Model-scale tests on ice-structure interaction in shallow water: Global ice loads and the ice loading process

Laboratory-scale experiments on ice-structure interaction process in shallow water were performed by pushing a ten-meter-wide ice sheet against an inclined structure of the same width. Seven experiments were performed in three series: In one of the series, the compressive and flexural strengths were both about 50kPa, in the two other test series the ice strength was two and four times higher. The ice thickness was about 50 mm in all experiments. The loading process showed two phases: the ice load on the structure (1) first increased linearly with a rate that was constant for all experiments, after which (2) the loading process reached a steady-state phase with approximately constant load. The magnitude of ice loads was not proportional to ice strength, as the weakest ice yielded higher loads than the ice having twice its strength. The ice rubble grounded in all experiments, but the bottom carried only a small portion of the load. The load records could be normalized by a factor combining the weight and the characteristic length of the intact ice. Based on the normalization, a model explaining the loading process was derived; the weight of the incoming ice has a dominant role during phase (1), while buckling explains the change in the process to phase (2) when the ice is strong enough. The loading process for the weakest ice was different from that for the other two ice types used. For example, instead of forming a rubble pile consisting of distinct ice blocks, weakest ice formed a dense pile of slush. The normalized ice load data highlighted the differences in the loading process.