Obtaining compaction parameters of soils by statistical analysis

Soil compaction has an important place in ensuring stability and durability for road embankments and earth dams.In the compaction method used to increase the density of the soil, the best compaction occurs at the maximum dry unit volume weight (γkmax) value corresponding to the optimum water content (wopt). These two parameters, which are very important for soil compaction, can be determined by compaction tests performed in the laboratory. Since determining these two parameters through laboratory experiments requires time-consuming and laborious procedures, estimating compaction parameters based on regression analysis using the index properties of the soil has been considered an alternative method for many years. In this study, a large data set including sieve analysis, consistency limits and standard proctor test results applied on 584 different soil types was used to estimate compaction parameters indirectly based on regression analyses. Using the physical properties of the soils in the data set, empirical equations based on regression analyzes were obtained to determine the optimum water content (w opt ) and maximum dry unit volume weight (γ kmax ) parameters. In the analyses, it was understood that the most effective physical parameter in estimating both compaction parameters was the fine grain percentage (FG%), and the weakest effective parameter was the sand percentage (S%). Additionally, it has been observed that the most successful prediction models in determining both compaction parameters are those in which liquid limit (LL), plastic limit (PL) and fine grain percentage (FG%) variables are used together.