets concerning manufacturability in deep-drawing processes

Abstract Multi-fidelity optimization, which complements an expensive high-fidelity function with cheaper low-fidelity functions, has been successfully applied in many fields of structural optimization. In the present work, exemplary cross-die deep-drawing optimization problem is investigated to compare different objective functions and assess performance a multi-fidelity efficient global technique. To that end, hierarchical kriging combined infill criterion called variable-fidelity expected improvement. Findings depend significantly on choice function, highlighting importance careful consideration when defining function. We show one based share bad elements forming limit diagram not well suited optimize example problem. contrast, two other definitions average sheet thickness reduction averaged violation diagram, confirm potential approach. They reduce computational cost at comparable result quality or even improve compared single-fidelity