Empirical Correlation of CBR and TVA Penetrometer Resistance for Coastal Soft Soil Assessment

This study aims to develop an empirical equation for predicting the California Bearing Ratio (CBR) value based on the results of the TVA Penetrometer (Hand Cone Penetrometer) test on soft soils, particularly in coastal areas. The soil samples were prepared by mixing bentonite clay, kaolin, and sand, then compacted and tested in the laboratory to obtain both CBR and QcTVA data. A stepwise regression analysis was conducted on four parameters: CBR, QcTVA, moisture content, and dry unit weight. The results indicated that only QcTVA was statistically significant as a predictor of CBR. The resulting regression model showed a coefficient of determination (R²) of 0.70 and a p-value of 1.90×10⁻¹⁷, with good accuracy based on the MSE value of 3.214, RMSE of 1.793, and MAE of 1.295. These findings suggest that QcTVA can be used as a practical method to estimate CBR values without requiring field tests involving heavy equipment. However, the model has limitations when applied to soil types with high CBR values. Therefore, its use is recommended only for soft soils or similar ground conditions, such as those commonly found in coastal regions