Planar Image-Based Reconstruction of Pervious Concrete Pore Structure and Permeability Prediction

Abstract

The permeability of pervious concrete is discussed in this paper. Transport properties of porous materials such as pervious concretes are inherently dependent on a variety of pore structure features. Empirical equations are typically used to relate the pore structure of a porous material to its permeability. In this study, a computational procedure is employed to predict the permeability of 12 different pervious concrete mixtures from three-dimensional (3D) material structures reconstructed from starting planar images of the original material. Two-point correlation (TPC) functions of the two-dimensional (2D) images from real pervious concrete specimens are employed along with the measured volumetric porosities in the reconstruction process. The pore structure features of the parent material and the reconstructed images are found to be similar. The permeabilities predicted using Darcy’s law applied to the reconstructed structures and the experimentally measured permeabilities of pervious concretes are found to be in reasonably good agreement. The 3D reconstruction process provides a relatively inexpensive method (instead of methods such as X-ray tomography) to explore the nature of the pore space in pervious concretes and predict permeability, thus facilitating its use in understanding the changes in pore structure as a result of changes in mixture proportions.