Characterisation of Saturated Loose Sand Samples prepared by Fluidization

Abstract

The following thesis has three major purposes: (1) to explore the static liquefaction phenomenon in saturated sand as a relevant geotechnical hazard, (2) to show the conventional testing procedures for index characterisation and strength testing of saturated fine sands at low stress levels, and (3) to look into the fluidization mechanism and the ongoing research into its applications in saturated sand testing, done at Delft University of Technology. The exploring on static liquefaction was performed by a literature study, which allowed to relate the macroscopic scale of this hazard in underwater slopes with the common practices performed in laboratory conditions. The literature found shows that a change in the framework of understanding saturated sand behaviour, accounting for anisotropy, may occur within the next years. In addition, the literature supports that improvements on the liquefaction prediction can only be done via: improvements of: the theoretical framework in which the sand behaviour is studied, the constitutive and numerical models used for prediction of the liquefaction mechanism and the testing and the physical modelling methods allowing for both sand characterisation and the determination of the constitutive model parameters. Some tests performed to characterise a fine sand were selected to see how the method can affect the obtained results and how the conditions of the sample preparation and control can affect the liquefaction potential on a saturated sand sample. The results support that idea that the amount of fines on the sand, indicated generally by a sieving test, can affect the results of the index minimum and maximum void ratio, differentiating the properties of a sand with fines and a sand without fines. In addition, several variables such as the degree of saturation, confining stress levels and membrane penetration can contribute to the development of pore pressures and the development of liquefaction. The literature study reveals that fluidization has been used on research on granular soils sedimentation, documented in the work by Allen (1984), and is not as new in earth sciences as it was thought to be. Based on the undrained triaxial testing procedure suggested by Jager and Molenkamp (2015a), where fluidization will be used for sample preparation, some sand batches were prepared, in which layers were removed to characterise the final product. While the results on void ratio distribution along the depth of the column are not sufficient to draw further conclusions, it was found that the discharge rate can affect the porosity distribution during and after fluidization. The study not only showcases two techniques that can allow a quantitative assessment of fluidization mechanism, but also that inclusion of this method in loose sample preparation could bring improvements into granular soils characterisation.