The random hierarchical capillary bundle (RHCB) model contains random input parameters for estimating the cementitious material's gas permeability. However, the impact of the probability density function (PDF) of these parameters on the PDF of the model's output data remains un-explored. Hence, in the present study, different PDFs (i.e. uniform and normal distributions) were given to the random number generator determining the pore segment's length, and their impact on the PDF of the output data was investigated. The covered experimental case was alkali-activated concretes (made with different dosages of fly ash and slag) at reference state and after sustained exposure to moderate temperature of up to 200°C. After conducting approx. 100 simulations in each case, it was found that when the random number generator possessed a normal distribution with the lowest tested standard deviation (i.e. 0.33), the average estimated permeability of the concretes (at both reference and damaged states) was the closest to the experimental value, with the lowest standard deviation (i.e. half a magnitude or more, lower than the average simulated value). This indicated that by controlling the PDF of the random number generator, the improved RHCB model could estimate the gas permeability evolution with an enhanced precision.
A.N.Omrani et al. (Mon,) studied this question.