Search Results

This paper presents a framework to incorporate the notion of reliability into crashworthy designs of automotive vehicle components. Optimal design for crashworthiness is a challenging task in itself as it involves time dependent complex interactions among bodies along with material and geometric nonlinearities. Maximum energy absorption in the structure is widely used as a design criteria in crashworthiness designs as long as penetration levels are kept under allowable limits. These designs behave well and are safe as long as loading conditions and material properties are deterministic however failure can happen due to excessive penetration under uncertain conditions. Hence a semi coupled reliability based crashworthiness design methodology is proposed in which independent reliability assessments are done on the designs at various intermediate levels during an iterative design cycle of a crashworthy structure.

In this investigation a robotic system's dynamic performance is optimized for high reliability under uncertainty. The dynamic capability equations allow designers to predict the dynamic performance of a robotic system for a particular configuration (i.e.,point design). While the dynamic capability equations are a powerful tool, they can not account for performance variations due to aleatory uncertainties inherent in the system. To account for the inherent aleatory uncertainties, a reliability-based design optimization (RBDO) strategy is employed to design robotic systems with robust dynamic performance. RBDO has traditionally been implemented as a nested multilevel optimization process in which reliability constraints require solution to an optimization problem (i.e., reliability analysis). In this work a robust unilevel performance measure approach(PMA) is developed for performing reliability-based design optimization which eliminates the lower level problem in RBDO.