CBR PREDICTIVE MODEL FROM SOIL INDEX AND COMPACTION PROPERTIES FOR QUICK SUB GRADE SOIL STRENGTH DETERMINATION (IN CASE OF FINE GRAINED SOILS OF DEBRE–TABOR CITY)

Abstract

This research is an effort to correlate CBR value with an index properties and compaction characteristics of fine grade soil. California Bearing Ratio (CBR) is a common and comprehensive field or laboratory test method currently practiced in the design of pavement to assess the stiffness modulus and shear strength of sub-grade, sub-base and base materials so as to determine the thickness of overlying pavement layers. But determination of CBR value is time consuming and needs large amount of material. As a result, this research evolves to develop an efficient simplified California Bearing Ratio (CBR) predictive model from soil index and compaction properties targeting quick sub-grade soil strength determination specific to Deber-Tabor fine grained soils. Specific to this research, NCSS-12 statistical software with subset selection option is employed to investigate the significance of individual independent variables with the soaked CBR. During multiple regression analysis, subset selection with interaction option of the NCSS-12 statistical software is used for the task of finding small subset of the available independent significant variables that does a good job of predicting the dependent variable. The normality of the data was checked and an appropriate data transformation when necessary is applied. One best model from each category with a very good statistical goodness of fit measures was selected. The laboratory results indicated that samples used in this research lie in MH categories based on unified soil classification system and in group A-7-5(16) based on AASHTO classification system. The developed models were validated against primary data. Moreover, the newly developed models are found to be by far better and can be used as a simple convenient tool to predict the CBR value of fine-grained soils in the study area, among the various parameters derived from index properties and compaction characteristics the Liquid limit and maximum dry density were found to be the most effective predictive parameter with R2 = 0.899. The comparative results showed that the variation between the experimental and predicted results for CBR falls within 7.86% confidence interval.