Multiple Linear Regression Analysis for Predicting Unconfined Compressive Strength of Fiber Reinforced Fly Ash Stabilized Lateritic Soil

Abstract

AbstractThe importance of the prediction models for any of the engineering properties is due to time saving and material conservation which or else will take a lot of material and longer periods of time to determine the same through the laboratory experiments. The lateritic soil stabilized with fly ash (FA) and reinforced with glass fiber might be an effective alternative construction material in application to the geotechnical structures such as embankments, pavement sub-bases, retaining walls and structural landfills. This paper deals with the multiple linear regression (MLR) analysis to develop the prediction models for unconfined compressive strength (UCS) of lateritic soil stabilized with fly ash and reinforced with glass fibers. The lateritic soil replaced with 10, 20, 30 and 40% fly ash and added with glass fiber contents such as 0.1, 0.2 and 0.3% is varied by weight of lateritic soil. In total, 17 combinations were tested for UCS at curing period of 7 days in un-soaked condition. The results obtained in the experimental investigation are used for developing the regression model using MLR analysis technique. The models thus developed were found to predict the UCS in fair to better agreement with the experimental values obtained through laboratory trials. The values of R2 range from 0.963 to 0.973 indicating the good co-relation between the UCS as dependent variable and contents of fly ash and glass fiber as independent variables.KeywordsLateriticsoilFly ash (FA)Glass fibersUCSMultiple linear regression analysis