Correlation Of CBR, MDD And OMC With Field DCP Test For Low Volume Road In Ethiopia

Abstract

The design of low volume road and evaluation of the existing pavement requires the value of California Bearing Ratio (CBR) as an input parameter. However this method is time consuming and demands significant effort to determine in the laboratory. On other hand, Dynamic Cone Penetrometer (DCP) is a simple, rapid, hand held and low cost device used to evaluate the strength of in situ soils. Correlating these two parameters helps the designer to determine the CBR values easily from simple field DCP test. There is lack of correlation between CBR and DCP in Ethiopia for different types of soils in different regions. In this study, forty two disturbed soil samples were collected from Amhara, Addis Ababa City Administration, Oromia, Southern Nation and Nationalities People's and Afar Regions. The experimental program was carried out for samples collected. Field and laboratory tests were carried out. The field tests were carried out at different sites of all regions for Dynamic Cone Penetrometer (DCP) to find the DCP value. The laboratory tests were carried out for the basic geotechnical tests and California Bearing Ratio (CBR) tests to find the Optimum Moisture Content (OMC), Maximum Dry Density (MDD) and Soaked CBR values. Specific to the present study, statistical software (SPSS) was employed to investigate the significance of the individual independent variables. The correlation between CBR and DCP, MDD and DCP, OMC and DCP have been proposed for different types of soil. Accordingly, equations such as ln(CBR)=-0.39ln⁡ (DCP) + 4.828 for GW &GP soils, ln(CBR)=-0.65⁡ln(DCP) + 5.415 for GC soils, ln(CBR)=1.692ln⁡ (DCP) -2.107 for ML & SM soils, ln(CBR)=-1.96ln⁡ (DCP) + 7.375 for MH soils, The developed correlations are based on small number of sample ranging from six to ten. This resulted in moderate correlation strength. It will be improved if more sample sizes are considered. Finally, comparison of developed equation with existing correlation and actual laboratory test values was made. The existing correlation developed is single equation for granular and fine soils. Thus, the value of existing correlation is under predicted for granular soils and over predicted for fine soils. It is not suitable to be used in Ethiopia for prediction of CBR values of soils in general. Hence, the present study proposed new correlation equations that are suitable for prediction of CBR values for different soil types in Ethiopia.