Probabilistic Risk Prediction for Aircraft Fatigue Life Management Using SHM Systems Considering the Effect of Inspection Correlation

Abstract

A numerical study was conducted to investigate the effect of inspection correlation of a SHM (Structural Health Monitoring) system on the risk prediction. The study was based on a new Probabilistic Structural Risk Assessment framework which accounts for statistical correlations among inspection outcomes from multiple SHM sensors mounted on the same structural component. The core of this framework is an innovative, computationally efficient probabilistic method termed RPI (Recursive Probability Integration). RPI is suitable for damage tolerance and riskbased maintenance planning using inspection results from either NDI (Nondestructive Investigation) or SHM systems. The RPI methodology is capable of incorporating a wide range of uncertainties including material properties, repair quality, crack/damage growth related parameters, loads, probability of detection, and inspection correlation. The result of the study demonstrates that: (1) the inspection correlation has significant effect on the risk prediction and cannot be ignored, (2) relative to independent inspections commonly assumed, the inspection correlation has a negative impact to structural risk, and (3) the stronger the inspection correlation, the larger the risk the structural component may experience.