Prediction of consolidation parameter using multiple regression analysis

摘要

Since existing empirical formulas have been proposed without clear verification of normality, they involve very high uncertainty in their design parameters. Therefore, in the present study, marine clay regions located in river estuary regions with deep soft layers were selected as study subject regions, the normality of those soil investigation data that were relatively highly reliable was verified, simple regression analyses of the data were conducted to propose prediction formulas, and the results from the prediction formulas were compared with the results from existing empirical formulas. Multiple regression analyses were conducted and the results indicated that regression models composed of G(S), e(o), LL were statistically significant in explaining compression indexes in both regions. It can be seen that the structures of the models presented in the present study changed depending on soil properties. In particular, because of diverse features of soil, G(S) had negative effects on C-C in the case of Pusan while having positive effects on C-C in the case of Kwangyang. In addition, the results of tests of equality between regional models rejected the null hypothesis indicating that parameters were statistically significantly different between the two regions.