RoPax Ship Collision – a Methodology for Survivability Analysis

University dissertation from Chalmers University of Technology

Abstract: Throughout the world, ships are continuously being declared as total losses and 10 to 15% of these accidents are collisions between ships. The consequences of a collision are diverse and depend on the ship type, e.g. oil outflow in the case of damaged tankers or loss of stability in damaged passenger ships. This thesis focuses on RoPax ships, which in damaged conditions are related to high risk due to the large number of persons on-board. The main objective of the work is to contribute to RoPax shipping’s further development of sustainable transport and maritime safety. The work contributes to knowledge and understanding of in what conditions a RoPax ship damaged in a collision will survive without capsizing and how these can be simulated accurately using numerical models. In order to determine the consequences of the survivability of a RoPax ship struck in collision from the shape and size of the damage opening in its side-shell, a computational methodology is presented. It is sequential (de-coupled) and incorporates a non-linear finite element (FE) analysis of a collision, followed by dynamic damage stability simulations due to flooding. By means of this approach the conditions for the survivability of a ship struck in a collision, which for a RoPax ship is the time to capsize, can be assessed. The influence of variations in input parameters to the computational methodology is studied. Uncertainties of parameters in the FE analyses include dispersion in material parameters, material failure criterion and its representation, model representation of the striking bow section, friction coefficient, collision angle and ship speed. The influence of these parameters on the shape and size of the damage opening area and time to capsize of the struck RoPax ship is assessed. Recommendations for a sufficient level of simplifications in the models and analyses for arriving at reliable results in a numerical simulation of ship collisions are made. A significant part of the thesis is dedicated to the model uncertainty that relates to a possible (user-related) insecurity in the selection of criterion for material damage and rupture in ship collision simulations using non-linear explicit FE analyses. Several criteria are compared, such as the Shear, FLD and FLSD criteria, and assessed by comparison between experiments and numerical analyses. Tensile tests are used to study the dependence on a length scale of the fracture of the material. A relationship similar to Barba’s law was established which relates the fracture strain of the material to the length scale (element size) in the FE analysis. Forming limit tests are used to study the dependence on a multiaxial strain state. A small-scale ship-like structure subjected to impact loading is used as a reference structure in the assessment of the criteria. Results from FE simulations are compared to and validated using experimental results and recommendations for procedures for a numerical analysis of collision simulations are presented. Conceptual crashworthy side-shell structures that follow either the ductility or the strength design principles are assessed with a conventional reference structure. The assessment is made by comparing the intrusion depth before rupture of the inner side-shell of a double-hull structure occurs, energy absorption during the indentation, the final damage opening area as well as the weight and manufacturing costs of each structure. The results provide a basis for the discussion of the potential and challenges related to the implementation of each structure.