Big data for damage characterization of aerospace structures

Structural health monitoring (SHM) of aerospace structures can ensure aircraft safety and keep maintenance costs down. The purpose of this work is to develop an effective damage characterization framework for identifying various features of damage inside aerospace structures, based on a large data library of damage scenarios generated by the nonlinear guided wave simulation tool UM/LISA. UM/LISA builds the large database of simulation cases with virtually all possible damage scenarios, and then matching pursuit method is adopted to correlate test signals with the database to obtain the information about the damage. Big data analysis will be performed on the database to validate the characterization framework and provide insights in the design of damage detection systems.