This paper presents the results of 3D discrete element modeling of monotonic constant volume simple shear test on Pea gravel. 3D DEM simulations were validated using results from large-scale stacked-ring simple shear laboratory tests on real soils, where each particle was accounted for and was characterized by size and shape using the translucent segregation table (TST) test. To acknowledge and incorporate both the irregularity and non-uniformity of particle shapes in real soil specimen and providing a realistic representation of soil assembly in the numerical simulations, a non-uniform distribution of rolling resistance (obtained from the particle shape characterization using TST) was assigned to the spherical particles in the simulated specimens. Different aspects of soil behavior at micro- and mesoscale such as non-coaxiality, stress-induced fabric anisotropy and validity of boundary measurements in evaluating the soil response were investigated. It is shown that boundary measurements (as generally done in laboratory) lead to a conservative estimate of the soil strength and generated pore pressure inside the specimen.

An important drawback of the hypoplastic model is the inaccurate prediction of the sand behavior under undrained monotonic loading conditions. The model is not able to reproduce the limited liquefaction type response widely observed in undrained tests on loose sand, and it often underestimates the initial stiffness and hardening rate of sand during the shearing. To address these issues, three novel modifications are introduced into a basic hypoplastic model to enhance its undrained predictive capability. Firstly, a new factor is added to the nonlinear term of the model, allowing the simulation of a purely elastic response at the beginning of loading. By doing so, the model can accurately capture the initial stiffness and undrained effective stress path of sand. Secondly, the characterized void ratios are related to an evolving state variable, enabling the model to reasonably reproduce the limited flow response and quasi-steady state. Furthermore, a new term is incorporated into the deviatoric part of the strain rate to adjust the hardening rate of the model. The model performance for undrained loading is significantly improved through the above modifications, as evidenced by the good agreement between simulation results and experimental data for tests with varying densities and confining pressures.

Geogrid stabilization has gained significant attention in recent years as an effective method for enhancing the performance of subgrade soils. However, the reinforcement effect of the geogrids under different loading conditions has not been thoroughly investigated, which hinders a comprehensive understanding of subgrade stabilization. Therefore, this paper aims to investigate and compare the behavior of a stabilized subgrade with geogrids reinforcement under cyclic loading and monotonic loading conditions. The experiments were conducted within a steel model box measuring 1.0 m (length), 1.0 m (width), and 1.2 m (height). The subgrade layer was consistently maintained at a thickness of 500 mm and strength of 2.5% California Bearing Ratio (CBR). A granular layer of high-quality material with a thickness of 200 mm was applied on top of the weak subgrade and geogrid was placed at the interface between the granular layer and subgrade. The tests were conducted in a controlled laboratory setting, specifically measuring vertical displacement in response to monotonic and cyclic loading. The results were then analyzed to evaluate ultimate bearing capacity, stiffness and rutting thereby estimating the effect of geogrids on stabilization of weak subgrades. These findings are anticipated to contribute significantly to the development of design guidelines for stabilized subgrade with geogrids reinforcement. By incorporating these insights, the design, and optimization of geogrid reinforcement systems for subgrade stabilization can be enhanced, ultimately resulting in improved performance and increased longevity of transportation infrastructure.

In this study, a hypoplastic model is developed to describe the mechanical behaviors of cemented sand under both monotonic and cyclic loading conditions. A state variable is proposed to qualify the bonding strength, and it is incorporated into the model to reflect the influence of cementation on the strength, stiffness, and dilatancy of sand. To reflect the bonding degradation, this variable evolves during the shearing following a simple evolution rule and may vanish after large deformation. The critical void ratio and friction angle are related to the initial cemented content to consider the variation of the critical state induced by the cementation. The model is subsequently extended to account for cyclic loading by incorporating the intergranular strain, fabric change effect, and semifluidized state. The capability of the model is demonstrated by simulating the behavior of cemented sand under both monotonic and cyclic loading conditions.

During service period, offshore wind turbines are subjected to both monotonic and cyclic loads, causing the rotation or translation of caisson foundations in the seabed. However, most of the existing studies focused on the performance of caisson foundations under monotonic static loading, and there are few studies about the effects of caisson-soil contact mode and soil strength reduction during installation in numerical simulations. This paper therefore systematically investigates the bending moment capacity and failure mechanism of caisson foundations under monotonic and cyclic loading in clay using finite element analyses. Three typical caisson-soil contact modes in different loading scenarios are considered, and the influence of soil strength condition, caisson aspect ratio on the bending moment capacity and failure mechanism of caisson foundations is explored. It is found that under monotonic loading, the bending moment capacity in the tensionless mode and the fully-bonded mode could be used as the lower and upper limit, respectively. Under cyclic loading, the fully-bonded mode always yields the highest moment capacity, while the frictionless mode and the tensionless mode produce the lowest in the case with small loading amplitude and the case with large loading amplitude, respectively. In addition, the behavior of cumulative angular displacement under combined load of wind and wave is also studied to provide insight for caisson foundation design.