ABSTRACT
Mathematical and numerical modelling of degradation of the microstructure presupposes a reliable description of the virgin microstructure. A possible way to characterize the microstructure is by performing porosity measurements. The interpretation of these measurement, however, is not free from difficulties [4]. Another line that can be followed is that of numerical simulation of the microstructure. This is not free from difficulties either, but has the advantage of flexibility is that these models can be refmed and improved by inserting huge amounts of information available in different disciplines (physics, stereology, (colloid)chemistry). One of the first aims of these numerical simulation models is to check their consistency with available experimental data and to which extent they are able to allow for the effects of changes in curing parameters like the curing temperature. Once this consistency has been established, these models can be used for further parameter and sensitivity studies. Against the background of this philosophy the potentialities will be discussed of the simulation model HYMOSTRUC to simulate the effect of curing temperature on porosity. The calculated porosity will than be used as a basis for the prediction of autogenous shrinkage of cement paste. The calculated micro-level properties of the paste will than be inserted in a meso-level model for simulatiing shrinkage-induced deformations and stresses in a real concrete.
