Correlation of CBR Values with Soil Index Properties by Regression Model using Soft Computing Techniques
Heena Malhotra1, Janmeet Singh2, Himanshu Jaiswal3

1Heena Malhotra, Research Scholar, Punjab Engineering College, Chandigarh (Punjab), India.
2Janmeet Singh, Research Scholar, Punjab Engineering College, Chandigarh (Punjab), India.
3Himanshu Jaiswal, M.Tech, Water Resource Engineering, Punjab Engineering College, Chandigarh (Punjab), India.
Manuscript received on July 02, 2018. | Revised Manuscript received on July 15, 2018. | Manuscript published on July 30, 2018. | PP: 9-13 | Volume-5 Issue-7, July 2018. | Retrieval Number: G1256075718
Open Access | Ethics and Policies | Cite
© The Authors. Published By: Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC BY-NC-ND license (

Abstract: The spatial variation of soil properties is beyond the designer’s control. Designer often feel discomfort before reaching at any conclusion and totally rely on the soil testing. This soil investigation takes much longer time and resources of a project. So geotechnical engineers usually attempts to develop empirical equations. But these empirical equations are more specific to the location & type of soil. However the empirical relation is useful for future projects coming in the vicinity. In road construction, civil engineers always encounter difficulties in obtaining representative CBR value for design of pavement. The type of soil is not the only parameter which affects the CBR value, but it also varies with different soil properties possessed by the soil. Laboratory CBR test requires relatively large effort to conduct the test and it is time consuming. Currently, many road construction projects and railway constructions are undergoing in the country. In light of this, the output of the proposed correlation will provide road authorities, railway authorities, consultants and contractors preliminary background information on the value of CBR, for a localized sub-grade material, from soil index properties with a benefit of time saving and without incurring any additional cost for carrying out laboratory CBR test. As a result, our present study aims to find the correlation between CBR values with soil index properties. So to develop correlation, Single line regression (SLR) & Multiple line regression (MLR) is done to correlate CBR value with soil index properties and their precision is examined by Statistical data analysis tool. Accordingly, 100 disturbed samples were collected from different location of Haryana district and required laboratory test have been conducted in order to establish an equation of CBR as a function of grain size parameters, atterberg limit by considering the effect of an individual soil properties and effect of combination of soil properties on the CBR value. The developed correlation lead to a regression value of R2 = 0.729, using SLR, while MLR generated relatively an improved value of R² = 0.650. After validating the established correlation with other empirical equation developed by other researchers, it was observed that correlation of CBR value with soil Indian properties is more applicable for preliminary characterizing the soil strength.
Keywords: California Bearing Ratio (CBR), Regression, Index Properties.