An integrated approach of laboratory testing and field monitoring for the stability analysis of a partially saturated river embankment

Partial saturation is a key-feature of the materials typically forming the river embankments, providing a significant contribution in terms of hydraulic retention and slope stability to such linear infrastructure. Although unsaturated soil related issues has become more and more investigated in geotechnical research, nowadays their adequate integration in the framework of flood protection and risk mitigation is still rather limited. Such circumstance can be mainly ascribed to the strong uncertainties that still affect their practical application, like the limited information on the actual soil suction distribution in the river embankments and the substantial lack of a suitable characterization of the materials under partially saturated conditions. This study aims at investigating the importance of a proper mechanical characterization of the unsaturated soil for the riverbank stability assessment. In order to introduce a more general and realistic approach, the paper presents the experimental and the numerical results obtained for a well-documented case study, concerning an embankment section of the river Secchia, a right tributary of the river Po. In particular, a detailed analysis of the safety variations induced by time-dependent hydraulic and atmospheric actions has been carried out by integrating the numerical model with the information collected on site from the extensive monitoring system purposely implemented. The stability analysis has been thus performed considering different soil failure criteria and then comparing the results of standard and more detailed predictive analyses. A suitable laboratory testing campaign has been also carried out, focusing on suction-controlled tests and including direct shear and oedometer tests, which enable to accurately describe the soil mechanical and retention behaviour at site-representative states. The final scope of the work is therefore to provide a contribution to the development of more realistic riverbank stability analyses and to bridge the gap between researchers and practitioners in the use of unsaturated soil mechanics.