Stochastic fatigue assessment of a container vessel
Examensarbete för masterexamen
Fatigue prediction of ships is always a challenge. For container ships, that which, by default, are sensitive to fatigue because of their large deck openings and because of their operating in the extreme conditions of the North Atlantic, this prediction is even harder to make. Further studies are trying to establish a new method that can predict possible material failure in a more effective way. This study aims at performing stochastic fatigue analyses using a finite element model of a 4,400 TEU container ship according to the procedure of DNV software SESAM. These analyses could become a basis for the comparison of results between the standard method of fatigue prediction and a new simplified method. The analysis contains frequency domain hydrodynamic simulation, linear elastic structural analysis and stochastic fatigue analysis. The finite element (FE) model of the entire ship indicated several regions where severe fatigue damage is expected. One of these areas was chosen to be represented by a more detailed FE model. The bilge in the midship region was considered most critical because being under the waterline level, cracks are not easily noticed. This local submodel indicated that the highest fatigue damage is located at the connection of transverse web frames with the double side. A detailed FE model represents the double side with a transverse web frame and cut-outs for longitudinal stiffeners. From this analysis, the largest damage is expected at the cut-outs. The hotspot method, according to DNV recommendations, was used to obtain stress concentration factors at two points, which are then applied to the fatigue analysis of the global FE model of the entire ship. The stochastic fatigue analysis of the detailed FE model gives unrealistically high fatigue damage for some local areas. The assumptions and limitations of this study suggest that the resulting fatigue damage will be overestimated, although they cannot explain this extent of overestimation. Stress concentration factors and SN-curves are therefore examined as for how modifications in their values affect the results.
Transport , Farkostteknik , Transport , Vehicle Engineering