| Shrinkages on drying from 100% to 80%, 60%, 40%, and 20% rh have been measured on well hydrated alite pastes, cement pastes and Vycor glass under ambient conditions. The effects of C(,3)A, gypsum, water/solid ratio, specimen geometry, microcracking, morphology and "heat" treatment to shrinkage also are studied. The presence of hydrates of calcium aluminates, sulfates, or sulfoaluminates does not affect shrinkage directly. Shrinkage is only affected by those parameters which alter the pore structure. Surface area, by nitrogen adsorption after the paste is solvent replaced, was found to correlate well with shrinkage. The specimen geometry only affects the drying rate and hence the rate of shrinkage, but not the magnitude of the final shrinkage. The measured total shrinkage is an unrestrained material property once the equilibrium condition is met whether or not microcracking has occurred during drying. During drying, shrinkage is dominated by capillary tension mechanism and the disjoining pressure which thins the pore water has no contribution to stress. During rewetting, swelling is dominated mainly by disjoining pressure and slightly by capillary pressure. The irreversible shrinkage is mainly caused by the stress difference between drying and rewetting processes. A mechanics model represents the paste structure and the idea of the dependence of the shrinkage on final state of stress field is proposed. Shrinkage, either total, reversible, or irreversible, can be represented by S(N(,2)).(VBAR)ln h(VBAR), the stability of pore structures and the deformability of the paste. |
