The existing IMO intact stability criteria (IS-Code 2008) do not generally provide sufficient safety against dynamic stability failures such as parametric rolling for modern ships. Therefore, new stability criteria have been developed by IMO / SLF. These so-called Second Generation Stability Criteria shall ensure sufficient dynamic stability. The criteria are structured in a three level approach, where the first level consists of quite simple formulae. If a ship does not pass the first level, it is assumed that the ship is vulnerable to the phenomenon addressed, and the second level of criteria shall then be applied. This level consists of computations which are a little more complex, but they still treat the problems addressed in a strongly simplified manner. If now the ship does not pass the second level, a third level shall be applied to ensure that the ship can be designed and operated safely. This third level consists of direct calculation methods which shall be applied, however no criteria or procedures have yet been developed for this third level. We have applied the level 1 and level 2 criteria to a reference ship where a direct stability assessment has been performed during the design. The results showed extremely large scatter in the required GM-values of the criteria, and none of the criteria showed GM values roughly comparable to the direct assessment. The paper shows why the application of the criteria is challenging for the design of RoRo-ships and why a third level (direct assessment) is urgently required before the first two levels are put into force. Some conclusions are also drawn for the possible treatment of the new criteria in a stability booklet.

There is an ongoing discussion on safety guidelines to be considering more recent developments in ship design. Numerical simulations of ship motions are considered as powerful tool for the safety evaluation of a given design. However, the consequent use of numerical codes calls for their thorough validation which has to be performed both qualitatively and quantitatively. This paper focuses on a code used and further developed by the Flensburg Shipyard. For its validation, the capsizing scenario in steep wave sequences is realized in the wave tank first. The dedicated computer controlled experimental technique ensures the exact phase correlation of wave excitation and resultant ship motions. Thus, the registered wave and the track of the ship model in the model test serve as input to the numerical simulation which results in the specific motion time traces. These are now directly compared to the motion registrations from the model tests. First results of the validation by direct comparison of time series have been presented in earlier publications, still with the restriction that only a few cases have been investigated. In this paper, the promising method is applied to another scenario in a long-crested sea state including steep wave combinations. Different aspects are discussed which results in the conclusion that the method is feasible for free running ships in stern and stern quartering seas.