Advances in structural damage diagnostic research: Architectural design, modeling and validation

摘要

Physical damages like thermally grown oxide (TGO) layer develop in the coated gas turbine blade during the aeroengine operation and severely limits the life and performance. This life critical damage data has been considered for the validation of a diagnostic fault analysis. High level architecture, statistical algorithm and experimental modeling of oxide growth data are discussed. Experimental measurement of the damage growth has been carried out under different time, temperature and processing conditions in coated blade to generate large volume of input data. Wilcoxon rank sum analysis is carried out to check its effectiveness. The proposed method appears to yield consistent result in operational data anomaly and damage diagnosis. The algorithm for prognostic analysis using exponential smoothing of damage data from gas turbine engine and nonlinear regression method will be applied for future work.