The main purpose of the project has been to develop a rational procedure for designing new crashworthy side structures for those ship types where it could be expected that a substantial improvement of the crashworthiness and the related safety could be achieved by careful consideration of the structural design. For a tanker vessel or other vessels carrying hazardous cargo, damage is not acceptable if it results in cargo outflow with disastrous consequences to the environment. Likewise, the foundering of a passenger vessel can be disastrous with loss of many human lives. A major challenge in collision and grounding analysis is the prediction of the onset of fracture and crack propagation in the shell plating. In simulations of accidental loading on ships it is crucial that fracture is determined correctly, as it will influence the global deformation mode and the amount of damage to the hull and thus determine which compartments will be flooded as well as the amount of oil outflow. The most commonly used failure criterion in large-scale FE-simulations has been the equivalent plastic strain. However, it is well known that equivalent plastic strain is not suitable as a fracture criterion when a structure is subjected to biaxial loading. The main aspect of this thesis has therefore been to study the fracture criteria available in the literature and to validate them against various experiments with varying stress and strain states.