TRANSPORT OF WATER DUE TO A TEMPERATURE-GRADIENT IN UNSATURATED FROZEN CLAY

As reported experimental data indicate, the net flux of water f in a fine-grained soil column was assumed to be given as: f=-D1(w,T) αw αx-ρD2 (w,T) αT αx where ρ{variant} is the dry density, T is the temperature (°C), w is the content of water in all phases, x is the coordinate, and empirical functions D1 and D2 are the properties of a given soil. Under this assumption a new experimental method was introduced to determine D2 of a soil with known D1. The D2 of Morin clay was determined as a function of w at several temperatures ranging between - 1.0 and 1.0°C. A common feature found is that D2 increases with increasing w, attains its maximum near or not far from a point where w is equal to the equilibrium unfrozen water content, and then decrease when T is negative. However, D2 increases with increasing w up to about 14%, and then remains more or less constant as w increases when T is positive. Because of this behavior of D2 a sudden change (or discontinuity) of D2 occurs near a point where T = 0°C when w is greater than 14%. The validity of the assumed functional description of the flux f is discussed based on some recent results of mathematical analysis on degenerate quasi-linear equations of parabolic type. © 1990.