Estimation of Unconfined Compressive Strength (UCS) for Soft Soil Stabilized with Cement Using Multiple Linear Regression Analysis

Abstract

This paper presents the development of precise linear regression model for predicting an unconfined compressive strength (UCS) for soft soil stabilized using cement in terms of main effective factors such as cement content and times of curing periods. The empirical UCS model was developed using the secondary dataset from the previous studies collected from the published data such as jurnals, technical papers, proceeding and theses. Meanwhile the multiple linear regression analysis was used to develop a UCS model. The results shows there is a strong significant relationship between cement content and times of curing periods to predict the UCS value and the results shows a reasonably good regression coefficient has been obtained for the UCS model which is R² = 0.93. The UCS model developed demonstrated significant accuracy with the fitted regression model between the predicted UCS and measured UCS with the determination coefficient value is R² = 0.97 and low value for mean square error (MSE) is 0.068. It shows the UCS model developed was present a good to estimate the unconfined compressive strength for soft soil stabilized using cement and it will be practical for a preliminary design of a project where there costs and time limitations.