Reliability-based design optimization of rotating FGM cylindrical shells with temperature-dependent probabilistic frequency constraints

In this study, the reliability-based design optimization (RBDO) of rotating functionally graded cylindrical shells (FGCSs) subjected to temperature dependent probabilistic frequency constraints is investigated. The uncertain parameters such as FG material properties and the shell thermo-mechanical loads are considered as random variable in RBDO. To decrease model uncertainty, the effect of initial thermal stresses are efficiently included in elasticity-based vibration equations of the shell and the powerful differential quadrature method (DQM) is employed to accurately determine the shell frequency parameters. An efficient RBDO framework based on the hybrid weighted average simulation method (WASM) and DQM together with the particle swarm optimization (PSO) method is then presented for design optimization of the FGCSs. A key feature of proposed hybrid DQ-WASM-PSO is that only one simulation run is required for WASM during entire optimization process of the FGCS, even if the distribution type of input variables and/or the system target reliability level be changed. The influence of temperature rise, temperature-dependence of FGM properties, annular velocity, PDFs of the random variables, and convection heat transfer coefficient of the shell inside fluid on the RBDO results are carried out. Parametric study indicates that exact evaluation of the initial thermal stresses, temperature-dependence of the FGM properties and convection heat transfer coefficient have considerable effect on optimization results.