Damage criticality assessment in complex geometric structures using static stream response-based signal processing techniques

The use of glass-reinforced plastics (GFRP) as a structural material is widespread because of their high strength and stiffness, low mass, excellent durability and ability to be formed into complex shapes. However, GFRP composite structures are prone to delaminations which can lead to a significant degradation in structural integrity. A number of non-destructive inspection methods have been devised to inspect such structures. One class of SHM system relies on the examination of the strain distribution of the structure due to its operational loads. This paper considers the strain distribution in a GFRP structure subject to loading. The strain distribution due to delaminations of various sizes and locations along the bondline of the structure has been determined by finite element analysis (FEA). A technique called the Damage Relativity Assessment Technique (DRAT) has been developed and implemented to process the data in order to amplify the damage detection process. An Artificial Neural Network (ANN) has been trained to relate this strain distribution to damage size and location. This ANN has been shown to predict the size and location of damage for a number of simulated cases. The extension of this technique is to detect multiple cracks in a complex structure with multiple loading sets. These studies will also be carried over for structures subjected to impulse loading. A major aspect of this effort will include the pseudo-automated assessment of the criticality of the damage. Results from computational and experimental work, in this regard will be presented and used in conjunction with the DRAT and the ANN techniques described above.