Theoretical Evaluation of the Mechanical Behavior of Unsaturated Soils

Abstract

An evaluation method for the mechanical behavior of unsaturated soils is studied in this paper. Although the mechanical behavior of unsaturated soils is complicated, a simple modeling is preferable in practice. This is because the soil properties are not homogeneous and ground data is limited when structures are being designed. In addition, in order to evaluate the reliability of the design, the physical meanings of the parameters applied in the prediction model should be clear. Firstly, the authors study the relationship between compaction curves and compression indexes in the unsaturated state that is used in the proposed constitutive model. Based on the constitutive model, the stress paths for constant volume shear tests are formulated under a constant void ratio condition and the stress paths for undrained shear tests are calculated under a constant water content condition. In the case of unsaturated specimens, the volume of these specimens changes with the shear deformation and the stress paths depend on the initial degree of saturation. The results of the calculation qualitatively describe the test results by considering the changes in effective confining pressure in the undrained condition and the water retention curves.

Modelling hydro-mechanical behavior for unsaturated soils

“…The experimental data of Honda (2000) and Jotisankasa (2005) indicates that NCLs for unsaturated states under constant suctions will gradually approach, but never cross, the saturated NCL.…”

“…Figure 15 gives an example to demonstrate the determination of the fitting parameter (a 2 ) in equation (18). This Figure shows the experimental results from suction-controlled isotropic compression tests performed by Sharma (1998), which are replotted in the plane of S e -v. The 100kPa suction data set (black square points) is used to calibrate the fitting parameter (a 2 ) in equation (18). As shown in Figure 15, this gives an a 2 value of 0.07.…”

“…In the following, the experimental results of a suction-controlled oedometer test from Honda (2000) are used as an example to demonstrate the calibration process for a 1 . For this series of data, the residual degree of saturation (S r res ) is unknown and assumed to be zero for simplicity.…”

“…As shown in Figure 15, this gives an a 2 value of 0.07. The remaining two data sets (at suctions of 200kPa and 300kPa) are used to validate equation (18). Both the calibration and validation curves are shown in Figure 15 (18) using Sharma's data (Sharma, 1998) …”