Double keel or single keel ?

Oh for goodness sake - are you trying to tell me that the angle of centreboards to the hull and direction of travel should be determined by the sailed course only?
Thats like saying a pilot should control lift only with the tail plane, and not use the ailerons on the main wings.
And to top it off, by implication you are saying that those boats that are using variable angle centreboards, or trimtabs (thank you Par), are not getting any benefit!

 

No, they are probably not getting any benefit compared to simply sheeting in more and pointing higher. Its still the angle between the keel and the sails no matter what the pushing from a fixed point in a tank tels you .That model is beiing pushed form a fixed point, not by sails.
Boat are boats , planes are planes . They have much in common , but not everything.
Brent

 

Brent, if you dont have any understanding of hydro or aero dynamic forces, and cant put those concepts into a readable form, you dont have to contribute to the discussion to fill in space.
If there is anyone else that has an interest and can contribute something, I look forward to reading it.

 

From an analysis perspective you could do a first order calculation treating the hull as a low aspect foil and the keel as a high aspect foil. I expect this approximation would be good for a long hull and short, high aspect keel. You will find the keel is substantially more efficient (L/D) than the hull simply by the higher aspect. So if the keel is given some toe-in such that the hull tracks true, resulting in no induced drag from the hull, while the keel develops the entire lift to ballance the sail force there should be a substantial gain.

The best L/D for the hull may be around 4 while the keel could easily be 20 or more. So makes much more sense to generate the lift from the keel. You see most modern aircraft have asymetric wings that provide lift when the fuselage is in line with the flow. I expect the majority of the lift comes from the wings not the fuselage.

The same thing could be achieved using two asymetric dagger boards. Also twin fixed keels angled out at say 30 degrees. The latter set up would tend to create a righting force with a downward componment that would increase hull drag. You are also contending with extra drag when running.

I can recall seeing a design with what looked like two large rudders instead of a keel. One well forwadr of centre as well as the stern. I am not sure if both could be rotated.

As I have said on other threads you can use JavaFoil to quickly get a quantitative insight into operation of foils.

Rick W.

 

Twin Keel Geometry

Some good points Rick. I think my line of thought as this stage is more the geometry of the control surfaces, than their lifting co-efficient.
The hull shape I have specified will probably produce very low co-efficient of lift, as it is designed to be a very boxy shape for maximum accomodation and power performance (a la macgregor etc)
From what little I can understand, the geometry of the fins will be more about pure resistance because of the relatively low sailing speed (7 knot average), rather than any induced lift factor.
Like a jet gets more of its lift from the sharp attack angle on takeoff at low speed, (like a kite), then at cruising speeds the co-effient of lift from curved surfaces is enough to hold it up in the air.
The attached picture is how I understand it works - angling the leeward fin to reduce the actual boat vector, while the windard fin ends up being in line with the improved vector, thus not producing drag.
I am tempted to make the fin casings oversize, and put some mechanism inside to adjust the angles while sailing. I suspect that varying sailing conditions (like running and reaching) might make this a usefull (and fun) feature.

 

I have attached some screen dumps from JavaFoil with your drawing in mind.

I am modelling the hull as a NACA0030 foil with AR of 0.2.

I have modelled a single keel as NACA0010 with AR 4.

If you assume the boat is designed with a single keel and design leeway is 5 degrees then the L/D for the hull is 1.9 while it is 18 for the keel. This means the keel will achieve roughly 10 times more lift for its cost in drag as compared with the hull. You would need to compare actual areas to determine the relative contribution.

With an asymetric keel you can get even higher L/D with the same AR. The third set of data is for a NACA3410-43 with AR of 4. It has a maximum L/D of 23 at 0 degrees AoA. So this could be set up to provide the entire lift without any leeway. You would need one on either side and be able to lift the windward one - a bit of a pain when tacking. Angle them outward so they get maximum draft when at design heal.

The point I am making is that it makes sense to generate all the lift from a high aspect foil than relying on any contribution from the hull. Any leeway is costing you in induced drag on the hull.

Also keep in mind that any wetted surface will add drag even if it is not lifting. You can see the Cd for the NACA 0010 at 0 degrees is .0055. Roughly 1/4 of the drag when it is working hard at AoA of 5 degrees. A big penalty to be dragging around all the time when not actually working.

One of the reasons I like small, fat, bottom hung rudders is that they will generate plenty of lift without the penalty of dragging a much larger blade through the water as in large, thin transom mounted rudders.

Rick W.

 

I couldnt follow much of what you were trying to get at. I think I already said that the hull for my purposes can be largely ignored as far as lift goes, and I dont have to be convinced that the foils will provide much more significant lift.
The last sentence puzzled me even more.
Small fat rudders dont generate much lift, not "plenty of lift" as you wrote!
I know the bigger the appendage, the more drag, but if you dont have a decent length to the foil, you dont get much lift because there isnt much leading edge, and also the laminar flow just drops off the ends of a shorter rudder as lost energy.
Thats why slow moving gliders have longer wings than fast moving powered planes - they need maximum lift at slower speeds, while powered planes can reduce wing size, and thus drag, because the high speed fluid flow can produce plenty of lift across a shorter leading edge.
At slow sailing speeds, short rudders or foils just wouldnt create any viable lift. And you do want lift in the rudders - its more efficient to produce steering pressure from laminar flow than from mere angle of attack, where often caviation occurs with resultant heavy drag penalties.
Thats why you never see anything but long foil rudders on sailing boats. So I dont know what you are getting at with the "small fat" rudder concept.

 

If you have leeway then the hull is producing lift and there will be induced drag as a result.

