Efficient Reliability Analysis and Reliability-Based Design Optimization of Mechanical Systems by Using Latin Hypercube Sampling

Considering uncertainty in engineering design is computationally more expensive than solving traditional deterministic problems. This challenge force researches to search for more efficient methods. For that purpose, in this work, the superiority of the LHS over MCS in the process of reliability analysis and RBDO of a mechanical system is investigated. Accordingly, the reliability analysis and RBDO process with both LHS and MCS is implemented separately on the tension-compression spring design problem. According to these results, in both reliability analysis and RBDO process, LHS was more stable in convergence compared to MCS. Moreover, LHS can be considered to be more efficient than MCS in RBDO of the spring problem.