Behavior of a double hull in a variety of stranding or collision scenarios

Abstract

A series of nine tests was conducted to investigate the behavior of a double hull in a variety of stranding or collision scenarios. Cones of five different nose radii were made to model accident scenarios ranging from grounding on a sharp rock to stranding on a relatively flat seabed or shoal, and collision with a sharp bulbous bow of a fast ship to collision with a large bow of a VLCC. Three sub-series were designed in which the cones pressed shell plating, main supporting members and intersections of main supporting members. The test results reveal that the nose radius and the location of penetration have a very strong influence on the behavior of a double hull. Therefore, careful definition of accident scenarios is of crucial importance to assess the strength of ship hulls in accidents, and it is necessary to base the assessment on probability of accidents. Characteristics of the response of structural members were identified and idealized as simple theoretical models. Analytical formulae were derived and discussed. Primary damage mechanisms include membrane stretching of shell panel, onset of rupture, crack propagation, folding of main supporting members, and crushing of intersections of main supporting members. The new plate punching model captures the phenomenon that the load-carrying capacity of a plate depends on the size of the striking object. The plate perforating model predicts the reduced strength of plates with cracks. It reflects the observed test phenomenon that loads do not drop to zero even after rupture occurs in shell plating. A simple analytical method was developed to calculate the global strength of a double hull. The method takes geometrical parameters of seabed rocks or bulbous bows into account, and can be used for a wide range of different accident scenarios. Calculations using this method compared satisfactorily with the test results. This method can be easily incorporated into a probability-based framework to properly assess structural performance for a variety of damage scenarios. Similar to the Wang et al. (J Ship Res 41 (1997) 241) paper on raking damage, which uses only four analytical models, this method also requires only a common calculator to carry out the calculations. (C) 2000 Elsevier Science Ltd. All rights reserved.