In most of mountainous area, particularly in France, the dominant weathering process lies in cracks propagation due to freezingthawing. To predict this risk in rock masses, it is necessary to know the stresses developed in freeze-thaw cycles in the natural fractures. A reproduction of this phenomenon has been carried out on artificially notched specimens of different limestones. During these experiments generated stresses were assessed along the crack inner walls. The deduced stress profiles show a timedependant maximum value, which moves in-depth in the notch. The limit reached by this stress in this process depends on intrinsic properties of the material. A simple model for this evolution has been established. The generated stress by freezingthawing in a crack comes from a coupling between thermal propagation, water phase shift and water flow in the notch and in the porous rock matrix. An analytical assessment of the stress intensity factor at the crack bottom was made by using the model established for the stress in the crack. It confirms the possibility to initiate a rupture according to the principles of linear failure mechanics. This study gives encouraging results, which could be used to establish a predictive model for the failure of a rock mass subject to freezing-thawing cycles.
The difficulty with the forecast of rocks falls in mountainous area requires risks management issues and also infrastructure maintenance . To define the temporal characterization of this risk of breaking in the rock massifs, we need to get a better knowledge of the alteration phenomena. Considering limestone cliffs, in mountain context, studies tend to indicate that frost is the main alteration phenomenon, see Figure 1. Freezing gets into the existing fractures network and propagates itself, see Figure 2: that is called macrogelivation. To estimate the spread of the rock cracking due to the frost and to set up a mechanical model representing this weathering mechanism, a model of the stresses exerted on this network is needed. However the volume increasing resulting of the change from liquid to solid phase is not enough to explain the stresses that appear during the freeze-thaw cycles: Djaballah Masmoudi  revealed water migrations at both solid and liquid phases between the rock and the inside of the crack. This makes the stresses generation by volume expansion simple model more complicated. A model of the stress exerted on a fissure by 568 freezing has been established from a simple laboratory simulation. Then, the limestone type and the notch geometry effects on the stress have been studied. Once the stresses model was entirely parameterized, it has been used analytically, to be validated, in linear failure mechanics laws.
The process reproduces the fissure progression due to freezing in a massif, started from the extremity of one notch in a limestone block.