Emissivity-independent spatiotemporal thermal anomaly detection in metal components using transmission-mode phase measuring deflectometry

Abstract

Non-destructive evaluation of components often relies on infrared thermography, yet metallic and highly conductive surfaces provide weak radiative contrast and require emissivity calibration. We present an emissivity-independent optical approach for rapid spatiotemporal heat-flow mapping and subsurface anomaly detection using transmission-mode phase measuring deflectometry (PMD). A sinusoidal fringe field is imaged through a transparent sensing medium in thermal contact with a thermally stimulated specimen. Thermal diffusion into the sensing medium creates transient temperature gradients; via the thermo-optic effect, these gradients induce refractive-index gradients that modulate the fringe field, enabling an estimation of the temperature-gradient field from the retrieved phase. A coupled experimental–numerical framework is used for quantitative validation: COMSOL-predicted and PMD-derived ROI-averaged vertical temperature-gradient show close agreement over the 0–3 s early-time window, with a root-mean-square error of 0.074 K/mm. Varying the applied thermal load from 5 K to 55 K yields an approximately linear phase response with a system responsivity of 0.0399 rad per unit applied ΔT_app (defined between the hot-water reservoir and the imaging chamber at t=0). Demonstrations on representative anomaly classes (dissimilar-material interface, localized low-conductivity insert, and void-type discontinuity) show clear localization of subsurface thermal anomalies within the first few seconds after stimulation, supporting rapid anomaly indication when radiative techniques are constrained.