Factors affecting the prediction of wave-induced iceberg motion

Abstract

Wave-induced motion can contribute significantly to the kinetic energy of small icebergs, bergy bits and growlers. In order to assist in the design of offshore structures resistant to wave-driven iceberg impact, several factors affecting the prediction of this phenomenon were investigated. Specifically, the influence of iceberg shape on its wave-induced motion, the use of linear superposition to obtain estimates of iceberg response in irregular waves, and the use of linear diffraction theory to compute iceberg response amplitude operators (RAO's) were extensively studied utilizing data generated in a wave tank. The validity of these experimental data was first addressed; it was found that model scale factor (i.e. viscous distortion) had little influence on the observed response in regular waves. A comparison between models of different shape but similar mass and characteristic length indicated that model geometry plays a much more significant role. For all models, motion response in irregular waves was found to be accurately predicted using the linear superposition of the measured responses in linear waves at different frequencies and the measured wave energy spectra. This is true in spite of the obvious nonlinear behaviour exhibited in high significant wave height model seas. Finally, the use of linear wave diffraction theory was found, on average, to underestimate surge RAO's by 11% and overestimate heave RAO's by 34%. However, very large scatter was seen in the heave results primarily due to overprediction of resonance amplitudes. It is concluded that an appropriate extension of design procedures to account for irregular iceberg geometry would be to assign probability distributions to iceberg significant motions as functions of characteristic length and significant wave height, rather than to treat this as a deterministic connection.