| Résumé | Grounded ice pads as a base for exploration drilling platforms are built by artificially thickening the ice cover, made from either level ice or ice rubble, at the desired drilling location. An experimental set-up was assembled with the purpose of addressing outstanding issues regarding the sliding resistance of these structures. These simulations were done on a clay seabed. The ice was contained in a bottomless rectangular box, called the 'wagon'. The bottom of the ice was in contact with the clay, with a 4 m2 footprint, and normal stress levels up to 23 kPa. The wagon was pulled by an actuator over distances of up to 100 mm per test while monitoring the horizontal forces required to do so. A total of 29 successful tests were conducted with level ice and another 32 with ice rubble. In the case of level ice, an increase in normal stress only resulted in a small increase in shear resistance, when enough time was allowed for the clay to consolidate. With ice rubble, an increase in normal stress systematically led to a significant increase in shear resistance, presumably because of shorter drainage paths for pore water evacuation, suggesting this material should be more suitable than level ice for this purpose. Vertical motion upon increasing and decreasing the normal stress provided evidence for clay consolidation and swelling, respectively. A peak response in shear resistance is attributed to dilatation. The clay's undrained shear strength, measured with a vane, was consistently higher than the sliding resistance of the ice in all tests. The vane data could therefore not be relied upon to provide a measure of sliding resistance. Also, this parameter progressively increased to levels well exceeding the maximum normal stress applied to the clay. An alternative approach, based on the soil's effective internal friction response, is assessed and proposed for estimating sliding resistance of real ice pads and other structures resting on a clay-rich seabed. |
|---|
