Development of a scanning acoustic microscopy-based structural prior for microtexture region characterization

Abstract

Nondestructive evaluation (NDE) plays a crucial role in ensuring aircraft availability. The current NDE paradigm often relies on mono-modal testing and signal-over-threshold criteria to provide robust defect or damage detection, not characterization. One example is found in the risk-based management of surface-breaking cracks in metal, where cracks of a given size can be detected by eddy current testing (ECT) with a calculable probability. Yet, there are cases where detection proves insufficient. Consider the case of microtexture regions (MTR) found in certain titanium alloys, which can increase the risk of cold dwell fatigue failure when found above a certain size and in specific orientations relative to the surrounding material. At present, the size and orientation of MTR cannot be characterized using only one NDE modality. In this work, a data fusion-based solution to MTR characterization is developed. First, two inspection methods—scanning acoustic microscopy (SAM) and ECT—are selected, where each method is individually capable of only partial characterization. Then, matching component analysis is used to develop a surrogate forward model relating MTR orientation to ECT output. This data is then inverted using boundaries provided from the SAM data as a structural prior.