Stéphane Sammartino

An image analysis contribution to the study of transport properties of low-permeability crystalline rocks

Abstract Transport properties of the Charroux–Civray tonalite matrix were studied by a two-dimensional multi-scale model, based on image analysis procedures. The estimation of a transport coefficient of the matrix was obtained by combining the 2-D image of the spatial arrangement of primary mineral phases with the aperture distribution of microscopic pathways. The study focused on one fresh facies and one fairly altered facies. Connectivity of the fluid flow network varied according to the studied facies. It is shown that pathways inside dark minerals and quartz, exert a lower influence on the transport coefficient than pathways inside feldspars. Strong assumptions used in the presented numerical method do not allow direct comparison between real transport coefficients and factors determined by the model. However, for both facies studied, comparison between simulated transport coefficients and experimental permeability suggests that the calculation hypothesis needs to be improved: the real aperture of microcracks must be taken into account, and a large number of classes of aperture are required for this kind of simulation.

Distribution spatiale de la porosité des matériaux cimentaires : Une approche méthodologique pour leur caractérisation

ABSTRACT Degradation of cementitious matrix in aggressive aqueous environments among others is accompanied by modifications of the spatial distribution of connected porosity. The application of the 14C or 3H-PMMA (polyméthylmethacrylate) impregnation method gives a quantitative and multi-scale representation of this distribution. The first application concerns a mortar, allowing to visualise and to quantify the in situ cement porosity. The second application is related to a cement paste degraded by organic acids. This tool has many potential applications in the field of cementitious materials. It allows a sharp characterisation of internal porosity variations frequently found during durability studies, and furthermore, it provides a realistic geometrical representation of pore space that can be used for solute transfer modelling.

Magnetic Resonance Imaging and Relaxometry as Tools to Investigate Water Distribution in Soils

Relaxation times and two imaging sequences (spin echo and single point imaging) were performed onto repacked soil samples to study respectively water distribution within the porosity and to measure water content profiles, distinguishing water contained in large pores from water contained in the whole porosity. These methods were applied to 25 samples of the same soil that was prepared to obtain aggregates of three different size, then repacked to five bulk densities. Samples were then equilibrated with water at five matric potentials. We found that T1 and T2 measurements present similar time distributions with essentially four peaks. We attributed the two shortest times to textural pores, and the two longest times to structural pores. The water profile measured with spin echo sequence was attributed to water contained in structural pores.