Optimal sampling for fast eddy current testing inversion by utilising sensitivity data

An adaptive sampling method based on simplex-mesh refinement is proposed for creating a forward database that can facilitate the inverse problem solution in non-destructive testing. The resulting mesh-database is used as a generic interpolator of the forward problem. The available sensitivity data are utilised for obtaining an optimal sampling with respect to the piecewise linear interpolation applied for data retrieval. The resulting database provides guaranteed quality of approximation using a relatively few number of samples. The method is illustrated by a crack reconstruction problem of eddy current testing.