No abstract

No abstract

Numerous studies in the literature are concerned with proposing new constitutive models for sands to simulate cyclic loading. Despite considerable progress in this area, there are various limitations on their simulation capabilities that are either overlooked or not communicated clearly by the developers. A number of these limitations are rather crucial for the end users, and therefore, providing discussion and analysis of them would be of great value for both applications and future developments. The present work is devoted to discussing seven characteristic limitations, which are frequently observed in cyclic loading simulations. Four advanced constitutive models are considered in this study: two bounding surface elastoplasticity and two hypoplasticity models-with the models in each category following a hierarchical order of complexity. Relevant cyclic loading experimental test data on Toyoura and Karlsruhe fine sands support the analysis. The key issues discussed include stress overshooting, one-way ratcheting in cyclic strain accumulation, liquefaction strength curves, stress attractor in strain-controlled shearing, hypoelasticity, cyclic oedometer stiffness, and effect of drained preloading. The presented results elaborate on the specific limitations and capabilities of these rather advanced models in simulating several essential aspects of cyclic loading of sands.

In the present work, a constitutive model for anisotropic clays is proposed and evaluated under monotonic and cyclic loading. To that end, the hypoplastic model for anisotropic clays by Mašín is extended to consider the effects observed on cyclic loading. The extension follows from the intergranular strain anisotropy (ISA) approach by Fuentes and Triantafyllidis. Similar to the intergranular strain (IS) theory by Niemunis and Herle, the ISA extension enhances the model in many aspects, such as the increase of the stiffness and reduction of the plastic strain rate on cyclic loading. Different monotonic and cyclic tests are simulated to evaluate the model's capabilities with the use of samples having vertical and horizontal bedding planes. Simulations obtained with the proposed model and the three-surface kinematic hardening model with transverse isotropic stiffness according to the work of Graham and Houlsby in 1983 (hereafter denoted A3-SKH) are directly compared and analysed. Some additional simulations with the same hypoplastic model extended through the conventional IS theory are also included. The results indicate that a better agreement of the pore pressure accumulation, and some effects emerging from the sample anisotropy, such as ‘inclined’ q–p cycles under undrained conditions, and the dependency of the pore pressure accumulation on the bedding plane direction, are provided by the proposed model. On the other hand, the A3-SKH model estimates with more accuracy the vertical strain accumulation on samples with vertical and horizontal bedding planes.