Developing Correlation between Index Properties and Swelling Pressure of Expansive Soils Case of Bishoftu Town

Abstract

When an expansive soil is subjected to moisture increment it will exert an uplift pressure on the structure resting on it. Expansive soil covers an appreciable part of Ethiopia. The study area, Bishoftu is one of the places where expansive soils are found in abundance and present a significant structural and geotechnical challenge, especially to light weight structures. Crucial factors that require identification when dealing with expansive soils are the swelling characteristics. Swelling pressure predictions via laboratory testing of samples is the most acceptable approach to estimate swelling pressure, but it is expensive and time taking, therefore empirical equations developed from correlations with index properties are an alternative means to predict swelling pressure. This study aims to correlate index properties and swelling pressure of expansive soils found in Bishoftu town. A total of 24 disturbed and 22 undisturbed representative samples were collected from 12 test pits, at a depth of 1.5m and 3.0m and the necessary laboratory tests were conducted to find out the index property and swelling pressure for all samples. The soil in the study area is classified as CH as per USCS and A-7-5 as per AASHTO soil classifications. The results were analyzed using MS-EXCEL and SPSS computer programs, taking swelling pressure as the dependent variable and results of index property test as predictors. The results of the single linear regression analysis showed that the swelling pressure, Ps of the study area could be best estimated from its dry density with R2=0.9172 and Ps= 785.24ρd-744.82. Besides dry density; the correlations with water content, Free Swell Index and Linear Shrinkage yielded an acceptable estimations of Ps with R2 values of 0.8976, 0.7888 & 0.6902 respectively. The results of multiple linear regression indicated that the regression analysis that included dry density and water content have better predictions of swelling pressure. The empirical equations that could best predict swelling pressure are Ps  24.790  400.977dry  737.845 with R2 =0.946 and Ps  15.389PL949.682LI 503.619dry 228.458 With R 2 =0.925.