Influence of salt on desorption isotherm and hygral state of cement mortar – Modelling using neural networks

The durability of most porous building materials is strongly related with moisture and salt transport mechanisms. The main storage parameter which describes the hygral state of any porous material is the sorption isotherm. The degree of saturation of pores with moisture phase as a function of relative humidity and salt concentration was determined by means of saturated salt solution method for cement mortar. Then, using feed-forward neural network the relation was approximated for the whole range of arguments. Moreover, both partial derivatives were calculated using specialized neural network approach. It turned out, that a surprisingly simple, layered network (2–4–1) fits well the experimental data. The appropriate subroutines were implemented in HMTRA_SALT code. Using it, the examples of drying of the wall with and without salt were calculated and the results were analysed.