Application of the Cramer Method for the Prediction of Geotechnical Parameters of Soils in the Intertropical Zone in Road Construction.

Abstract

Geotechnical road construction work requires many tests that integrate several parameters, which can be reduced by predictions. This paper establishes the prediction through the Cramer multiple linear regression technique of the modified Proctor parameters and California Bearing Ratio (CBR) parameters in function of Fines and Atterberg limits, lateritic materials developed on metamorphic rocks. The parameters such as Maximum Dry Density (MDD), Optimum Moisture Content (OMC) and CBR index as a function of unknown %fines (˂80µm), Plasticity Index (PI) and Liquid limit (LL), were determined. This study in soil mechanics shows that, the CBR (13 to 53) and MDD (1.5 to 2.1 g/cm3) increase as the percentage of fines (75 to 12%) in the soil decreases. These CBR values can be used for backfill and foundation layers of traffic T1 to T2/T3. The Principal Component Analysis (PCA) reveals the presence of three classes C1 (LL, PL and OMC), C2 (˂80µm) and C3 (CBR and MDD), representing two families of related parameters in C1 and C3 respectively. The multiple determination coefficients obtained are 0.998 with R2 = 0.996 for MDD; 0.991 with R2 = 0.982 for OMC and 0.98 with R2 = 0.97 for a CBR at 95% of optimum respectively, indicating an excellent correlation. This prediction helps to handle smaller quantities of materials, save time, reduce energy expenditure and the cost of analyses and field campaigns.