A New Approach for the Reliability-Based Robust Design Optimization of Mechanical Systems under the Uncertain Conditions
Khodaygan, S. - Sharafi, M.H.
A mechanical system inherently affected by the conditions, factors, and parameters of uncertainties. Without including the uncertainty effects in the design procedure, the designs may not be robust and reliable. Robust design optimization (RDO) method is a procedure to find the insensitive design with respect to the variations. On the other hand, reliability is measured by the probability of satisfying a specific design criterion. Therefore, a reliable design is a design that satisfies the specified criteria even with some uncertainties in variables and parameters. Reliability-based design optimization (RBDO) is an optimization procedure that incorporates reliability requirements to find the proper design. Since RDO and RBDO are usually the expensive computational approaches, the Reliability-Based Robust Design Optimization (RBRDO) may be difficult to apply. In this paper, a new model for the reliability based robust design optimization is introduced. First, two new factors that are called "chance - penalty function" and "reliability multiplier" are introduced. Based on these new factors, the optimality, the robustness, and the reliability functions as three objective functions are modeled. Then, a combined model for the RBRDO of the mechanical systems under the uncertain conditions is proposed. Finally, the application of the proposed method is demonstrated through the design of two case studies under uncertain conditions, and the computational results are compared to the obtained results from classical methods. © 2018 SAE International. All Rights Reserved.