A GIS-based image processing approach to investigate the hydraulic behavior of mortars induced by volcanic aggregates

Since Roman Age, the reactiveness of the mortars was empirically sought using different aggregates added to lime to confer hydraulic properties. With the advent of microanalytical studies, great strides have been made toward a better understanding of how the aggregates can catalyze the formation of hydraulic phases. However, due to the small scale and the different distribution speeds of the reactions, the techniques routinely used are not fully performing for answering this task. In this work, an innovative semi-automated image processing procedure based on the multivariate statistical analysis of X-ray maps is adopted to highlight and quantify, for the first time, the hydraulic behavior of mortars in terms of reactivity between binder and aggregate. The method proposed has been applied to two different volcanic-based mortars (known as “ghiara- and azolo-mortars”) collected from historic buildings in Catania (Sicily, Italy). Obtained results allowed identifying, in both mortars, four distinct compositional sub-zones within the binder as well as mapping the distribution of the Hydraulicity Index (HI). This latter resulted to be quite higher for the ghiara mortar (HI from properly hydraulic lime to cement for ∼50% of the binder) with respect to the azolo one (HI from properly hydraulic lime to cement for ∼10% of the binder).