The shear strength of the soil is an important factor to know the internal resistance of soil against external loads causing shearing forces. Shear strength parameters are mandatory for the analysis of load bearing capacity of the soil, the stability of Geotechnical structures and in analysing stress and strain characteristics of soils. The undrained shear strength is one type of shear strength parameter. This parameter is conducted by undisturbed samples. But due to handling, transportation, release of overburden pressure and poor laboratory conditions, it is difficult to obtain accurate undisturbed samples. So, prediction of undrained shear strength parameters (cu) for cohesive soil with the help of compaction characteristics provides a good alternative to minimize this problem.
Therefore, this study was conducted to develop the correlation between undrained shear strength values with soil compaction characteristics specifically located in Burayu town. The study was carried out using thirty samples collected from the 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 Atterberg's limits, Grain size analysis, Specific gravity, Compaction tests and unconfined compression test. The test procedures were based on AASHTO and ASTM laboratory test standards. These parameters are used to establish equations of correlations between undrained shear strength values with soil compaction. The soil type found in Burayu town was highly plastic red clay soil.
Based on both single and multiple linear regression analysis relatively good correlation is obtained by combining undrained shear strength (qu ) with maximum dry density and optimum moisture content of red clay soil. From the correlation analysis the equations developed are qu = - 3105 + 1625 MDD + 40.9 OMC with coefficient of determination of R2 =0.828 for multiple linear regression and qu= - 1473 + 57.8 OMC and qu= - 4861 + 3910 MDD with coefficient of determination of R2=0.787 and R2 =0.601 for single linear regression respectively.
Generally, the intended correlation obtained from the study area fulfil the basic requirement of regression.