The ability of a DP-vessel to keep its position depends highly on the performance of the DP system. The thrust efficiency of the DP-system depends on the efficiency of the individual thrusters, but also on the interaction of the thruster wake and the hull of the vessel. This thruster-hull interaction becomes even more important when the vessel is a semi-submersible vessel; the thruster wake of the thruster on the upstream pontoon might impinge on the downstream pontoon resulting in high losses in efficiency and reduced DP-capability. Heerema Marine Contractors has two DP-semi-submersible crane vessels; the Thialf and Balder. An assessment of the thrust efficiency of the DP thrusters of these vessels has been made by comparing CFD computations with dedicated model tests.
In previous benchmark studies CFD is used to assess the current loads as well as thruster-hull interaction without current on a semi-submersible vessel. The logical next step is to perform a numerical study on a thruster-hull interaction with current. Similar as the previous benchmark studies the numerical data are validated with a series of dedicated model tests. The model test data include the global forces, the forces on each individual pontoon and the forces of each individual thruster, including the nozzle thrust and propeller thrust. The comparison between the CFD and model test data shows that CFD is able to predict the relevant force components within a sufficient accuracy for engineering purposes.
At present not much is known about the extrapolation of model scale DP-thrust efficiency to full scale DP-thrust efficiency, neither for model test results, nor for CFD results. Scaling CFD from model scale to full scale is not trivial; it involves a significant change in Reynolds number, a different description of boundary layer and poses challenges to meshing and grid. Therefore, validation is required. A first validation study is performed based on data acquired during a transit of SSCV Thialf in Q4 2011. In preparation, CFD simulations are performed for different thrust combinations. These results are compared to full scale observations and, where possible, improvements to the numerical modeling are assessed.
The paper addresses lessons learnt to improve the CFD computations as well as practical aspects and limitations of thrust efficiency modeling, including all interaction effects, using CFD from an engineering perspective.