Reliability-based design optimization method of turbine disk with transformed deterministic constraints

Abstract

To improve the computational efficiency of the reliability-based design optimization (RBDO) of a complex structure with nonlinear and implicit limit-state function, the single-loop-single-vector (SLSV)-limit-state factor (LSF) (SLSV-LSF) method was developed by fully considering the advantages of the SLSV approach and the LSF method to transform uncertain constraints into deterministic constraints. The mathematical models of SLSV and LSF were established and the basic RBDO process of the SLSV-LSF method is presented. The shape optimization of an aeroengine turbine disk was completed based on the proposed method. From the reliability sensitivity analysis of the turbine disk, it is revealed that an uncertain constraint of average circumferential stress can be transformed into a deterministic constraint and material density can be regarded as a deterministic variable. Through the min-mass shape design of the turbine disk based on different approaches, it is demonstrated that the developed method maintains high computational speed and efficiency while keeping maintaining computational accuracy, which validates the feasibility and validity of the SLSV-LSF method in the RBDO of aeroengine typical components.