Reliability-Based CCF Damage Analysis for Gas Turbine Blade With Thermal Barrier Coatings

Abstract

In this work, a dynamic surrogate modeling approach is presented for reliability analysis of turbine blades with thermal barrier coatings (TBCs) under combined high and low cycle fatigue (CCF) loadings. Initially, a three-dimensional model encompassing TBCs, turbine blades and flow fields is built to investigate the stress distribution at turbine blade surface using numerical analysis method of fluid–thermal–solid coupling. Following that, an improved seagull optimization algorithm-based backpropagation neural network (ISOA-BPNN) is developed by integrating the strengths of seagull optimization algorithm (SOA) and BP neural network. Furthermore, the probabilistic CCF estimation of turbine blades with TBCs is considered as a numerical case to evaluate the developed approach under the consideration of the uncertainties in material properties and loading conditions. The results reveal that the application of TBCs reduces the maximum stress at the blade mortise position, and the proposed ISOA-BPNN holds great prediction accuracy and computational speed for reliability analysis.