Sail CFD - Help interperate my results

[m_liddell]

I've been running CFD analysis on the 29er main and mast using fluent and the results have a weird effect. Those show here are using k-omega sst but I tried a run with k-epsilon (with enhanced wall functions) and the results were almost identical, with the vortex present.

Angle of attack 12 degrees in 11knts apparent wind.

(NB Due to difficulties meshing a small section of the top of the mast and main are not included)





The solution converged in 600 interations using second order discretization where possible.

At first I thought that the results were seriously flawed but the fact I got almost identical results with k-epsilon is interesting.

[jehardiman]

Why shouldn't there be a vortex? As a sailor you should know that if the windward LE tailtale on the main is flying (i.e. in the seperated flow), then you have too much draft in the sail. Put in some outhaul and more mast bend to flatten the draft and try it again.

Additionally, see Hoerner FDL for a discussion on the lift and flow characteristics of very thick foils (i.e. the mast) and you will see some surprizing results (negative lift, seperation at low AOA, etc).

[webbwash]

So why all the games with CFD -- you will never in your lifetime duplicate the conditions you have set herein --

Just go out and sail the bloody thing and enjoy the discovery of what can be done with wind on your sails and water under your hull --

Have fun --
RW

[m_liddell]

I'm a final year mechanical engineering student at southampton university and this is my final year project. Believe me I'd rather be out on the wire sailing my I14

[koen]

This picture shows a similar flowstructure.

http://arnone.de.unifi.it/ICAD/galle...it/sydney3.gif

I always have doubts if RANSE is very usefull for this kind of situations.

[jehardiman]

Originally Posted by koen

I always have doubts if RANSE is very usefull for this kind of situations.

It is always my experience that CFD for hydro/flow at most sailing speeds (say less than 10 knts) is more than useless because it gives a false impression of exactness. When velocities are very large in relation to disturbances then CFD can get you in the ballpark (+/- 20%), but that's the same as a good coefficient analysis. When disturbances are on the order of the velocities (which is the case for most sailing craft) then most CFD numerical answers are meaningless (especially with the invicid, irrotational, isodense assumptions) but can be enlightening about flow patterns.

[tspeer]

What you're seeing is real. There is a large separation bubble on the windward side behind the mast. If you put several telltales in a row in the chordwise direction you can see it. The telltales near the mast will be reversed and the telltales farther back will be streaming. You can see the bubble shrink as you sheet in, and expand as you sheet out.

What shows up as a separation bubble in 2D flow, with a trapped volume of fluid, becomes a vortex when you add some spanwise flow and allow the bubble to "drain" out the tip.

[m_liddell]

Thanks for all your replies guys, it's given me some good stuff to write about in my report on the project.

[Pieetry]

as a 29er sailor i can tell u that we generaly keep the outhall cranked upwind. also the the way the vang is set up (on top of the boom pushing) comdined with the very flexy mast lead to a far more curved mast both for/aft and side/side then your images show (self adjusting rig i belive its called) leading to a flater sail. have a look at some pictures of a well trained boat going upwind and you will see.

[Pieetry]

http://29er.org/img/1/4800,520,800,80.jpg

[Tim B]

It looks to me like there should be a more evident vortex at the top of the sail and the bottom of the sail. If I were you, my first move would be to use more cells in the grid in those areas. Also, check the mast geometry, as solving with the wrong geometry may make the result look worse than it actually is. Further, I agree with Pieetry, the twist on the sail does not look radical enough.

Good Luck,

Tim B.

[yokebutt]

Monsieur Lidell, I,m rather surprised your model doesn't show any tip-vortex at the bottom, especially considering that you seem to have carried the full draft all the way down to the foot. I'll hasten to say that I'm just a bloody-minded sailor that's used to looking at tell-tales, and not an aerodynamicist, any help to elucidate what I seem to be seeing would be very much appreciated.

Yokebutt.

[Skippy]

I second that question (actually 3rd). Why is there no air flowing under the boom?

[m_liddell]

In response to the questions asked:

Why is there no air flowing under the boom?
Not totally sure what you mean - air spilling from HP side to LP side?

With regard to the sail twist/mast bend, the CAD model was made from measurements with the boat just rigged up with no load on the rig. I then deformed the shape to correspond with the 'ideal' shape for 6-8knts in 'High Performance Sailing' by Frank Bethweite. I'd love to have had the resources to make the model from photographs from the masthead and from behind while the boat was sailing.

The only commercial turbulence model in existance that could cope with the low velocity flows/separation around sails is V2F which the university didn't have a licence for and refused to buy one

[Skippy]

Right, HP to LP. I thought there would be a lot of flow & turbulence around the boom, since it's such an abrupt edge at the lower end of the sail.