Fluid Structure Interaction on a Porpoising Boat
ANSYS Fluent | CFX | Structural | Drop test
I carried out this simulation while working for a boat designer who wanted to identify the regions of stress concentration on a preliminary design of a boat that he was planning to build. CFD was chosen for this study because the empirical guidelines used by mainstream boat builders, do not apply to boats capable of running at 60 knots made of carbon fiber.
Initially the CFD simulation was planned on ANSYS Fluent because of it's highly reliable simulation results when used correctly. A user defined function (UDF) was used to model the physical properties of the boat and the torque at the propeller. Since it was a two-phase simulation there were quite a few problems with respect to convergence on ANSYS Fluent.
Thus we shifted to ANSYS CFX which was known to give fewer problems when it came to converge. Once a basic mesh was made and the pre-processor was setup, we decided to validate CFX against two experimental results for the peak pressures obtained on the underside. The first one was a simple wedge vertical drop test (Fu et al. 2012) where the authors first verified their solver NFA (numerical Flow Analysis) against an experiment and then went ahead to simulate a high speed boat as well. The experimental validation was based on dropping a wedge from a certain height and observing the peak pressures. When performing the wedge simulation it dawned on us that to capture realistic pressure peaks the closest mesh element length would have to be less than a millimeter and the time step necessary would have to be in the order of micro seconds. If this would have to be scaled on to the 15 meter boat, the elements necessary would be in the order of billions which was not possible for the computational resources we had at hand. To verify if this claim was accurate, we verified against another validation which was more a more complicated case of pressures observed on the bottom of sea-planes' landing gears (Smiley R. F., 1952) which was chosen because it had comparable speeds to the high-speed boat. We observed that the pressure peaks were being under predicted for the sea-plane too and thus we had to find a new plan to observe the pressures on the underside of the boat with our limited resources. Finally, as what some would call a stroke of genius, we Thus we computed the pressure peaks on a relatively coarse mesh and scaled it according to our knowledge from the wedge simulation and empirical theories particularly the DNV (Det Norske Veritas) rules.
Finally these pressures were transferred over to ANSYS Transient Structural and the thickness of the various structures of the boat was corrected accordingly.
At this moment the boat is under construction and will be tested in late 2017.
 A Detailed Assessment of Numerical Flow Analysis (NFA) to Predict the Hydrodynamics of a Deep-V Planing Hull, Fu, T. C.and O'Shea, T. T. and Judge, C. Q. and Dommermuth, D. G. and Brucker, K. A. and Wyatt, D. C., 29th Symposium on Naval Hydrodynamics, 2012
 A theoretical and experimental investigation of the effects of yaw on pressures, forces, and moments during seaplane landings and planing, Robert, S. F., National Advisory Committee for Aeronautics, Technical Note 2817National Advisory Committee for Aeronautics