Probabilistical distribution statistical analysis of Tianjin soft soil indices

In the design of soft ground treatment, the soft soil indices are the basic design parameter, and their changes have a strong geographical feature and significant uncertainty. Today, the soft soil indices statistical analysis content is mainly in the correlation and distribution regularity of physico-mechanical indices, which lack systematic study in different regions for different types of soft soils, and the statistical samples are from indoor tests and in-situ tests, without considering the impact of different test methods. In this paper, taking the soft soil in Tianjin as an independent statistical unit body, based on studying the change confines, variability, correlation, probabilistical distribution model of soil indices, some statistical relationship empirical formulas and probabilistical distribution models for physical indices, compression index, strength index were obtained; moreover, the correlation of shear strength results by straight shear test, triaxial test, vane shear test are analyzed; and a method for using random factor to evaluate the uncertainty involved in different test methods of soft soil parameters is proposed. The study results provide a basis and reference for defining the values of geotechnical parameters for soft soil.