Frost action in clay soils. II. Ice and water location and suction of unfrozen water in clays below 0°C

Abstract

The relation between temperature, unfrozen water content (UWC) and shrinkage of a number of clays, and the distribution of ice and water within the clays, was studied. The clays included sodium and calcium Wyoming bentonites, < 1 pm and 1-2 pm kaolinite and three English subsoil clays. On freezing blocks of these clays (initial size up to approximately 1 cm') over a period of months, they shrank progressively with reduction in temperature studied except for the < I pm and 1-2 pm kaolinites, which ceased shrinking at -0.24"C and -0.73"C respectively. During shrinkage, the water in the clay blocks migrated to freeze outside the blocks where ice was already nucleated. After equilibration for a month at -0.19"C and at -0.73"C, the water contents of the finer clays, excluding ice bands, equalled the UWCs measured previously using a differential scanning calorimeter. Two stages of ice formation were apparent in the curves of UWC vs temperature and were attributed to the different locations of the ice in the clays. Initially, ice formed in bands outside the clay matrix, accompanied by clay shrinkage (Stage 1). After a shrinkage limit had been reached ice began to form within the pores of the clay (Stage 2). Comparisons of the results of these freezing experiments with conventional moisture-release curves showed that the suction of the unfrozen water in clays was, within experimental error, as predicted by the Schofield equation. Results are also discussed in terms of ice-water and air-water interfacial tensions.

Morphological instability of a non-equilibrium ice–colloid interface

“…As the temperature is lowered, the ice phase grows and the particle volume fraction of the clay increases. At a critical breakthrough temperature T B and volume fraction φ B , ice is able to enter the pore space between particles (Brown & Payne 1990;Shanti et al 2006). For kaolinite clay, the ice-entry temperature is T B = −0.77…”

“…• C (Brown & Payne 1990). As far as we are aware, the ice-entry temperature has not been measured for bentonite, though Brown & Payne (1990) found no pore ice down to at least −8…”

“…For example, in soils composed mainly of silt particles, the ice usually forms nearly planar layers, called ice lenses (Taber 1929;Watanabe & Mizoguchi 2000), whereas in a wide range of clays and colloidal suspensions, the ice can grow in several directions, producing a remarkable variety of patterns (Taber 1929;Chamberlain & Gow 1979;Brown 1984;Deville et al 2007; figure 1). Recently, we proposed a mechanism for ice segregation in freezing colloidal suspensions based on the concept of morphological instability studied in alloys quantitatively applied to the ice-suspension interface (Peppin et al 2006(Peppin et al , 2007(Peppin et al , 2008.…”

Frost Heave in Colloidal Soils

“…At a given temperature the concentration φ of the soil in equilibrium with pure ice is dictated by the T f (φ) curve. Eventually the soil becomes fully consolidated (φ → φ p ) and a critical ice breakthrough temperature T E is reached [7,42]. At this temperature ice enters the pore space between particles, and interstitial freezing proceeds.…”

“…A combination of factors allow the partially frozen soil to segregate, revealing a new lens of pure ice [26,39]. Remarkably, frost heave experiments on highly compressible laboratory soils such as colloidal silica found that there is no pore ice near the ice lenses [6,7,45,46]. These results confirm early observations of Beskow that in clays and fine silts the soil between ice lenses is soft and unfrozen [5,28], but have yet to be explained theoretically.…”