Evaluation of a semi-empirical numerical simulation method for level ice breaking

Abstract

A semi-empirical numerical method for level ice breaking is evaluated in this thesis by full scale data from ice trial of S.A. Agulhas II. The data are analysed and organised systematically for evaluation use with certain accuracy. The original model together with some improved versions are evaluated in terms of ship resistance, time history of speed and average speed. Additional resistance source is identified. A comparison between empirical formulas and simulation methods is conducted to find the advantages and deficiencies of this method.

Results have shown that this model generally gives acceptable prediction on average resistance and velocity. However, there are obvious deviations in time history of velocity. Compared to empirical formulas, numerical method gives more ship-specific results. Besides, shoulder crushing is identified as an important phenomenon which introduces uncertainty in results. Randomness is suggested for ice breaking pattern to eliminate this problem.

As suggestion, a combined modification including dynamic bending and random breaking radius is recommended. Resistance due to ice rubble under level ice is required to be included in ship resistance. A combined ice thickness measurement including electromagnetic method and stereo camera is suggested to grasp more information of the ice. Model and full scale tests are required to investigate more about ice breaking process, especially breaking pattern. This model can be improved when some empirical coefficients are replaced by conclusions from test results.