Extensive experimental studies have demonstrated the time-dependent mechanical behaviors of frozen soil. Nonetheless, limited studies are focusing on the constitutive modeling of the time-dependent stress-strain behaviors of frozen clay soils at different subzero temperatures. The objective of this study is to numerically investigate the time-dependent behavior of frozen clay soils at a temperature range of 0 degrees C to - 15 degrees C. The Drucker-Prager model is adopted along with the Singh-Mitchell creep model to simulate time-dependent uniaxial compression and stress relaxation behaviors of frozen sandy clay soil. The numerical modeling is implemented through the finite element method based on the platform of Abaqus. The constitutive modeling is calibrated by a series of experimental results on laboratory-prepared frozen sandy clay soils, where the strain hardening, the post-peak softening, and stress relaxation behaviors are captured. Our results show that both the rate-dependent model and creep model should be adopted to characterize a comprehensive time-dependent behavior of frozen soils. The rate-dependent stress-strain behaviors heavily rely on the rate- and temperature-dependent hardening functions, where the creep strain provides a very limited contribution. Nevertheless, the creep strain should also be adopted when a long-term analysis or stress relaxation behavior is involved.

Ground improvement is necessary in many flat areas and landfill sites in Japan because these areas have soft ground and are highly susceptible to serious damage such as long-term consolidation settlement and liquefaction. The deep mixing method (DMM) is an in-situ soil treatment in which native soils or fills are blended with cementitious and/or other materials. Ground treated by DMM has higher strength and lower compressibility than untreated ground. However, there are quality problems in this method due to a condition in which a mixture of soil and materials adheres to and rotates along with the stirring blades without performing efficient mixing. Therefore, our purpose is to improve quality of improved columns produced by the mechanical mixing method using vertical rotary shafts and mixing blades. In this study, five cases of small-scale model experiments were conducted with changing in the blade rotation number, the incident angle of agitating blades, and the agitating blade angle. Strength tests were conducted using unconsolidated samples at different depths to investigate strength distribution, and needle penetration tests were also conducted. From the results, the effects of the blade rotation number, the incident angle, and the agitating blade angle on improvement quality were discussed.