Model-Augmented Pattern Recognition for Damage Detection in Aerospace Materials

Currently successful laboratory non-destructive methods are impractical for inspection of integrated aerospace structures. To resolve this issue, the Metis Design Corporation (MDC) has developed structural health monitoring (SHM) system sensors and hardware. SHM is an emerging technology leading to the development of systems capable of continuously monitoring damage with minimal human intervention. During Phase I research MDC used piezoelectric actuator/sensor pairs to perform Lamb wave testing on composite plates with drilled holes, delamination and impact damage. Subsequently pattern recognition methods were used to classify the presence, type and severity of damage. This method proved very successful, demonstrating 100% accuracy in determining damage presence and type, and 99.9% for rough damage severity across 9000 tests. Phase II research leverages these results to scale this methodology to real built-up structures. MDC will work in parallel with Boeing on the AFRL/VA funded “Hot Spots” program, focusing on the application of pattern recognition methods to fault-prone composite and titanium components on ageing F-16’s. The first goal will be to evaluate the limitations of this methodology, including minimal damage size and discretized severity ranges. The second goal will be to develop analytical and/or FEA techniques to minimize the required physical training.