Main, WSEAS World Congress, Malta, September 2008

Plenary Lecture

Critical Review of the Physical Foundations of Unsaturated soil Mechanics

Professor Rafi Baker
Co-author: Professor S. Frydman
Department of Structural Engineering & Construction Management
Faculty of Civil & Environmental Engineering
Technion - Israel Institute of Technology
Haifa, 32000, Israel

Abstract: Most constitutive models for unsaturated soils are based on identification of matrix (or metric) potential (energy per unit volume) ψ with it's the capillary component ψ_cpl= (u_a - u_w) where {u_a, u_w} are the pore air pressure and (negative) pore water pressure receptivity. This identification ignores the contribution of adsorption potential ψ_ad to ψ. Identification of a potential with stress (or suction), is questioned, since these quantities have different physical significance despite their common dimensions. Moreover, (u_a - u_w) is a valid expression for capillary potential only in the simplified and non realistic model of a pore space consisting of a collection of cylindrical capillaries. In reality however the structure of clay soils consists of collection of pods built up of closely spaced parallel plates, and the pods themselves are arranged randomly in space. The water inside these pods is well within the range of adsorption forces, and the adsorption potential should not be ignored. It is noted that unlike the capillary potential, the adsorption potential has no direct interpretation in terms of pressures, and it can not be incorporated directly into mechanical constitutive equations.
Note also that (u_a - u_w) are not measurable quantities and the only measurable variable is the potential ψ which to a first approximation can be considered as a sum of capillary and adsorption potential, i.e. ψ = ψ_{ad +}(u_a - u_w). Therefore, neglecting the adsorption potential results in an over estimation of the capillarity component, for a given measured ψ value.
All techniques for measuring ψ are based on the principle that at equilibrium it is the water potentials rather than water pressures that are equal in the soil and the measuring device. It is impossible to measure water tension greater than approximately 0.8-1.0 atm due to cavitation of the water in the measuring device. In order to overcome this technical difficulty, most potential tests in geotechnical engineering utilize the axis translation technique which applies an external air pressure to the sample. This technique "translates" water pressure to the positive range thus preventing cavitation and making it possible to perform the measurement of u_w.
Both thermodynamic considerations and direct measurement of ψ using psychrometeric techniques indicate that at low water content ψ is of the order of 10.000 atm. Moreover under usual field conditions air pressure is atmospheric (i.e. u_a= 0). Combining the above considerations yields u_w= -10.000 atm. In a heterogeneous and cavitation nuclei rich medium like unsaturated clays, the tension stress in the water can not exceed 0.8-1.0 atm due to cavitation of the soil water. Consequently the extremely large tensile stresses in the water implied by the geotechnical approach, are not realistic, and result from neglecting the adsorption potential which does not have a mechanical interpretation, and from the use of the axis translation technique. Introducing such unrealistic water tensions into mechanical constitutive equations is not justified, resulting with various conceptual problems. It is noted that preventing cavitation by applying an elevated air pressure to an unsaturated soil sample, modifies its behavior.
Consequently it is doubtful whether constitutive formulations based on experimental information obtained by the axis translation technique are relevant to actual field behavior.
Capillary potential is shown to account for only a small part of matrix potential, the major contribution resulting from water adsorption inside the soil pods, particularly in soils having large specific surface areas. Consideration of double porosity models as well as the adsorption potential appears essential for proper interpretation of unsaturated soil behavior.
The present talk does not present a complete framework overcoming the above mentioned difficulties. However, reference to the strength of unsaturated clay soils supports the above criticism, resulting also with an alternative and simpler formulation than the conventional geotechnical approach. The main purpose of the present talk is to emphasize the distinction between the terms water potential ψ (energy per unit volume), and the stress variable (u_a - u_w). The confusion between these two terms in the common geotechnical framework is probably the main element preventing the construction of rational and consistent theory describing the mechanical behavior of unsaturated clay soils.

Brief Biography of the Speaker: Baker completed his first and second degrees in the faculty of Civil Engineering of the Technion I.I.T. The 2nd degree was done in the Geotechnical Engineering Department. He did his PhD in the faculty of Agriculture of McGill University, Montreal Canada, in the Dept. of Soil Physics. The subjects of the speaker's 2nd and 3rd degrees dealt with the interface between soil mechanics and soil physics. As a result, of this history he follows both the geotechnical and soil physics literature. The present talk is a result of this dual interest.
He received the G.J. Zeitlen price from the Israeli Association of Engineers and Architects, and twice delivered the Kassiff memorial lecture, which is the most prestigious geotechnical lecture in Israel (some of the previous distinguish lecturers include: G.A. Leonards from Purdue Univ., J. Bear from the Technion, S. Marchetti from L’Aquila Univ., D. Fredlund from Saskatchwan Uni., J.B. Burland from Imperial College, R. Goodman from Berkley, M. Randolh from the Univ. of Western Australia, F.H. Kulhawy from Cornell Univ., and I. Moore from Queen’s Univ. ).
During 1995-1998 he was a member of the International subcommittee on Slope Stability. During 2000-2002 he served as the head, of the Division. of Geotechnical Engineering at the Technion I.I.T.
During 1980-1981 he was a Visiting Professor at VPI & SU. Virginia, USA. During the summer of 1981 he was a Visiting Scholar, in the Geotechnical Research Center of McGill Univ. Montreal, Canada. During the summer of 1983 he was a Visiting Scholar in the Univ. of Arizona, Tucson, Arizona, USA. During 1985-1986 he was Visiting Professor in Carleton Univ. Ottawa, Canada. During the summer of 1997 he was a Visiting Professor, at Kobe Univ. Kobe. Japan. During 2004 he was a visiting professor in North-Carolina University at Chapel-Hill USA (interrupted due to sickness).