Speaker
Description
Subsurface corrosion inspection using active infrared thermography (IRT) has been extensively investigated in the literature, with the majority of studies focusing on rear-side inspection configurations. In these approaches, depth estimation is facilitated by the fact that the thermal response is dominated by the defect closest to the observation surface. In contrast, front-side inspection has received less attention, and existing studies have focused on defect detection in specific scenarios, without performing an analysis of the influence of inspection parameters or addressing depth estimation.
In many real-world applications, such as ship hulls, offshore structures, and large steel components, access to the rear side of the structure is not possible. As a result, front-side inspection often represents the only feasible configuration, although depth discrimination becomes more challenging, as all subsurface defects are located at similar distances from the observation surface. This work presents an experimental study evaluating the capabilities and limitations of front-side subsurface corrosion inspection using active IRT, with a focus on identifying configurations suitable for efficient integration into autonomous inspection systems.
The study is conducted using a long-wave infrared (LWIR) camera combined with halogen lamps under the long-pulse thermal (LPT) technique. Artificial corrosion defects are introduced in steel specimens to enable a controlled and reproducible evaluation. Defects with diameters of 10 mm, 5 mm, and 2 mm are produced, with depths ranging from 0.1 mm to 2.0 mm. This enables the analysis of the influence of both defect size and depth on the front-side thermal response.
All experiments are performed in reflection mode, with the camera and heating sources placed on the same side of the specimen. Several inspection configurations are evaluated by varying key parameters, including inspection distance (40 cm, 60 cm, and 100 cm), heating power (1700 W, 970 W, and 500 W), and excitation duration (20 s, 10 s, 5 s, and 2 s). Defect detectability is quantitatively assessed using the contrast-to-noise ratio (CNR). In parallel, the temporal evolution of the thermal response is analyzed to evaluate its sensitivity to defect depth, using the critical time at which the thermal responses of defect and sound areas converge.
The results show that, under suitable inspection conditions, the use of an LWIR camera with the LPT technique enables both the detection of subsurface corrosion and the estimation of depth-related information. However, as thermal excitation decreases, these capabilities degrade. They are not coupled: depth-related information is lost first, while the defect may remain detectable.
Overall, this work provides an assessment and practical guidelines for front-side subsurface corrosion inspection, clarifying the achievable detection limits and depth estimation. The results help address a significant gap in the literature and support the use of front-side thermographic inspection in scenarios where rear-side access is unavailable.
