ABSTRACT
Baker and Frydman (2009) presented an extensive discussion of soil water suction. Although this will not be repeated here, it is worth again emphasizing the major point that suction is a state of energy or potential, and not a stress, although one of its components, the matrix potential (ψm), includes a capillary component, ψcap, equal to (ua − uw) which is a stress. However the potential also includes an adsorption component, ψad, which results from physiochemical interaction between the water and the clay minerals, and cannot be considered a stress. In fact considering that the tensile strength of pure water is of the order of 150 MPa, (Zheng et al. 1991). While matrix suctions may reach up to 1000 MPa, it is obviously not possible to equate the two. Consequently, matrix suction cannot be considered, or treated, as a stress in constitutive modeling of unsaturated soils. Monroy et al. (2011) expressed this principle most succinctly in their discussion of matrix suction measurement using the miniature tensiometer: “When making a measurement with the miniature tensiometer on an unsaturated soil, the internal energy of the water filling the tensiometer’s reservoir reaches a state of equilibrium with the soil’s pore fluid. However, this does not mean that the tensile stresses in the tensiometer and soil are similar, since both capillary and adsorbed components of potential are present in the latter case.” Since matrix potential/ suction contains these two components, of different physical character, it would be expected that each may have a different effect on the mechanical behavior of the soil. At the same time, although we are able to measure total matrix suction, it does not appear possible, presently, to separately measure the individual components. This makes our attempt to study, quantitatively, the influence of matrix suction on mechanical behavior (e.g. to develop constitutive models in terms of matrix suction) a formidable task. Gens (2010), in his Rankine lecture, recognized this difficulty, stating: “The fact that the two components of matrix suction are lumped together in a single variable, and that in addition, they are difficult to separate experimentally, should not lead to the conclusion that their effects on mechanical behavior are equivalent. Actually such effects should depend on the relative importance of each of the two components, and thereby on the types of soil being examined”. A way forward may be offered by considering the double porosity nature of unsaturated clays (inter-aggregate larger pores, hereafter termed macropores, and intra-aggregate smaller pores, hereafter termed micropores). Romero (1999) and
Romero & Vaunat (2000) observed that at low moisture contents, clay samples compacted at the same moisture content but to different dry densities (and so, also, to different degrees of saturation) were found to have the same matrix suction-see, for example Figure 1.
