Helicopter Landing on a Moving Frigate
The goal of this study is to demonstrate the feasibility of using the Vorcat code for accurate and real time simulation of ship board helicopter landing flows.
In order to achieve said goal, we utilized the Vorcat software to:
Develop and solve a realistic numerical model for landing scenarios.
Model & simulate physical experiments carried out at Old Dominion University (ODU) and compare CFD results to the experimental data.
Analyze results and devise a predictive tool based on accuracy and speed.
Frigate wind tunnel model and grid:
Sample Simulations Results :
(Complete data sets are available upon request)
Frigate – no rotor
time step=877, T=2.1, 8.6M vortons (50% shown), 680K filaments
Top View
Side View
Velocity fields: computed (right) vs. experimental (left)
y/d=0.4
y/d=-0.3
Frigate – with rotor
B.1. rotor (5/5) located at y/d=0.5
Time step=900, T=1.36, 7.0M vortons (50% shown),
1.17M filaments
Top View
Back Side View
Rotor Only vVorticity elements
Top View
Side View
Side View
Velocity fields: computed (right) vs. experimental (left)
y/d=0.4
y/d=0.2
y/d=-0.5
Rotor model resolution parameters (15 points – left vs. 5 points – right): y/d=0.4
Tracer Particles
No Rotor
Rotor placed on side
No Rotor (left) vs. Rotor on Side (right)
Rotor placed in the center
Experimental data vs. CFD results: General Comments
Experimental results are based on relatively long time averages compared to CFD stats (roughly X100).
Experimental measurements are conducted on 1,152 points per cross section compared to about 110,000 computed points per cross section.
The experiment included a physical rotor and its shaft that were not accounted for by the CFD model.
Experimental data vs. CFD results: Comparisons
The CFD results closely matched the experimental data.
We found excellent agreement for the no-rotor case. The differences between Vorcat averages and experimental data are on the order of the differences between experimental data sets measured at symmetrical locations.
In the frigate+rotor case, the CFD averages are reasonably close to the experimental data – both qualitatively and quantitatively - and they get closer as more computed data sets are made available for averaging.
For the application considered here - the dynamic simulator - the data sets provided by Vorcat are physically consistent and represent realistic time-dependent large-eddy turbulent velocity fields. In particular, the numerical results depict the flow asymmetries due to:
Advancing/retreating blades (rotor at y/d=0.5 vs. y/d=-0.5)
Ground and corner effects (rotor at different heights and distances from a fully inviscid geometry; results not shown here).
Interacting rotor-wake with ship wake (comparison between rotors above inviscid and viscous frigate - results not shown here - and comparisons with no-rotor case).
Time-dependent velocity fluctuations associated with turbulence (no-rotor case on a symmetrical geometry).
Conclusions
The Vorcat software has been applied successfully to the simulation of rotor-wake coupling in the complex turbulent flow produced by a model rotorcraft landing on a moving ship.
Results suggest that the velocity field at the rotor plane can be modeled, accurately and fast so that it can be utilized in a dynamic simulator.
Adaptation of the present methodology to a variety of realistic scenarios including a helicopter body and other boundaries/obstacles is straightforward and practical.