Prediction of maximum dry density of local granular fills

Abstract

The paper presents a relation. developed to predict maximum dry density (MDD) in terms of the solid density and the gradation coefficients that characterize the grain size distribution of locally employed granular fill materials. For this purpose, two geologically different soils commonly used as selected fill materials are blended. A large number of compaction tests as required by statistical principles are conducted using the modified Proctor compaction procedure. The gradation envelopes, within which the ((ontrolparameterS are varied, are selected on the basis of local specifications for sub-base materials. A strong correlation is found between the MDD and the specific gravity of the solid grains. In contrast, the correlation between the MDD and the gradation coefficients including the coefficient of gradation, Cg and the uniformity coefficient, Cu, is weak. The developed predicting formula is compared with existing methods. It is observed that the developed relations predict well the MDD of soils of the nature studied in this work than do the existing methods. A test on a soil of relatively high solid density revealed that the developed relation looses its power of prediction when applied on soils of solid density beyond the range studied.

Keywords: Compaction, gradation coefficients, granular soil, maximum dry density, mooified Proctor test, solid unit weight, optimum moisture content