| Popis: |
California Bearing Ratio (CBR) is a useful method to assess the strength of different pavement layers by comparing them with the strength of standard crushed rock The main objective of this research is to find correlation between California Bearing Ratio with classification test parameters/index properties these are liquid limit, plastic limit, plasticity index, Optimum Moisture Content (OMC) and Maximum Dry Density (MDD) of soils taken from different places of burayu town so as to obtain accurate CBR values in less time. The research is descriptive research, CBR tests were performed on forty one soil samples by adopting primary and secondary data quantitative method of research approach was applied. Regarding to the quantitative data; Soil samples were disturbed and collected at a depth of 1 meter then tests are performed on SW-SM type (Well graded sand containing silt) soil in laboratory according to AASHTO and ASTM procedures to obtain the CBR value experimentally and on the other hand the correlation is established. Then prediction of CBR value by using index properties of the samples were done using multiple nonlinear as well as linear regression analysis, predictive equation by means of Artificial Neural Network in addition to STATA estimating with CBR as the dependent variable plus explanatory variables are liquid limit, plastic limit, plasticity index, Optimum Moisture Content (OMC) and Maximum Dry Density (MDD) The equations are developed by using genetic programing toolbox in MATLAB software as well as linear regression model using STATA software empirical correlation was found with a correlation coefficient R2 = 0.9954 with average mean square error value of 0.043 for Neural Network then again R2 = 0.7927 and average mean square error value of 1.0763 for STATA therefore satisfactory empirical correlation was obtained by using Neural Network than linear regression. The results from analysis showed that all the explanatory variables were predicted with the explained variable better in Neural Network than in STATA. Among these variables, maximum dry density (MDD) is highly predicting for CBR value |
