GEOLAB - RES FLUCTIS project: Assessing River Embankments Stability to FLoods through Unsaturated Centrifuge Testing In transient Seepage conditions

River floods rank among the most significant natural hazard in Europe, causing substantial economic and human losses and are frequently due to severe damages endured by water retaining earthworks, under extreme weather events. Forecasting models suggest that climate change will be a determining factor in intensifying the hydrological cycle, with a far-reaching impacts on runoff regimes of streams and watercourses, thus significantly reducing the return times of droughts and flood events. This means that existing water retaining infrastructures, such as earthen river embankments, are expected to withstand extreme stresses, being subjected to hydraulic loading hardly experienced in the past, in terms of duration and intensity.  

In this framework, a reliable assessment of the existing river embankments safety conditions, represents a key aspect to enhance the resilience of these critical infrastructures (CI). A satisfactory solution to this problem cannot disregard the partially saturated state of the earthfill, neither the role of river stage fluctuations on the seepage process within the embankments. Nevertheless, in the current engineering practice, these aspects are frequently neglected, mainly due to the difficulties in estimating the actual suction distribution and shear strength, thus providing erroneous conclusions on the effective safety margins towards potential failure mechanisms.

The RES FLUCTIS (Assessing River Embankments Stability to FLoods through Unsaturated Centrifuge Testing In transient Seepage conditions) project aims at contributing to a better understanding of the effect of time-dependent hydraulic loadings on the stability of a river embankment model, subjected to simulated high-water events, through a series of centrifuge tests.

The experiments were carried out at the Schofield Centre of the University of Cambridge (UK), on the 150 g-ton Turner beam centrifuge, with a nominal radius of 4.125 m and a payload of 1000 kg. In the first centrifuge test, water level was gradually increased, up to three incremental elevations, and each of them was maintained until a steady-state flow regime was established within the embankment body.

A sequence of hydrometric peaks and drawdowns of constant intensity and duration, was applied in the second test, with the aim of replicating wetting-drying cycles, which river embankments typically experience under site conditions.

The small-scale physical model, tested under the enhanced gravity field of 50 g, was characterized by a compacted embankment, made of a natural silty sand, representative for a typical embankment section of the Alpine and Apennine riverbank systems of the main river Po (Northern Italy), which have recently experienced multiple overall collapses and breaches. The earth structure was founded on a fully saturated homogeneous Speswhite kaolin layer, 1D consolidated.

To investigate the hydro-mechanical behaviour and the possible failure mechanisms of the CI model, induced by stationary and transient hydraulic boundary conditions, in-house built miniaturized tensiometers, pore pressure transducers, displacement sensors, high-resolution cameras and flow meters were installed.