Predicting the Compaction Characteristics for Malaysian Sedimentary Residual Soil Mixed Bentonite as Compacted Soil Liner by Regression Model

Abstract

This paper presents the regression model for predicting the maximum dry density (MDD) and optimum moisture content (OMC) for sedimentary residual soil mixed with bentonite. The empirical models were developed in accordance with the statistical validity steps and conditions. The laboratories were conducted to determine the physical properties of mixed samples with bentonite such as Atterberg limit and particle size distribution. Meanwhile the compaction testing was performed on the mixed soil samples with bentonite to determine the MDD and OMC values at different compaction energies. The data obtained from laboratories and secondary data were used to develop an empirical model. The results showed there is significant relationship between the particle size, compaction energy and bentonite content to the OMC. Meanwhile there is a relationship between the OMC, bentonite content and compaction energy to the MDD. Reasonably good regression coefficients are obtained in case of both model which R2 = 82% for MDD model and R2 = 72% for OMC model. Meanwhile the determination coefficient and mean square error (MSE) for validated model between predicted model and empirical model were given R2 = 89.5% with MSE = 0.065% for MDD model and R2 = 79.5% with MSE = 5.9% for OMC model. Therefore, the models developed present a good predicting for MDD and OMC values.