Part_IICFD_Applications

1
Part_II-gtCFD_Applications

• Required skills
• Appreciation of flows involved with sailing
  yachts
• Appreciation of numerical methods
• Knowledge of classical fluid dynamics

2
Summary

• Use aero results
• Cunning use of experiment and CFD
• BEM vs BEM with Kutta condition vs Inviscid FVM
  vs Laminar Viscous FVM
• Sails
• Keels and rudders
• Hulls
• The complete package!

3
Use aero results! - I

• Aerodynamics has been computationally based from
  the start
• Aerodynamics has far more resources than sailing
  yachts
• Similarities are closer than you think
• But realise that compressibility results in
  different numerical problems
• See Ferziger and Peric (2002), pp 309- 328

4
Use aero results! - II

• Eg the DPW II had this at the wing body interface

5
Use aero results! - III

• Eg I have seen this on bulb/wing interface

6
Use aero results! - IV

• Eg which turns into this with too much sweep

7
Cunning use of detailed experiments

• More than just force measurements, eg from DPW II

8
BEM vs BEM with Kutta condition vs Inviscid (?)
FVM vs Laminar Viscous FVM

• Test case is a NACA 4412 at 20 degrees angle of
  attack in 10 m/s
• Stalled, unsteady flow

9
BEM

• Force 0.0!
• Not a very interesting solution

10
BEM with Kutta condition

• Lift exists
• Stall by empirical means
• 2D drag by emperical means
• 3D drag by analytical means

11
FVM, laminar viscous

• Lift, drag and stall all exist
• Flow here is fundamentally unsteady

12
BEM vs BEM vs FVM
13
FVM, laminar viscous
14
FVM, inviscid?
15
Resources

• BEM
• 0-100 software
• 2.8 GHz P4 for lt1 second
• 20 minutes work
• FVM
• 1000s software with annual license fees
• 4 x 3 GHz P4 cluster 4 hours
• 2-3 hours pre, 1 hour post

16
Sails - I

• Upwind sails through BEM (attached), in 2D

17
Sails - II

• Upwind sails extended to 3D with vortex lattice
  method, really just predicting the tip vortex
  strength

18
Sails - III

• Vortex lattice codes virtually free

19
Sails - IV

• Downwind or cross wind sails through FVM
  (detached), extension to 3D restricted by
  computing power

20
Keels and rudders

• Broad results from BEM and vortex lattice, eg AR
  investigation
• Detail requires shift to FVM (look to aero), eg
  bulb/wing interaction, section shape

21
Hulls

• BEM very good
• Needs lift effects for upwind wave resistance
• Seakeeping predictions possible
• FVM just coming of age

22
The complete package!-I

• Foil can have a free surface (using FVM with
  interface tracking/capturing)

23
The complete package!-II

• Free surface can be pushed to breaking wave
  (using FVM with interface capturing)

24
The complete package!-III

• Free surface can be pushed to breaking wave
  (using FVM with interface capturing)

25
The complete package!-IV

• Free surface can be coupled with motion (using
  FVM with interface capturing)

26
The complete package!-V

• Interface capturing FVM models the air as well as
  the water
• It might be possible to model aerodynamics and
  hydrodynamics in one!
• Similar problem to developing field of wave
  prediction through viscous wind action
• Will require resolution of variety of time and
  space scales, maybe adaptive griding will be
  required
• ?Realm of research for some time