Estimating Water Retention Characteristics of Sands from Grain Size Distribution using Idealized Packing Conditions

Abstract A new approach is presented for estimating the soil-water characteristic curve (SWCC) of granular porous media along a drainage path using the relatively simple measurements of grain size distribution (GSD) and mass-volume relationships (void ratio). GSD measured using mechanical sieve analysis is converted into an equivalent population of smooth, spherical particles. Pore size distributions representing relatively loose and relatively dense packing conditions are calculated from the geometry of idealized two-dimensional unit pores formed among particle groups randomly assembled from the simulated particle population. The Young–Laplace equation is used to quantify the amount of pore water retained in the form of thin films, liquid bridges, and completely filled pockets, thus allowing the SWCC to be modeled over the entire saturation range. The measured void ratio is used to constrain the modeled SWCC via the theoretical consideration of the work done to expand air-water interfaces throughout the matrix. Nine sets of results for sands and glass beads are used to evaluate the model’s performance. SWCCs are most effectively predicted in the capillary and funicular saturation regimes (20 % < S <100 %), with more deviation observed in the pendular regime (S < 20 %). Air-entry pressure is predicted within an average error of 8 %. Assumptions and constraints required in the framework restrict the general applicability of the approach to poorly graded materials with predominant grain sizes in the sand- to silt-sized range.