Parametric Evaluation of Key Constitutive Components in a Bounding Surface Model for Unsaturated Soils

The lack of consensus on appropriate formulations for effective stress, the soil-water characteristic curve (SWCC), and the critical-state void ratio has plagued the constitutive modeling of unsaturated soils. This paper presents a numerical parametric investigation of the aforementioned modeling components using a rate-independent hyperelastic bounding surface plasticity model for unsaturated granular soils. The constitutive model was calibrated against experimental results for a mid-Atlantic silty sand, and stress–strain and volumetric response simulations were conducted using alternative formulations for Bishop’s effective stress parameter, SWCC, and critical-state void ratio. The study indicated that the choice of SWCC has the greatest influence on model accuracy, particularly in predicting volumetric behavior, with formulations that employ more parameters providing greater fidelity. It was observed that variations in the definition of effective stress have a comparatively limited impact, whereas accurate representations of the critical-state void ratio improve overall predictive capability. The findings are based on a single silty sand, an SWCC estimated from grain size distribution data, monotonic loading conditions, and a limited critical-state database. Accordingly, the conclusions should be interpreted within the context of the investigated material, constitutive framework, and stress range considered. The findings highlight the relative importance of key modeling components and guide the selection and calibration of constitutive models for unsaturated soils.