Comparing rock shape models in grounding damage modelling

Groundings are among the most common and destructive maritime accidents. Sea bottom shape influences greatly what kind of damage the ship structure suffers and whether this leads to loss of water tightness. Sormunen et al. [21] presented and statistically compared rock models used in grounding damage analysis with detailed bottom shape data from two Finnish harbour fairways. The results were promising in terms of statistical fit especially for the binormal rock model, which also showed a wide range of flexibility in representing different types of sea bottom shapes. However, this measure does not explicitly tell if the model rock in grounding damage analysis results in similar damage size as real rocks cause. To test this, this paper develops a framework for studying, testing and evaluating rock models in terms of resulting grounding damage. FEM is used to analyse and compare grounding damage of rock models to the actual rock using otherwise identical grounding scenarios. Analysis is performed for four different real rocks with each rock being modelled with four different analytical rock models as well. The results show that rock models with good statistical fit did not necessarily result in similar grounding energy compared to results using the real rock. Differences in energy are caused especially by the rougher surface of the real rocks. For the similarity in rock area and the damaged ship structure element volume the results are much better, especially for the binormal model. As such another criteria for evaluating rock models for grounding damage analysis is needed. The results show that the damaged material volume is strongly linearly dependent on the rock area- and volume metrics. A similar linear dependency exists between the damaged volume and the energy dissipated in grounding. Knowledge of these relationships can be used towards estimating grounding damage of ships in future investigations, but rock surface unevenness should be evaluated as well.