Two Improvements to Gardner's Method of Measuring the Hydraulic Conductivity of Non-saturated Media: Accounting for Impedance Effects and Non-constant Imposed Suction Increment

Gardner's (1956, https://doi.org/10.2136/sssaj1956.03615995002000030006x) transient method of measuring the hydraulic conductivity function of unsaturated media has been largely used, together with the improved graphical method proposed by Kunze and Kirkham (1962, https://doi.org/10.2136/sssaj1962.03615995002600050006x) to account for the impedance effect resulting from using a low hydraulic conductivity ceramic disk in porous plate testing. These methods are nowadays seldom used, since they have been replaced by numerical back analysis and methods for parameter optimization. Based on tests carried out on a specific device allowing to determine the water retention and transport properties of water in a coarse granular media at low suctions (up to 50 kPa), it was found necessary to account (i) for impedance effects and (ii) for the effects of nonconstant suction increments, as is often the case when using the hanging column technique. A new method is proposed to account for impedance effects, based on an analytical solution of the equations governing water transfer. The validity of this method is tested by considering experimental data from three distinct materials: a coarse green roof volcanic substrate, poorly graded sand, and undisturbed silty clay. Compared to the graphical method Kunze and Kirkham's method, it is less operator-dependent and hence more objective. It is also simpler than numerical back analysis methods, since it does not require any use of numerical code or parameter optimization algorithm, providing a more direct and reliable access to the hydraulic conductivity. An analytical solution is also proposed to solve the problem resulting from the application of a nonconstant suction increment.