2022 Volume 76 Issue 1 Pages 211-219
A model was proposed to evaluate protected paste volumes around air voids. The area where each air void should provide protection was modeled by the Dirichlet tessellation associated with the planar point pattern of air voids. Changes in the Dirichlet tile area distribution due to the random distribution of air voids were represented by Monte Carlo simulations. The distances to achieve a specific volume for enough protection were defined by the area fraction corresponding to an upper confidence limit of 95％ in the simulation. The distances estimated in the proposed model were compared with the proximity of cement paste to the surface of air voids in the conventional models including the spacing factor of Powers. The results showed that the average area protected by the voids was consistent with the volume fraction estimated by the Attiogbe model. The characteristic distance to characterize point patterns in the proposed model was also equivalent to the mean tile area. The Philleo factor was found to overestimate the ability of protection at a given air content so that the evaluated distance was much smaller than that in the proposed model. The distances to cover the most of paste region estimated by the proposed model were almost equivalent to the conventional spacing factor. The conventional spacing factor was found to correspond to the maximum distance covering the local area where air voids were dispersed more sparsely compared to the average density of air voids in random spatial distribution.
