On hydrostatics and matristatics of swelling soils

Abstract

We review and re-examine the development of the hydrostatics of swelling soils, starting with the pioneering work of John R. Philip. His early attempts to formulate the overburden potential of the soil water as an additional component to the pressure and gravitational potentials did cause some confusion, until it became clear that there is no additional component and that the actual pressure potential can be split into two parts: the (non-actual) pressure potential the soil water would have if the system were unloaded, and a remaining part, which may be called the overburden potential. Using fundamental thermodynamics it is then shown how the overburden potential can be calculated from the slopes of the bundle of load-pressure-dependent shrinkage curves. A set of experimental data published by Talsma is used to test the Groenevelt-Bolt equation of state for the bundle of shrinkage curves and for the numerical values of the overburden potentials for Talsma's clay paste. Subsequently, the Groenevelt-Bolt equation is used to interpret the physical significance of the extremes of residual shrinkage curves, for which shrinkage data from the literature are used. The description of the static behaviour of the matrix of swelling soils is termed "matristatics".