EXPERIMENTAL VALIDATION OF SOIL PERMEABILITY PREDICTIONS BASED ON GRADATION CHARACTERISTICS

Permeability is a fundamental parameter in geotechnical design, influencing seepage, drainage, and stability in earth structures, foundations, and embankments. An empirical formula using the area under the soil gradation curve to predict the hydraulic conductivity of coarse-grained soils offers a promising numerical model. This study experimentally tests and extends this model using soils from the Eastern Province of Saudi Arabia with varying proportions of gravel, coarse sand, and silt. Constant-head tests were conducted to evaluate hydraulic conductivity as a function of median particle size (D50) and gradation curve shape. Falling-head tests were applied to samples with substantial fine content. The results broadly corroborate the trend predicted by the model: soils with flatter gradation curves and more fines generally display lower permeability. However, in specific constant-head cases, measured permeabilities deviate from the empirical predictions; these discrepancies are attributed to factors not considered in the formulation, notably particle shape, roughness, and mineral composition. These findings suggest that while the gradation curve area is a strong predictor, accurate permeability estimation requires incorporating additional soil physical characteristics. This work bridges numerical prediction and laboratory evidence, contributing to more reliable permeability modeling for geotechnical engineering.