Teardrop End Plate On NACA Flat Bottomed Fin Keel ??

Is it worth adding a teardrop shaped end plate to the bottom of a NACA flat bottomed fin keel on our O'Day 37 center cockpit sloop designed by C.R. Hunt? (See attached figure for end plate shape)

 

Why do you want to add the end plate? What do you hope to accomplish? These are real easy to screw up and rarely worth the effort to get an extra little performance

 

I've read articles ( i.e. www.boatus.com/goodoldboat/keeldesign.htm ) which state "a large part of the resistance of a keel is created by the vortices, similar to miniature whirlpools that form when the water flows across the bottom of the keel from the high-pressure (leeward) side to the low-pressure (windward) side." They claim that "70's design" keels, like the O'Day 37, suffers from such high keel resistance. This all makes sense from an engineering perspective.

The question is, if I design a plate according to their recommendations, and bond it to the flat bottom of our fin keel, will it result in significantly reduced keel vortices, and therefore increase our max attainable (vs theoretical) hull speed by say 1/2 - 1 knot?

We do a lot of cruising and racing, where the extra 1/2 - 1 knot would be a big help.

Is "the gain" worth "the pain" based on other's experience in this upgrade?

 

It is true, that "...a large part part of the resistance of a keel is created by the vortices...", but remember that the resistance of the keel is just a small part of the total resistance of the entire hull.

If you decide to put an end plate on your keel, the resistance of the keel will decrease, but you won't gain anywhere near 1 knot - maybe not even 1/2 knot.

Besides less resistance, the keel will most likely develop a tiny bit more lift (because of the increased effective aspect ratio), so youl'll be able to point a tiny bit higher, but not much. The endplate will probably lower the center of gravity (unless it's made of wood), so the stability will also be a tiny bit better.

So, is the gain worth the pain? You'll point a tiny bit higher and sail a tiny bit faster at a tiny bit smaller heel angle. I'd say this depends on your personal threshold of pain...

 

As sorenfdk says, don't expect any shocking increase in performance. I took the trouble to see if I could find a planform and what I found shows a highly swept, moterate tapered, low aspect foil. And she has a huge skeg/rudder (any weather helm?). I doubt that adding an endplate to this keel would be noticable except in decreasing downwind speed. I can't tell, but is the keel tip squared off? There is some work with the tip cap that can improve L/D ratios, and some small flow fences may help with VMG down in the hundreths of knots.

To see any major performance gains would most likely require an entire new keel design/ballast change/accomidation rip-out/rudder change which would really screw your rating or gain a hefty penality for racing and would most likely increase your draft. PHRF rating now is almost even with a J-24 or Ericson 30 or 32 so we're not talking real speed demon here.

I personally would stick with the standard go-fast stuff here; good smooth and fair bottom, low-stretch lines, new sails, and well tuned rig.

 

Yeah but:

When the boat is heeled, will the endplate not act to reduce leeway? So, the boat may point higher AND make less leeway? That could add 1/2 - 1 knot VMG!

 

A more efficient keel should reduce leeway and thereby reduce the resistance. However, I don't think you'll gain as much as 1/2 - 1 knot, but only a real-life, full scale test can show.

 

Not that simple. All symetric foils increase in drag as you increase lift for a given aspect ratio. All end plates increase increase drag as Angle of attack (AOA) increases and effectively also increase aspect ratio....to a limited AOA. Aspect ratio increases lift per degree of AOA and decreases stall angle. In the final analysis maximum Cl may have more to do with section fineness (i.e flow attachment or stall angle) vs sail power than most people expect. Leeway is a function of the L/D curve and side/driving force curve. Therefor as you increase lift and make less leeway, you generate more drag and go slower provided you don't stall either the sail or the keel (i.e. the old plank on edge low aspect sail and foil high leeway slow slog to weather vs the modern hull high aspect sail and foil low leeway foot off and go fast, both having the same VMG to weather). A very fine trade off that cannot be well determined. Both CFD and small scale models are not up to this trade off (i.e. both have failed to predict correctly on occasion).

 

Try half a plate to see what the benefits are ??

Interestingly the Boat US article talked about testing 1/2 a plate (i.e. the Port Side only) to run a quick comparison.

Does this make sense?

Captain Rich

 

Worth a try?

Your points all seem like it's worth a try if a) it doesn't take much effort and b) it's reversible.

I figure that making a wooden male mold, vacuum bagging it in fiberglass and bolting it to the bottom of the flat fin keel (with a gasket in between), shouldn't be that "big a deal".

I guess my 30 years as in R&D is showing.

Captain Rich

 

Gain Not Worth The Pain ??

I've included a photo of our O'Day shortly after we took her down to the gel coat, put three layers of barrier epoxy on, and then three layers of ablative copolymer bottom paint, making her hull a lot smoother.

We have new sails, roller furling, and low stretch lines, which made a world of difference in sail handling.

The inner forestay was moved aft to provide a cutter rig configuration. This allowed us to point 10o closer to the wind, and allowed us to reef down and still keep the helm close to neutral.

She still has a lot of weather helm, as you mentioned, keeping the autopilot working in gusty conditions.

The only thing left that I can think of is a) to move all halyards and topping lifts to the leading edge of the mast and b) reduce keel and rudder drag, if possible.

Surprisingly in our first ocean race this year, we won second in class during the Bermuda Ocean Race, finishing just 8 minutes behind the leader after 7 days of ocean sailing.

Your help is much appreciated, since now I'm into the fine tuning.

Captain Rich

 

Only you are the boss of you...

Hello...

If you want to do it - just do it...

If you think it might work - it just might work...

If you have the time to sacrifice - it don't mean nothing...

If you fail miserably - don't do it again..

If you can afford it - think of all the steaks you could buy instead...

If you like to tinker - blame the Meccano company...

Nice boat - good resale value as Hunt designed it...

SH.

 

Positive Attitude !!



What a great attitude!

You made my day!

Captain Rich

 

It would not be worth it. from looking at your sketch and keel planform, upwind gain would be not noticeable or maybe even hurt. Downwind would definitely be hurt. You're better off with good sails, clean bottom, tune rig, etc.

 

Sean's got the right idea, ....if you want to do it, just do it. But make it fixable if it doesn't work.

If you want to size the end plate, the transverse distance it sticks out should be slightly (20%?) larger that the calculated boundary layer at maximum speed at that point all along the foil. I would not have the plate extend much forward of the quarter cord point myself to avoid AOA issues. Just an start the plate with an elliptical expansion to the calculated boundary layer thickness. Thickness of the plate should be 10-20% of the span with slightly broken square edges. Trailing edge of the plate should extend 1 span length aft of the trailing edge of the keel. Lacking detailed data, trailing edge shape is personal preference, but I would avoid having it cut square.

Remember, it is well pointed out that it is more efficient to extend the foil than it is to add end plates (Hoerner, Fluid Dynamic Lift, Chpt. 3). The original mystique of end plates was formed by a rule limitation and a bit of legerdemain. You would do well to review your design limitations to assure yourself that this in fact is the only way to improve performance.