Numerical Determination of Hydraulic Conductivity of Aggregate Drainage Layer

Abstract

In practice, the actual hydraulic conductivity (k) of an aggregate drainage layer in a pavement structure is often not known, and is estimated empirically. This leads to estimation errors of unknown magnitude. The main difficulty encountered in practice is that a standard highway laboratory does not have the required facilities for k determination, and road contractors are typically not equipped to conduct in situ k measurements. Other problems include the large size of test cylinders required for either falling- or constant-head tests, the difficulty in laboratory testing of sample preparation to achieve the desired porosity, and, in the field, the interference of underlying materials of lower k. One solution is to derive k theoretically and eliminate the need for experimental measurements. This research proposes a numerical simulation method to theoretically determine k of a given aggregate blend. The study first employs the Laguerre tessellation method to randomly generate a virtual microstructure of an aggregate blend with known gradation and porosity. Next, the lattice Boltzmann method is utilized to calculate k by simulating the process of a falling-head test. The application of the proposed numerical method is demonstrated by comparing the predicted k with experiment-measured values of standard AASHTO aggregate drainage materials. The proposed method is applicable for calculation of k for any aggregate drainage material in the laboratory or field with known gradation and porosity (or density), both of which can be more readily determined by experiment than k.