From algorithm to engineering judgment: navigating topology optimization's limits in crash applications

The non-linear nature of crash applications has led to many designs being developed through extensive trial and error based on the intuitions of the design engineer, as such using topology optimization for crash applications can provide major improvements in cost, weight, and passenger safety. Topology optimization is known for creating stiff, lightweight structures, however its application to crash scenarios must be handled carefully. Compliance minimization, the most common optimization objective, can yield misleading designs that prioritize undesirable qualities when developing structures for crash applications. In this paper the design optimization process of a passenger seat assembly subject to both an enforced displacement phase into a crash deceleration phase is discussed Due to the conflicting nature of compliance minimization and enforced displacement, the design was split into two types of regions; sacrificial, which are regions manually designed to absorb the majority of the enforced displacement and crash energy, and structural, regions designed with optimization tools to maximize stiffness and reduce mass. Often a component that failed during the crash phase was remedied via the removal, rather than addition, of material in key sacrificial locations through alleviating stresses on the component during the enforced displacement phase. Identifying and using these design guidelines was shown to greatly improve seat performance metrics, and reduced design times as failure mechanisms were clearly defined and designed, mitigating the issue of cascading failures across an assembly as individual failure modes are addressed.