A Review of Constitutive Modeling of Unsaturated Soils
نویسندگان
1 Associate Professor, Civil Eng. Dept., School of Engineering, Razi University, Kermanshah, Iran
2 Ph.D. Student, Civil Eng. Dept., School of Engineering, Razi University, Kermanshah, Iran
doi
10.30499/ijg.2025.514578.1687چکیده
The mechanics of unsaturated soils are integral to understanding and predicting the behavior of diverse geotechnical and geo-environmental systems, including natural slopes, engineered embankments, landfill covers, and agricultural fields. Unlike saturated soils, unsaturated soils exhibit a complex interplay among air, water, and solid phases, in which suction and partial saturation are pivotal in governing stress-strain responses, volume changes, and fluid flow processes. Over the past several decades, extensive theoretical, experimental, and computational efforts have culminated in sophisticated constitutive models, reflecting the need to include moisture content and suction as additional state variables.This article comprehensively reviews these modeling endeavors, tracing historical developments from empirical extensions of saturated soil models to contemporary elasto-plastic, multi-scale, and data-driven frameworks. Emphasis is placed on the evolution of stress state variables, the role of hydraulic hysteresis, and the bidirectional coupling between mechanical and water retention behavior. Special attention is given to recent models that incorporate bound water structure and dehydration mechanisms in expansive clays, highlighting their influence on retention properties, suction, and thermomechanical responses. The paper also explores advances in small-strain stiffness modeling (which is crucial for predicting seismic behavior of unsaturated soils), machine learning integration, and coupled thermo-hydro-mechanical-chemical (THMC) processes. Practical challenges, including parameter calibration, are examined. The paper also offers examples of prominent constitutive models, detailing their mathematical formulations and underlying assumptions. Current trends, including the integration of machine learning, are evaluated, and future research directions are proposed, underscoring the importance of interdisciplinary collaborations and long-term monitoring to refine and validate constitutive models for unsaturated soils.