A damage sampling method to reduce A-index standard deviation in the probabilistic assessment of ship survivability using a non-zonal approach

The present SOLAS damage stability regulations for passenger and dry cargo ships address vessel survivability after flooding due to collisions with a probabilistic framework. This concept has been extended to other possible hazards responsible for flooding of a ship, such as groundings (bottom or side). Therefore, probabilistic distributions have been provided for damage locations and dimensions, enabling ship survivability assessment to be based on Monte Carlo (MC) sampling of pertinent distributions for generation of damage breaches using a flexible non-zonal approach. Such a method introduces randomness into the process, leading to a dispersion of obtained A-indices within different batches of generated damages. In the present work, a Quasi Monte Carlo sampling method is applied to generate multiple sets of bottom grounding damages on a reference test barge available in literature. The obtained A-index has a significant data dispersion reduction compared to standard MC samples of equivalent size, reducing the number of cases necessary to obtain an engineering significant value for A-index.