Using grain size to predict engineering properties of natural sands in Pakistan

Laboratory determination of strength and deformation behavior of clean sands and gravels has always been challenging due to the difficulty in obtaining their undisturbed samples. An alternative solution to this problem is to develop correlations between mechanical properties of cohesionless soils and their gradation characteristics. This study presents database of 3 natural sands with 11 varying particle size gradation curves to allow investigating relationships between mean particle size, maximum and minimum void ratio, relative density and shear strength of the test soils. Direct shear tests were performed at relative densities of 50, 75 and 95% to explore the effects of gradation and density on the angle of internal friction of the modeled sand samples. It is found that the mean grain size D₅₀ bears good correlations with void ratio range (e_max-e_min) and peak angle of internal friction ϕ'_peak. The generated regression models are in good agreement with published literature and can be considered as reliable for natural sands in Pakistan. These empirical correlations can save considerable time and efforts involved in laboratory and field testing.