Multi-characteristic combination based reliability enhancement of optical bidirectional thermal wave radar imaging for GFRP laminates with subsurface defects

Abstract

In this paper, the detection reliability of optical bidirectional thermal wave radar imaging (BTWRI) for glass fiber reinforced polymer (GFRP) laminates with subsurface defects using various imaging algorithms was quantitatively investigated. A set of GFRP laminates with artificial defects were prepared and inspected. Three frequently-used imaging algorithms (cross-correlation, CC; chirp lock-in, CLI; and Hilbert transform mean, HTM) were applied to construct characteristic images. An analysis for probability of detection (POD) was carried out based on the hit/miss data obtained by comparing the defect contrasts and noise thresholds of characteristic images. A multi-characteristic combination (MCC) method integrating the advantages of each algorithm was proposed. The reliability assessment of optical BTWRI for inspecting GFRP laminate defects was compared by the defect diameter-to-depth ratio (r90/95) at 90% POD with 95% confidence level and detection rates (DRs). The comparison results show that the MCC method exhibits enhanced reliability with smaller r90/95 and higher DRs at a series of determination thresholds compared with CLI, HTM, and CC algorithms.