Developing correlation model for UCS from DCP and index properties of soil (A case study of Agaro Town)

Abstract

Characterizing field material properties by using laboratory tests is an ongoing problem in the discipline of geotechnical investigations. It is difficult to collect and test representative samples, and because of this, there is a discrepancy between laboratory test results and in situ soil conditions of environment. This usually leads us to use of unreliable designing methods. To avoid such problems Correlations Dynamic Cone Penetration Index (DCPI) with Unconfined Compressive Strength (UCS) and index properties that make it interesting alternative, due to operating quickly, very light, versatile and user-friendly property of DCP . The dynamic cone penetrometer (DCP) which is a simple test device that is inexpensive, portable, and easy to operate and understand. Unconfined Compressive Strength (UCS) is one of the type of laboratory testing. This laboratory testing is conducted by undisturbed samples. It is difficult to obtain accurate undisturbed samples. Even some designers or companies do not conduct soil investigation for building. Some of these predictions may leads to unexpected failures of structure or uneconomical design. So, prediction of Unconfined Compressive Strength for cohesive soil with the help of Dynamic Cone Penetration Index (DCPI) and index properties of soil provides a good alternative to minimize this problem. To this end, the present research aimed to develop single and multiple correlations of UCS, DCP and index properties of soil and also compare, validate and evaluate the developed model using the controlled test and with related existing model. This research consists of field testing, laboratory testing, and analysis of the results for 30 samples from 15 test pit collected from different location of town. By using the test result regression based statistical analysis was carried out to develop the intended correlation. The parameters considered for this study are Natural moisture contents, Atterberg limits, dry unit weight and specific gravities. The test procedures were based on ASTM laboratory test standards. Regression models were develop using SPSS software and Microsoft Excel for this study to enable the prediction of UCS values. From the study predicting Unconfined compressive strength (UCS) is obtained from multiple linear regression analysis and given by UCS = 202.211-0.673 DCPI +6.03ɣdry -0.406LL -1.511NMC, coefficient of determination (R²) = 0.918 .The results are expected to have wide application in the construction sector