Five year drying of high performance concretes: Effect of temperature and cement-type on shrinkage

This experimental study imposes limited relative humidity (RH). gradients to small mature concrete samples, at a constant temperature T = 20, 50 or 80 degrees C. Mass loss and shrinkage are recorded until stabilization at each RH and T, for up to 1991 days. Firstly, our mass loss data are consistent with those presented in former research (on different samples of the same batch). After presenting and analyzing shrinkage kinetics, experimental data are fitted with usual models for shrinkage prediction, at each temperature of 20, 50 and 80 degrees C. An adequate match is obtained by combining capillarity (i.e. Vlahinic's model coupling poro-elastic constants and water saturation level) and desorption (Bangham's equation). Subsequently, relative mass variation (RMV) is plotted against shrinkage epsilon(dry)(sh) data. Three distinct phases are obtained at 20 or 50 degrees C and down to 30%RH; up to four distinct phases are observed at T = 80 degrees C and down to 12%RH. The latter are confirmed by experiments on (60 degrees C; 7%RH) dried concrete. The four phases in the (RMV, epsilon(dry)(sh)) diagram are interpreted against shrinkage data on mature cement paste dried at 60 degrees C; 7%RH and against the literature.