New Frostblast Theory
The hitherto common textbook opinion on the crushing of rock by the expansive force of water during freezing may need to be revised. According to Julian Murton of the University of Sussex in Brighton and his colleagues, the current mechanism does not work under natural conditions, at least at latitudes near the poles.

When water freezes to ice, its volume increases by nine percent - enough expansion energy to widen and open pores, for example. However, according to the geoscientists, the rock would have to be completely water-saturated and iced up from all sides at the same time, otherwise the moisture in the rock could escape or even escape completely. The same applies to fissures: either the liquid rushes through them or the ice grows out of the fissure. In addition, frost cracking, at least in laboratory tests in sand and similar rocks, only sets in well below zero degrees Celsius, which also speaks against pure freezing.

In their experiments, however, Murton's team found that frost cracking is based on the formation of so-called ice lenses, which are also typical of permafrost soils. The ice crystals that form when the temperature falls below zero attract water from the direct rock environment or from outside – favored by the capillary replenishment. Since the water molecules are connected to each other and to the rock and ice surface by adhesive forces, water continues to trickle in until the available supply is exhausted - the thickness of the ice lens continues to grow for just as long. They first build up the pressure which, together with the flow of water in the pores, ultimately shatters the rock.
The researchers feel confirmed by a comparison of their experimentally generated frost damage with that from permafrost regions, where the formation of ice lenses is a daily process. The depth and geometry of the fractures were nearly identical.