The Unloading Modulus of Soft Soil: A Field and Laboratory Study
Improved infrastructure in urban areas often necessitates underground construction, where the unloading behaviour of the soil is of importance for predicting the amount of heave and how long it will take for the soil to swell. It is of particular interest to study the need for the installation of tension piles, as well as whether the tunnel will be subjected to settlement or heave in the long term. The objective of this study was to investigate the unloading behaviour of soft clay, and thereby reducing the uncertainties associated with this type of construction in future projects.
This paper focuses on the unloading modulus of soft clay and includes results from a field and laboratory study. Previous studies on the unloading of clay have mainly employed laboratory methods to determine the unloading modulus, and these results are compared with the results of this project. The field study was conducted at an excavation for the Götaleden tunnel in Gothenburg. Extensive measurements of heave and pore pressures were performed at the construction site during a period of over one year. Several piezometers, earth pressure cells measuring total pressure and pore pressure, heave gauges and extensometers were installed at the centre of the excavation. The results were analysed and the unloading modulus was estimated from the measurements.
The laboratory tests mainly consisted of incrementally loaded oedometer tests and drained triaxial extension tests. Bender elements were used in the triaxial apparatus to measure the initial soil stiffness. The results show a very high initial stiffness, which corresponds well with field measurements (Andréasson, 1979). The laboratory test results reveal that it is possible to capture the small strain behaviour in the clay, using advanced triaxial test equipment and measurement techniques. However, results from the field study indicate a noticeably stiffer behaviour, than ordinary laboratory tests. Differences in the results were expected due to the difficulty of recreating the in-situ characteristics of the clay specimen in the laboratory.
Proposed curves for the variation of the unloading modulus with stress level are presented for normally to slightly overconsolidated clays. The study shows that the unloading modulus has a large initial value, which drops considerably with the decrease in stress. The creep in the clay prior to unloading is a factor that has a major influence on the initial stiffness. The field results indicate an initial stiffness three times greater than that observed in the Bender element tests, which is likely to be an effect of creep or stress path rotation.
In the excavation at hand, the heave process seems to have terminated within three months of completion of the excavation, coinciding with the time at which steady state pore pressures were reached, which was considerably faster than originally expected.