ABSTRACT
In general, the lower effectiveness of air voids in blast-furnace cement concretes is explained by the fact that because of the very dense microstructure and its very low permeability the water can hardly reach the expansion spaces, even if the spacing factors are small [4],[9]. In contrast to this hypothesis experience has shown that the effect of air-entraining agents in blast-furnace cement concrete may be improved by longer curing and the resulting denser microstructure [4]. In an attempt to explain the different effect of air-voids on portland and blast-furnace cement concretes Brodersen [10] refers to the fact that the gradients of moisture and chlorides in blast-furnace cement concretes are steeper because of the very dense microstructure in the core. Therefore greater stresses may occur near the surface of these concretes. But Brodersen himself points out that this cannot be a satisfactory explanation of the low effectiveness of air-entraining agents in blast-furnace cement concretes, since according to this statement also blast-furnace cement concretes without airentrainment should behave in a more unfavourable way than the respective portland cement concretes. According to his own investigations, however, this is not the case. This means that at the moment a satisfying explanation of the different behaviour of air-entrained portland and blast-furnace cement concretes under freeze-deicing salt attack cannot be given.
