Abstract
The old theory that frost heaving is due to change in volume of water frozen was based on experiments with closed systems. Field observations and recent experiments indicate that soils, when subjected to freezing under normal conditions, usually behave as open systems. When the freezing of saturated soils results in little or no heaving, part of the water is forced through the soil voids below the zone of freezing, compressing or expelling air. Excessive heaving results when water is pulled through the soil to build up layers of segregated ice. These ice layers grow in thickness because water molecules are pulled into the thin film that separates the growing ice crystals from underlying soil particles. Since heavy surface loads may be heaved and much force is required to pull water through impervious clay, the water is put under high tension. Heaving is limited by the tensile stress that may be developed in the water and by downward growth of ice crystals in soil voids. These two factors also probably explain the rhythmic banding due to alternating layers of ice and clay. In well-consolidated clays the surface uplift equals the total thickness of the ice layers, the water content of the clay between the ice layers remaining approximately constant; but heaving is continuous and regular instead of intermittent. Clay is soft near the lowest ice layer because much of the water is unfrozen, the hardness increasing higher up where the temperature is lower and freezing has gone on for a longer time. Additional evidence has been obtained by freezing in open systems other liquids than water.