Keel Foils (optimizing a 70's keel)

[ErikG]

I have an old IOR designed 26 foot sailboat that I race when I have the time.
I'm thinking of optimizing the fin-keel.

The original design have a foil that somewhat looks like Naca 0009 section. At least that's the closest I've found so far. The difference to a 0009 is that it is thicker/blunter at the leading edge. The max width seems to be around 20-25% of the cord. Max thickness is around 150 mm and Cavg is 1000mm.

It does not look (in any way) like a good foil.
What could be a reasonable way to attack the current foil?
The keel is of lead and covered in polyester/mat/roving. This is also the main way of keeping the keel attached to the hull (cheap-ass solution :-)
Additional keelbolts are in place but were not in the original design.

I'm thinking of lengthening the cord just a bit to get the max width a little bit "further aft" and also have the leading edge a bit narrower more like the proper 0009 profile.

What are the drawbacks of the 0009 section compared to f.ex. 63&64 series?. I've bought "Theory of wing sections" but have yet to recieve it.

Any suggestions or ideas?

Erik

[tspeer]

Originally posted by ErikG
...The original design have a foil that somewhat looks like Naca 0009 section. At least that's the closest I've found so far. The difference to a 0009 is that it is thicker/blunter at the leading edge. ...
It does not look (in any way) like a good foil.
What could be a reasonable way to attack the current foil?
...

One way would be to use XFOIL to design a custom section. Take many measurements of your current keel, especially near the leading edge, to create a set of coordinates for the existing shape. Then load this into XFOIL to see what the predicted performance is. The use the inverse design capability (QDES) and geometric design (GDES) menus to modify the shape until you get something that performs well and is close enough to your existing keel.

What are the drawbacks of the 0009 section compared to f.ex. 63&64 series?. ...

Probably none, especially with only 9% thickness. The problem with the 6-series sections for keels is that the drag bucket is narrow and centered on zero lift coefficient (because of symmetry). Except for downwind, the lift coefficient will probably be outside the bucket and you lose any performance gains you might have expected. It's unlikely you would have seen the drag bucket anyway, because it takes a very smooth surface to maintain laminar flow (depending on the Reynolds number) - look at the curves for "Standard Roughness" completely wipe out the laminar flow drag bucket. Finally, the 6-series sections can have a nasty leading edge stall.

The 0009 will probably have a lot of laminar flow at low angles of attack if the surface is smooth and the Reynolds number moderate to low. It will be better behaved at stall - like when coming out of a slow speed tack - and will probably have lower drag at the lift coefficients expected when going to weather.

[ErikG]

You're the man!

The more I look at the different foils I realise that the number of "ready made" symmetric foils seems to be only a few.
The four digit series i.e. 0009, 0012 etc. is it just thickness/cord relationship that are different? Overall shape seems to be pretty identical.

Larsson & Eliasson claims (in a way) in Principles of yacht design, that unless you're very knowledgeble in regards to foils and flow that you shouldn't try your own sections. What's your opinion on this?

Erik

[tspeer]

I don't know how to become knowlegeable on sections except to start playing with them! I suppose it depends on how much time you're willing to invest in the effort. There's a definite learning curve to XFOIL, but I found it rewarding once it all clicked in my head.

For some ready-made symmetrical sections, try the Eppler E520 (15% thick), E521 (13.8%), E297 (11.4%), E836 (12.7%), and E837 (16.1%) from "Airfoil Design and Data", Springer Verlag, 1990.

Also Wortmann FX 71-L-150/20, FX L III-142 K 25, FX L V-152 from Stuttgarter Profilkatlog I, Friedr. Vieweg & Sohn, 1981. There's a second volume of the Stuttgarter Profilkatalog, too, and there may be additional symmetrical sections there.

The Wortmann sections have wind tunnel data for them, while the Eppler sections are theoretical predictions using the Eppler code. XFOIL is more accurate than the Eppler code, so it would probably be worthwhile to use XFOIL as the common basis for comparison.

Of these, I think I like the FX L V-152 (15.2%) the best, as it has a wide flat drag bucket and a gentle stall with a maximum Cl of 1.2. XFOIL would tell you how well it would perform if you were to shrink it down to a thinner section.

[Tim B]

Hi,
I may voice just a small note of caution. The NACA-4 digit series are what are known as turbulent flow sections, that is, they utilise a turbulent boundary layer aft, and a laminar boundary layer forward, as anyone who has run X-foil for viscous cases will know. Running inviscid cases is worse than useless, because we are really interested in good L/D at high angles of attack. The other thing to look out for is cavitation, which I beleive occurs around Cps of -2 (don't quote me on that). However, X-foil is a remarkably useful prediction program. I suggest that you pick up at least version 6.94 (I think it's PC-ONLY, available off the web), which appears to give similar results to CFX with a wing between two symmetry planes. At the moment I'm running Version 5.4 (mainly) as well as 6.94, and the older version on a Strong-ARM risc PC. I'm currently getting about 20 seconds for a 2D viscous case, but 6.94 is a lot faster on a good PC. The other thing I suggest you try is looking through the UIUC aerofoil database. There should be plenty of sections to keep you going there.

NB. I prefer GDES in V5.4 for RISC-OS, but it is quite useable in V6.94.

Good Luck,

Tim Brocklehurst

[SuperPiper]

On a different note:

What would be the effect of adding a flat plate to the bottom of the keel? I do not mean a bulb or added ballast. I suggest a 1/4" thick S.S. plate lagged into the bottom of the foil. The section view of the plate would somehow match the pressure gradient around the keel. The purpose would be to eliminate the losses around the bottom of the keel and provide some "bite" when the boat is heeled. (Let's ignore the additional wetted surface area for now).

[Milton Thrasher]

I am interested in your comments about "adding a flat plate at the bottom of the
keel". I have done that on my Irwin 23 and am experiencing a big improvement
in pointing ability and reduced hobby-horsing. This is design by trial and error
and is costing virtually nothing as I have a hoist and time to make many more
prototypes to find what works best.

Please see my web page at
http://www.angelfire.com/fl4/mft/wing.html
for an explanation and photos of my
plywood wing keel prototype.

I have proved that the worry about running aground is not well founded for a
thin plywood keel. We were able to get off ground by simply heeling the boat
sharply. The thin plate keel apparently cut through the sand with very little
friction. Now I am thinking of using a hard rubber plate in case of runing
aground on rock or hard packed sand.

[SeaDrive]

I have read (I don't remember where) that smoothing the surface is worth a lot, but small changes to the foil section (i.e. templating) is not. It may (repeat MAY) have been in a discussion of Dennis Connor's preparation of a PHRF boat.

[SailDesign]

Originally Posted by seadrive

I have read (I don't remember where) that smoothing the surface is worth a lot, but small changes to the foil section (i.e. templating) is not.

That's what I've always found, but you have to be careful - a smooth POS foil is still a POS foil.....
And the way I heard it was "there is more to be gained from smoothing the foil than from using a better foil that isn't smooth". Of course, this somewhat contradicts what I say above, so I'll shut up now, and leave you in the _more_ than capable hands of Mr. Speer.

Steve

[Stephen Ditmore]

I also would defer to Tom -- but I was in a position to hear reports of 1992 America's Cup research, and the 00-- series sections and similar foils seem to do pretty well at a thickness of 7-10%. I think the most important thing is that you choose some decent foil and do a good job of fairing and wet-sanding your keel.

Best Wishes,
Stephen