Re‐evaluation of the centrifuge method for describing the unsaturated hydraulic functions of porous rock and till soil samples

Abstract The determination of hydraulic characteristics as a function of the pressure head is laborious and data ranges are restricted. Indirect methods to estimate K ( h ) are introducing additional uncertainty and the method‐dependent restricted data ranges that can hamper a bimodal curve fitting. The objective of this study was to re‐evaluate the centrifuge approach for directly measuring the unsaturated hydraulic properties of porous carbonate rock and glacial till soil samples. Hydraulic conductivity was determined by means of evaporation (EVA), quasi‐steady centrifuge (QSC), and double‐membrane steady‐through flow (DMSTF) methods. Retention curves were obtained by using the EVA, QSC, mercury intrusion porosimetry (MIP), suction table, and pressure chamber. The tested samples belong to two rock lithotypes coming from the Apulia region, southern Italy, and three soil clods of dense glacial till parent material collected in northeastern Germany. Unimodal and bimodal hydraulic functions were used to fit hydraulic conductivity and retention data. The QSC data of the highly dense soil clods qualitatively demonstrated the different effects of bulk density and rigidity on the soil hydraulic functions. The MIP data showed the existence of a bimodal pore size distribution. The combination of data obtained by EVA, QSC, and DMSTF methods improved the fitting of the bimodal hydraulic conductivity functions to data of both rock and soil samples. The results confirm the uniqueness of the QSC method for measuring the hydraulic conductivity in a wider range of matric potential value and its limitation for nonrigid samples even if they are of high density.