Optimization of infrared thermography investigations for identification of subsurface anomalies in plastered masonry walls

This study investigates the optimization of Infrared Thermography (IRT) for detecting subsurface anomalies in plastered masonry walls typical of heritage buildings, including those restored with traditional, modern, and innovative coatings. Despite widespread application of IRT in non-destructive testing (NDT) within cultural heritage diagnostics, standardized protocols for evaluating plastered and upgraded surfaces—particularly involving materials such as reinforced concrete (RC) and Textile Reinforced Mortar (TRM)— are lacking. Despite widespread application of IRT in non-destructive testing (NDT) within cultural heritage diagnostics and the availability of research studies on the subject, standardized protocols for evaluating plastered and upgraded surfaces—particularly involving materials such as reinforced concrete (RC) and Textile Reinforced Mortar (TRM)—are lacking. To address this gap, three full-scale masonry mock-ups (rubble stone, brick, and mixed masonry) were constructed and coated with traditional fresco layers, hydraulic lime mortars, or cement-based plasters. Controlled defects—including Teflon inserts, wet sponges, metallic elements, and timber inclusions—were embedded at depths ranging from 3 to 50 mm. Independent IRT assessments were conducted by two research units under both passive (solar) and active (artificial) heating regimes. Results demonstrate that passive IRT, driven by a solar-induced thermal gradient of approximately 11°C over 10 h, effectively identified deeper defects (up to 25–30 mm), such as wooden elements, moisture accumulations, and delamination in lime-based plasters. Conversely, active IRT employing convector and infrared lamp heating with 20–35 min of exposure and surface temperature increases between 5 and 12°C allowed detection of near-surface flaws during transient heating and cooling phases, albeit with limited observation windows of 2–5 min. Detection depth was influenced by plaster type, with lime-based coatings permitting defect visibility up to 25–30 mm, whereas cement-based plasters limited detection to 15–20 mm due to lower thermal conductivity. Reinforcement materials, including steel and basalt fiber meshes, were generally undetectable under both passive and active regimes, particularly when embedded beneath coatings thicker than 30 mm. Key parameters affecting defect visibility included plaster thickness, the magnitude of the thermal gradient, and heating uniformity, with thin fresco layers exhibiting the highest defect contrast. Based on these findings, a simplified and replicable IRT protocol is proposed, tailored to common heritage masonry typologies. The study emphasizes the importance of precise calibration of acquisition timing and thermal input to optimize diagnostic accuracy. While confirming IRT strong potential for qualitative assessment of plaster detachment and heterogeneity in historic masonry, caution is advised when interpreting results in the presence of modern retrofit materials.