Cement vs aggregates and textures of aggregates in a mortar: Comparative image analysis methods and analytical protocols

The types, abundance and texture of phases are crucial for reconstructing the manufacturing of historical and contemporary construction materials such as mortars. Commonly, these data are obtained on polished mesoscopic and/or thin section surfaces and imaged with several techniques. Here, a thin section from an already well-characterised mortar was analysed to unveil the amount (area%) of cement paste vs aggregates, plus the textural features of the aggregates. The thin section was imaged by a high-resolution scanner (HRS), by transmission optical microscopy (TOM) and scanning electron microscopy (SEM). The single HRS image discriminates only quartz (qz) from cement+af+ss (af: alkali-feldspar, ss: sheetsilicates). The stitched TOM image distinguishes cement from aggregates, i.e. qz+af+ss, whereas the stitched SEM image discriminates cement, pores, qz and af+ss. The amount of cement vs aggregates and the area of the different aggregates determined by SEM is more accurate since it reflects chemical attributes. 2D Fuller curves were constructed considering different types of 2D dimensional parameters extracted from both TOM and SEM digital images. Intermediate dimensional parameters for both TOM and SEM had the best match with the 3D sieving Fuller curve. SEM shows the most detailed and adaptable recognition of cement to aggregate ratio and quantification of aggregate clasts, but it is time-consuming and expensive. HRS is rapid but only crudely accurate. TOM is more accurate and less expedite than HRS; by contrast, TOM is faster than SEM but unable to distinguish different aggregate clasts with similar optical features. • A mortar sample was analysed using three of the most used image acquisition technique. • Quantitative textural parameters of aggregates were extracted, compared and evaluated. • SEM imaging is more accurate on the discrimination of aggregates phase composition. • Either TOM or SEM imaging may retrace the aggregate grain size distribution curve. • On average intermediate textural 2D dimensions fit better the 3D sieve grain size.