Computational Modelling of shrinkage in repaired concrete

Abstract

This study aims at development of a computational model and experimental study of moisture diffusion and shrinkage in cementitious repair materials to evaluate the suitability of a cementitious repair mortar for shrinkage performance. Time-dependent moisture loss in a cementitious repair materials was modeled using moisture diffusion theory. The problem of moisture diffusion is highly non-linear with the diffusion theory. The problem of moisture diffusion is highly non-linear with the diffucivity Kc depending on the moisture content C. Two finite element codes MSTDIFF1 and MSTDIFF2 have been developed which handle one- and two-dimentional moisture diffusion through cementitious materials, respectively. The relationship between the diffusivity Kc and moisture content C, required for the predictin of moisture loss, was established using experimental data of changes in moisture content C at different depths from the drying surface for different drying times. Free shrinkage strains esh in the repair material were determined experimentally and related to the moisture loss M. The relationship established between free shrinkage strains esh and moisture loss M can be used to determine the free shrinkage strains for different values of moisture loss in the repair material.

Free shrinkage strains esh were fed into a finite element based stress analysis program STRSRSYS developed to predict the shrinkage stresses in the repair overlay/substrate concrete system. The suitability of a cementitious repair material for shrinkage performance was evaluated by comparing the stresses developed in the repair material with the premissible maximum stresses. The results of this study show that use of a moisture dependent diffusivity Kc simulates the problem of moisture diffusion through the material more accurately. Free shrinkage strains were noted not to vary linearly with moisture loss, rendering the practice of solving directly the shrinkage diffusion equation questionable. The commercial repair material used yielded high free shrinkage strains resulting in high tensile and shear stresses. The ultimate free shrinkage strain of the repair material had the most significant effect on the level of induced stresses in the repair overlay/substrate concrete system. The restraint provided by the concrete substrate as well as the elastic modulus of the repair material had a significant effect on the level of induced stresses.