Résumé On a prouvé expérimentalement et théoriquement l'existence de degrés critiques de saturation en relation avec le gel des matériaux de construction, poreux et cassants. Le degré critique de saturation est une constante du matériau qui a une valeur bien définie pour un type et une qualité donnés de matériau. On a démontré qu'il se produisait des détériorations après un gel si le degré de saturation actuel excède la valeur du degré de saturation critique. On peut choisir un matériau en fonction de sa résistance au gel en comparant, pour différents matériaux, le degré critique de saturation et les degrés de saturation effectivement atteints dans une situation donnée. L'utilisation du «degré critique de saturation» permet de faire disparaître le manque de corrélation existant dans certains cas entre les résultats obtenus par les expériences de gel et les résultats obtenus lors de l'utilisation du matériau.

Summary It has been proved theoretically and empirically that critical degrees of saturation exist in connection with freezing of porous and brittle materials of different types. It has also been proved that damage occurs after one freezing if the actual degree of saturation exceeds the critical, which means that frost deterioration is not a phenomenon of fatigue but a typical fracture phenomenon. This in turn means that the problem of frost resistance in practice is of a statistical nature (fig. 9). p The “critical degree of staturation”, SCR is a material's constant that adopts individual values for each particular type and quality of material. In different environments and points of time, different “actual degrees of saturation”, SACT, are reached. Comparisons between the critical degrees of saturation of different materials and those reached in practice in a given situation provide a possibility of choosing materials with regard to frost resistance, viz. the frost resistance of different materials used in the same situation is expressed: Degree of frost resistance=SCR−SACT The analogy with a normal design in regard to strength is clear (figs. 1, 2, and eq. (1)). By using the concept “degree of saturation”, two fundamental deficiencies of recent methods of estimating frost resistance can be removed to a great extent: The defective correlation between the results of laboratory experiments and the behaviour of the tested material in practice; the difficulties of the choice between different materials with regard to frost resistance. The article also shows how the concept “degree of saturation” can be used in dealing with the frost resistance of a layered construction (figs. 3, 4). A laboratory test for determination of frost resistance can be divided in two parts: Part 1 Determination of S CR . This can be done empirically by usingexactly the same freezing method for all materials. It can also be done theoretically, viz. the existence of a critical degree of saturation is only a geometrical consequence of the existence of a critical distance between air-filled recipients. Possible empirical and theoretical methods are accounted of in the article. Part 2 Determination of S ACT . This can only be done, from a theoretical point of view, by use of advanced moisture mechanics or by tests reproducing exactly the actual environment. However, as a substitute, even simple water-uptake tests can be used. The different ways of using a material can be taken into account by changing the procedure at the determination of S ACT . c