Analysis of Foreship and Iceberg Impact Loads Based on Bayesian Networks

This paper presents an application of Bayesian Networks (BN) to the estimation of foreship and iceberg impact loads and corresponding damages. The realized BN models the energy dissipated in the collision between a foreship structure and an iceberg. The energy is estimated based on a simplified formulation of the external dynamics of the ship and iceberg collision. At present, the analysis focuses solely on collisions occurring in the front area of the ship, as this scenario is believed to be the most probable. BN requires probability distribution functions (pdfs) of the variables entering the energy formulation. The pdfs utilized are based on the literature available and the resulting pdf of the energy level is illustrated. The internal mechanics of the collision has been modeled by means of the non-linear finite element analysis (NLFEA) program LS-DYNA, with a user-defined material model of the iceberg. From the simulation, the energy levels modeled in the BN are thereafter associated with extents of the damage in the ship. Finally, conclusions and discussions for further improvements are presented.