Segmentation of experimental eddy current testing data via matching component analysis

Microtexture regions (MTRs) are collections of grains with similar crystallographic orientation. When present in aerospace components, they can potentially limit component life. As such, a non-destructive evaluation (NDE) method to detect and characterize MTR is desired. One potential solution is to use an electromagnetic NDE method known as eddy current testing (ECT), which is sensitive to local conductivity variations associated with MTR. Recent work has shown that MTR boundaries and orientation can be determined from ECT data using a variant of matching component analysis (MCA) combined with a regularization method originally developed for image deblurring. However, this method has only been demonstrated on simulated ECT data. In this work, we apply the previously developed method to experimental ECT data of a large grain titanium specimen. We show that we are able to determine grain boundaries and orientation from experimental ECT data, serving as a first step to full MTR characterization. This article is part of the theme issue ‘Frontiers of applied inverse problems in science and engineering’.