UC San Diego

This paper aims to use structural health monitoring (SHM) on fiberglass reinforced plastic (FRP) panels subject to fatigue loading. The testing procedure will assess the structural integrity and predict the remaining serviceability of a panel. Fiber optic sensors embedded into the specimens collect readings after discrete fatigue cycle damage level increments which are then transformed into multidimensional damage sensitive features. The features are then placed in a mahalanobis distance space where they are examined under a positive feedback relationship. Constraints on the model, linearize the method, and finally, a data driven approach produces probabilistic time of failure predictions at each damage phase.