Numerical Modeling for Prediction of Compression Index from Soil Index Properties in Jimma Town

Abstract

The compression index is one the compressibility characteristics concept to make estimates of soil responses, when one cannot conduct sufficient soil tests completely characterize a soil at a site. In this study, correlations are developed to predict compression index (Cc) from index tests so that one can be able to model Jimma soils with compression index using simple laboratory tests. The objective of the study is to predict the compression index from soil index properties in Jimma town. Undisturbed and disturbed soil samples from fifteen different locations of Jimma, where different clay soil is found, are collected. Laboratory tests like specific gravity, grain size analysis, Atterberg limit and one dimensional consolidation test for thirty test samples (at 1.5 m to 3.0 m depths per each of fifteen test pits) are conducted. From this test, compressibility soil parameters compression index (Cc) and swell index (Cs) are determined. From the results of limited tests, an indicative good correlation is observed between compression index and liquid limit, plastic limit, and plasticity index. However, a poor correlation has developed between compression index (Cc) and plastic limit (PL) when related to the other parameters. The developed correlations will be important inputs in modeling Jimma clay soils with regression analysis and artificial neural network model using simple index tests. The proposed model that obtained from the correlation between Cc and LL, PI is given as Cc = 0.0018(LL) + 0.0004(PI) +0.1231, R2 = 0.847 with 0.012 of standard error through multilinear regression analysis. In addition, the results of this study can serve as a basis for further study of such correlations on different clay soils in the country. The compression index of soils was mean 0.274, at least 0.227 and at most 0.33 and depended on clay soil class. The results showed that the correlation coefficient (R 2 = 0.912 and R 2= 0.841) was determined by neural network and regression method respectively. By comparing the values of R-value and error square mean (MSE) by two using methods, it was revealed that artificial neural network has the least error and the most accuracy. As a result, for estimating the compression coefficient in the study area, this method should be used