LAUSR

Flaw profile characterization from nondestructive evaluation measurements is a typical nondestructive testing (NDT) inverse problem. Flaw profiling, particularly in aerospace industry, is important, as it is a decision tool to evaluate the air worthiness of the aerostructures. Accurate flaw profiling thus ensures aircraft safety through the timely implementation of cost-effective replacement/repair actions. Eddy current (EC) NDT data acquired at multiple frequencies contain complementary information about the flaws due to the skin effect phenomenon. However, finding the exact contribution of each measurement mode while determining the solution to the inverse problem is considerably challenging. In the reported research work, a novel multifrequency EC data-based polynomial model is formulated to solve the inverse problem. The weight (coefficient) of each measurement mode in the polynomial model is computed using nonlinear optimization techniques. Actual NDT data of a retired aircraft containing mutually exclusive training and test databases has been used as a case study. Data fusion (DF)-based inversion results are compared to the results acquired using individual measurement mode data. The improved results with DF indicate the efficacy of the proposed technique.