I think you are confusing length with depth. In my reference frame, length is measured along the logitudinal axis of the hull. Depth is measured below the hull. So a short rudder has a short chord length. The higher the aspect foil (deeper relative to length) the more efficient.

Once you get Re# over about 50,000 you get reasonable foil performance. A 150mm long rudder at 6kts is well into effective regime. I also think you are confusing streamline flow and laminar flow. A fat section rudder like NACA0020 will carry streamline flow up to about 20 degrees. I have attached JavaFoil polar curve for a rudder with AR of 4. You will see that the Cl peaks just under 1.1. Using a NACA0010 section of AR of 2 the Cl peaks at 0.4. So for a draft constrained rudder the fatter rudder only needs to be about 1/3 the area of a thin section rudder like NACA0010 to generate the same lift. The in-line Cd for the fatter section is less than twice the thinner section so there is an overall reduction in drag by using the fatter section. I have never taken the time to determine what the best rudder section is for a given condition but NACA0020 is not far off. It might be as high as NACA0025.

If you are thinking about toe-in or asymetric keels a similar sort of analysis is required to get the optimum result. The aim should be to eliminate leeway so there is no induced drag on the hull. I expect you need a symetrical steerable keel or two drop keels with asymetric section to get the optimum.

Rick W.

 

rwatson
the type of rudder that Rick refers to is not 'short'. It is small but still high aspect, and fat when you look at it from above. It produces a lot of lift and must be protected from cavitation/ventilation.

 

Just been reading the thread properly. Rwatson....you were a bit harsh on Brent don't you think? I wouldn't like it if someone brushed me off like that...and I'm sure I have said things that don't make sense! Be nice

 

geeez I sound like a woman

 

Its not my fault Richard - Honest Darling !!!
Its only because people keep trying to confuse me - like
I know what Laminar Flow is - its "streamline flow" when people are trying to confuse me!
"Laminar flow, sometimes known as streamline flow, occurs when a fluid flows in parallel layers, with no disruption between the layers" according to google definitions.
And "small, fat" means I am "confusing length with depth"
In my school a rudder would be "small" as in short, not small as in length, so of course I assumed "fat" was the other axis.
But now thats all sorted out, Rick really wasnt incorrect - He just wasnt talking about my area of interest - which is twin keel geometry, not foil sections - which was another attempt to confuse me!
But for all that, I am going to have to think about foil sections as well I guess - but that might be too confusing. The suggestion of having asymetric sections will probably result in more confusion.
I just need the web address for the "Twin Fin Geometry Corporation", and then I can go and buy a couple of kilos of "Twin Fin Configurator" to splash in the water around the boat. is that too much to ask??????

 

To ease the confusion I should be refering to attached flow rather than streamline flow.

Foils of the scale used on a full size sailing boats operate in the turbulent boundary layer regime with "attached" flow over most of the surface. The flow over the majority of the surface is NOT laminar as there is a turbulant boundary layer but the flow remains "attached" to the surface. At high angles of attack the flow becomes "detached" and the foil stalls.

This article might help the understanding:
http://www.sportpilot.org/magazine/feature/2006 - 07 July - In Pursuit of Airflow.pdf
It covers streamlines, laminar flow, turbulent flow, attached flow and flow separation. You can see how turbulent flow is induced to improve performance of the foil in certain circumstances.

I have attached an output screen from JavaFoil that shows the flow transition points and the flow separation points for a NACA0020 at 5 degrees AoA and Re# of 300000. You can see the transition point on the upper surface occurs at 20% along the surface. Flow separation is only occuring near the trailing edge.

Keels, rudders, sails, propellers and wings are all foils and JavaFoil provides a wealth of insight into their operation and good numbers for design purposes. With a little effort it will be your Twin Fin Geometry Corporation and provide the necessary insight.

Rick W.

 

Not quite. The total side force, of course, has nothing to do with the keel design or toe-in. The total side force from the keels and hull will be a function of the sail trim. The leeway angle will adust to match the force from the sails, so you're right about that.

Toe in has the effect of shifting the load from the windward keel to the leeward keel. At a given leeway angle, the leeward keel will have a greater angle of attack than the windward keel.

Since the windward keel operates much closer to the surface, the induced drag due to lift is significantly higher for the windward keel. Plus, the normal force from the windward keel is not oriented to oppose leeway nearly as well as the leeward keel. So you want the leeward keel doing all the heavy lifting and the windward keel along for the ride. Ideally, the zero lift line of the windward keel would be aligned with the incident flow at the leeway angle.

Then when you tack, the load shifts onto the other keel. So, although the wetted area is not reduced, the effect on induced drag is like raising and lowering bilge boards on every tack.

 

case for angled fins

Thanks Tom, that confirms my understanding, and I have modified my working diagram to clarify these concepts in my own mind.

I am leaning towards Ricks hint at using asymetric foil sections, and angled boards as Brent was talking about. (In fact the angled boards could be incorporated in the angle of the cabin sides, making it a nice neat building solution).
If a chord of 150mm in a fin can produce lift at 6 knots using Ricks example, I would expect a slight attack angle would help keep the flow attached, and also provide directional assistance, thereby reducing the leeway angle.

And I think your sentence "the zero lift line of the windward keel would be aligned with the incident flow at the leeway angle" means that one should try to keep the angle of the windward fin as close to the boats true heading (not steered heading) as possible.

That is probably the answer to the mystery of the optional toe-in angle for a boat, and it sounds like it would vary a fair bit from one boat to another depending on hull/sail/size factors.
If I dont master the math, I might just incorporate adjustable foils and experiment in real life conditions.