Abstract:
This paper presents a methodology for solving a worst-case design optimization problem, specifically focusing on head impact in automotive design. The core of the method is an optimization algorithm capable of finding saddlepoints in the design space. This is achieved by simultaneously maximizing the objective function with respect to "worst-case" design variables and minimizing it with respect to "nominal" design variables. The approach is validated using two analytical functions, one of which includes a random noise component and an integer variable, to simulate the complexities found in real-world engineering problems. The methodology is then successfully applied to a crashworthiness head impact problem, demonstrating its power and effectiveness in the field of automotive occupant safety design.
