Moisture is one of the major causes of problems in buildings, and it can compromise their performance. Infrared thermography (IRT) is a non-destructive testing technology that can be used to assess the humidification phenomenon and, thus, prevent some of the problems caused by moisture. The images obtained by IRT reflect the thermal patterns of the surface under study and can be evaluated using a quantitative approach, which allows not only the traditional visualization of the thermal patterns but also quantification of surface temperatures and/or their differences. The relevance of this work is related to the discussion of the strengths and weaknesses of several methods to quantitatively assess the humidification phenomenon using IRT. For that purpose, the partial humidification by the bottom surface of a lightweight concrete specimen was considered as a case study. To evaluate the thermal gradients, the evolution of the thermal imaging throughout the measurement period and the definition of the areas particularly affected by moisture, a methodology that included a pre-processing phase for data reduction, followed by a data processing phase, were implemented. In the data processing, different statistical and numerical methods were tested. The results of the statistical descriptive analysis highlighted the time variation of the surface temperature, both when considering the entire specimen and when considering only specific areas. The variability of the temperatures at certain moments of the experiment could be observed in the box-plot representation. The image subtraction proved to be an interesting technique to quantify the temperature differences if the first image was used as reference. A thermal index, TI, was proposed to assess the cooling rate. The index highlighted the initial instant when the effect of moisture on the surface temperature was detectable.
Keywords: descriptive statistics; infrared thermography; moisture; quantitative analysis; thermal images subtraction; thermal index.