Multi-field coupling effect of deep excavation is investigated under the stress path of unloading. First, based on two surfaces model, an elasto-plastic model for clay under different stress paths is established. Second, the constitutive model is introduced into the Biot consolidation equations. Then the incremental frame of governing partial differential equations is established to be suitable for this problem. Third, a physical equation for pore water pressure and the fundamental solutions are presented for the plane strain problem. By use of the fundamental solutions, a semi-analytical and semi-numerical method is presented. Fourth, the method has been applied to numerical analyses of consolidation in deep excavation. It is indicated that this method is reasonable, which can reflect the effect of unloading stress path on horizontal displacement, settlement and pore water pressure.
Mechanics character of soil depends not only on the stress history, but also the current stress state and following stress path. Lamber (1967, 1979) put forward the stress path method for strength and deformation characteristics of soil under different loading conditions. After that, some tri-axial tests of stress path show mechanics characteristics of the soil are affected significantly by different stress paths (Charles 1999, Zhang & Chen 2002).
In numerical analysis of soil excavation, some constitutive model under some loading paths is chosen and unloading path is always achieved by controlling the life and death of the cell (Qian & Yin 2000). To some extent, this method can be more desirable for numerical analysis under unloading problem. However, there may be some discrepancy for the actual working conditions.
Soil sample of the experiment is taken from a deep foundation in Huangshi. Because the samples have been disturbed, the soil should be remolded as the test samples. The testing of basic properties of the soil and preparation of remolded soil is on the basis of soil test procedures. physical indicators of the sample are shown in Table 1.