Ship design traditionally relies upon towing tank testing, which is costly and requires scaled down models. This method is prone to inaccuracies, especially when testing unconventional designs and fast vessels. Numerical simulations based on potential theory, albeit extremely cheap and fast, suffer from accuracy limitations when dealing with these vessels and wakes. The advent of faster computers and more developed RANS-based (Reynolds-Averaged Navier-Stokes) methods, has begun a new era of naval design.
For a RANS-based simulation to be of value to the designer, it has to meet the following requirements: in hydrodynamic cases it has to capture the deformed free surface, take into account dynamic sailing attitude (especially for fast vessels, and is even better if it can simulate the vessel motions in incoming waves 6 DOF motions), and it has to be computationally efficient. In aerodynamic cases the simulation has to deform the sails and rigging and converge to the 'sail flying shape' while remaining computationally efficient and handling very large models. To avoid the main problem of towing tanks and wind tunnels, both aerodynamic and hydrodynamic simulations need to be run at full scale.
Over the years, CHE has developed an efficient methodology that meets all of these requirements and more. It is based on the well-known multi-purpose solvers from CD-adapco. We have extended and coupled these solvers with other programs and systems to simulate 6 degree-of-freedom motions and obtain the flow over deformed sails and rigs.
The CD-adapco solvers should be credited for their ability to efficiently handle free surface flows. They were among the first to implement a free surface model (Volume of Fluid or VoF method) that was specifically tailored to problems in naval hydrodynamics, including wave breaking, spray, cavitation and complex geometry.
Additionally, the method developed at Cape Horn Engineering to couple the flow with the solid body motion has proven robust and efficient. It has been applied to several dynamic cases, showing that large amplitude motions, even ship capsizing, can be simulated. Cases for slamming, water entry, wave piercing, water on deck and planing craft jumping waves have been simulated, and the results have been successfully validated with towing tank tests (see validation for more information). The robustness and efficiency of our method is mainly due to the simplicity of tracking the vessel's motion without deforming the mesh or using complicated multi-mesh strategies. There are no restrictions on the vessel's speed or hull and appendage shape. Any hull configuration, including multi-hulls, or hulls with lifting surfaces, can be analyzed.
The solvers use a finite-volume method with fully unstructured meshes, allowing local refinement, non-matching grid blocks and moving grids with sliding interfaces. Meshes can utilize a new type of control-volume, the polyhedral cell, giving total flexibility in mesh generation and very accurate results. For sail and rig deformations, we rely on polyhedral meshes and re-gridding. In this case re-gridding is the best solution to handle the large deformations that occur, especially in downwind sails. Re-gridding using polyhedral cells and our automated technique is very fast and flexible.
Like most marine CFD programs, the first validations of these methods were done for the Wigley and Series 60 benchmarks in 1997-1998. These were followed by other cases more relevant to the industry, as shown in the list below. When validating these methods, the focus was not only on the forces, mainly resistance, but also on the dynamic sailing attitude (sinkage, trim, squat), and on the predicted motions for ships in regular waves. Check the Publications for complete information on validation.
More recently, our simulations have been extensively validated during the 32nd America's Cup campaign with challenger BMW Oracle Racing. Validation was done using more than 20 large models towed at the Institute for Ocean Technology in Canada, one of the most prestigious towing tanks for America's Cup Class yachts. The accuracy and benefits of the simulations were demonstrated so thoroughly that it was decided to cancel the last model tests of that campaign. We are very proud that today, in our current high profile design campaigns, the team designers are relying completely on our simulations.
© 2008 Cape Horn Engineering. All Rights Reserved.