Modified random hierarchical bundle model for predicting gas permeability of cement-based materials.

• A modified random hierarchical bundle model for predicting the permeability of cement-based materials was presented. • The intrinsic permeability of cement-based materials can be accurately simulated. • The apparent permeability and Klinkenberg constant of different gases can be predicted. Permeability is a fundamental property indicating the durability of cement-based materials, predicted according to the material's pore size distribution (PSD). The random hierarchical bundle model is a promising approach based on a random hierarchical assembly of capillary segments to estimate the permeability of porous media. However, the original model neglected the slippage effect during gas penetration—a common occurrence in cement-based materials and other low-permeability porous media. Thus, we considered this effect in a modified model proposed herein, and the permeability of cement-based materials can be directly predicted using the PSD obtained via mercury intrusion porosimetry. Compared with data from experiments and literatures for various mortars and concretes with permeabilities ranging over 3 orders of magnitude, the results revealed that this modified model can accurately predict the intrinsic permeability and Klinkenberg constant. Moreover, this model can predict the intrinsic permeability and apparent permeability of various gases and porous media. [ABSTRACT FROM AUTHOR]