Why do we need CFD calculations for planing boats?

High speed planing hulls can be characterized by a flow with a large spray, a dry transom at design speed and a flow separation at the chine with a possible reattachment along the sides. Its characteristic stagnation line is associated with the large pressure gradients. State of the art CFD techniques can compute the performance of planing boats taking all these elements into consideration.

VOF RANS software typically base their recommendation for the time step on the boat’s speed: the higher the boat speed, the smaller the time step should be.

How can CFD calculation benefit from Savitsky prediction?

The Savitsky prediction is directly embedded into the C-Wizard. Hence, the C-Wizard can estimate the position of the fast boat (trim and sink) allowing the CFD simulation to start directly from the dynamic equilibrium position of the planing hull for the given speed. This means we get a mesh that is already aligned with the flow and we do not need to use the acceleration ramp to reach final speed and position. The main benefits are the gain in convergence time, higher numerical stability and accuracy.

How does the automated C-Wizard work?

Referenced case description          

One of the validation studies of the C-Wizard Planing regime approach is based on the experimental results of E. Thornill et al [1]. A set of bare hull resistance tests were performed on a 1/8 scale model of an 11.8 m long planing hull. The model is a generic quasi-prismatic hull which suits Savitsky prediction conditions. Main parameters of the model can be found on the Fig. 1 below:

Fig 1. Parameters of the hull model

Results

The generated automatic mesh is shown here below and contains 1.343.000 cells including the viscous layers. Simulations have been performed on a standard workstation with 9 CPUs reaching the convergence around 1,2 sec of the flow time and 8 hours of simulation time. General representation of the mesh can be found in the Fig.2.

Fig 2. General mesh representation: side and aft view

A fully automated mesh generation procedure was implemented in C-Wizard, based on the refinement dictionary and completed with the input physical condition. The refinement dictionary has been created based on the experience and validation results of the evaluation tests. For the current study the mesh was generated including extra refined zones of the free surface and the aft-zone of the boat.

The result is a mesh of the high quality, especially in terms of orthogonality and viscous layer insertion, preventing the non-physically diffused, streaked mass fraction on the hull surface. At the same time the obtained solution benefits from the high accuracy numerical scheme, aimed to eliminate the high diffusion of the mass fraction from the solver side.

The full study has been performed over a range of speeds between 3 m/s and 7 m/s for the ballast conditions in calm water. Experimental results  provided in [1] allow us to compare the Towing force, running Trim and Sinkage obtained with the FINE™/Marine C-Wizard setup. Results of the comparison can be seen on the Fig.4-6.

Fig 4. Total resistance: FINE™/Marine results  against the experimental data of  E. Thornill et al [1]

Fig. 5: Sinkage: FINE™/Marine results  against the experimental data of  E. Thornill et al [1]

Fig.6: Dynamic trim: FINE™/Marine results  against the experimental data of  E. Thornill et al [1]

Conclusions

FINE™/Marine C-wizard Planing regime simulation shows a good correlation with the experimental data for Total resistance, running Trim angle and CoG sinkage;

The automated procedure from the inputs to the start of the simulation include all the steps from mesh generation to solver parameters preparation, based on the physical conditions: mesh specifics consider the flow around the stern area and body motion parameters are respected by the C-Wizard Planing regime setup;

The Savitsky prediction included into the procedure provides a good estimation of the final position for the planing hull and helps to get a faster converged simulation: it is possible to save up to 50% of the CPU simulation time.

References              

Planing Hull Model Tests for CFD Validation, E. Thornill et al, 6th Canadian Marine Hydromechanics and Structures Conference, 23-26 May, 2001 Vancouver BC