Computation of Scale Effects in Free-Surface Flows near a Ship’s Transom

In the European project VIRTUE, an EC-funded project under the 6th Framework program, several participants have cooperated in carrying out research to advance the role of CFD in ship hydrodynamics and design. One of the 5 work packages dealt with the computation of the steady flow around a ship hull in still water. A prominent theme in this work package was the improvement of the computation of free-surface viscous flow around the hull and the wave pattern and their scale effects. Strong improvements were obtained in the numerical accuracy, which was clearly demonstrated by the fact that in a VIRTUE workshop in 2007, in which participants were asked to submit computations of the viscous flow around a container ship, the 'Hamburg Test Case‟, two methods that differ in almost every respect gave very similar results as well in the computed wave patterns as in the wake fields, resistances and scale effects.
In this paper we focus on the computation of the viscous free-surface flow including scale effects for a more difficult test case: a container ship with a partly wetted transom. This wetted transom poses some extra challenges for a computational method: the transition between the wetted and dry regime and the details of the flow and the free surface (possible wave breaking) behind the wetted part of the transom are hard to capture. The challenge is to compute accurately the flow both on model scale and full scale. Scale effects near transoms can become very important when at full scale a smaller part of the transom is wetted than at model scale, since this can have a strong effect on the resistance. We will present mutual comparisons of wave elevations computed by three free-surface RANS methods, together with a comparison with freesurface measurements at model scale. We will focus on the results and scale effects in the free surface near the transom. There are strong differences between the computational methods in the modelling of the viscous flow and the free surface, as well as in the iterative method. The question is whether or not we can get good agreement with measurements at model scale, and how well the computed scale effects are in common agreement.