Assessment of the impact of climate change on an instrumented embankment:
an unsaturated soil mechanics approach

Abstract

Climate change has the potential to affect any transportation network that comprises embankments and cuttings built with soil material in an unsaturated state that is exposed to the climate. The BIONICS project (BIOlogical and eNgineering Impacts of Climate change on Slopes) aims to investigate how climate change will affect the serviceability and safety of earth structures. Part of the BIONICS project was to build a full-scale highly instrumented embankment combined with an automated climate control system.

Studies on the fill material used in the construction of the BIONICS embankment were carried out to understand the hydro-mechanical behaviour of the material, which is a sandy clay of medium plasticity. This involved the determination of the soil water retention behaviour and the mechanical behaviour under unsaturated conditions. Soil water retention curves (SWRC) were determined by a series of tests performed on compacted samples comprising various techniques (filter paper, psychrometer, high capacity suction probe and pressure plate). Total and matric suction SWRC following primary drying paths from 25% of water content were determined. In addition, a series of tests with the filter paper on samples at lower water contents (15%, 20% and 22%) was also performed. The SWRC following drying paths showed behavior similar to scanning curves intercepting the primary curve around 3000 kPa (11% water content). However, SWRCs that followed wetting paths showed atypical behaviour by intercepting the primary drying curve.

For the investigation of the mechanical behaviour a series of constant water content triaxial tests were carried out in double cell triaxial cells on as-compacted samples, and also samples wetted and dried from as-compacted conditions of 15%, 20% and 22%. A test series of samples tested in a saturated state was also performed to provide a reference state for the unsaturated tests. The unsaturated test series showed that the slope of the critical state line (CSL) in deviatoric stress space (M) was found to be similar for all water contents. The slope of the CSL in ν-ln(p-uw) space (λ) was found to be similar for all water contents, however the CSL shifted position due to variation in the intercept, Γ. Since specimens were at high degrees of saturation, calculations based on effective stress showed a reasonable interpretation of the data. However, a better agreement was achieved using the Bishop’s average stress assumption.

A new field measurement system to continuously measure pore water pressure at different depths using high capacity suction probes has been developed. This system was installed at the BIONICS embankment in two different panels (well and poorly compacted). In the well compacted panel pore water pressure behaviour had the tendency to increase with depth, always recording values that were slightly negative at shallower depths and positive at greater depths, showing profiles roughly parallel to the hydrostatic line suggesting that the material was close to saturation. In the poorly compacted panel the behaviour was found to be more variable showing abrupt reactions from the probes to weather events.

The differences in behaviour between the well compacted and poorly compacted panels could be related to the laboratory investigations. The well compacted panel was more homogeneous and less permeable (10-11 m/s). The poorly compacted panel was more heterogenic, more permeable and hence, during monitoring, showed more dramatic changes in pore pressure compared to the well compacted panel.