Identify structural flaw location and type with an inverse algorithm of resonance inspection

Abstract

To evaluate a structural component’s fitness for service and quantify its remaining useful life, aging and service-induced structural flaws must be quantitatively determined in service or during scheduled maintenance shutdowns. Resonance inspection (RI), a non-destructive evaluation (NDE) technique, distinguishes the anomalous parts from the good parts based on changes in natural frequency spectra. Known for its numerous advantages, e.g., low inspection cost, high testing speed, and broad applicability to complex structures, RI has been widely used in the automobile industry for quality inspection. However, compared to other contemporary direct visualization-based NDE methods, a more widespread application of RI faces a fundamental challenge because such technology is unable to quantify the flaw details, e.g., location, dimensions, and types. In this study, the effectiveness of a maximum correlation-based inverse RI algorithm on a variety of common structural flaws, e.g., stiffness degradation, voids, and cracks, either in monotype or the coexisting form, has been systematically investigated. The prediction results are found to be able to accurately locate the damages and quantitatively measure the physical characteristics of the defects, which can effectively help retrieve the actual state of health of the engineering structures in a computationally efficient way.