Rudder and daggerboard profiles.

Hi,

Actually designing appendages for an international 14 dinghy, i wonder what kind of profiles i could use out of the classic 0012. Maybe some kind of laminar section. I heard about "DAG" profiles designed specially for sailboats but could find no documentation on those.
Could anyone help?

[tspeer]

The best approach is to download XFOIL (http://raphael.mit.edu/xfoil/) and design a section for your specific requirements. The NACA 6-series laminar flow sections are not especially well suited for sailboats because they are prone to leading edge stall at low Reynolds numbers. The same leading edge separation problem would lead to ventilation when used as a rudder at high speeds. But you could use XFOIL to tailor the leading edge profile and improve them considerably.

Other sections to consider would be the Eppler E520, a 15% thick symmetrical section originally designed for sailplane vertical tails; and the E836, E837, and E838 series, which were specifically designed as hydrofoils and range in thickness from 12.7% to 18.4%. Personally, I think Eppler placed too much emphasis on cavitation for these hydrofoil sections, and I suspect they will also have difficulties at low speeds. But they could be good starting points for a new design. There are also some Wortmann sections that would be good candidates.

[DavidG]

Four digit NACA sections in my experiece are pretty good for sailing dinghies. Bucket sections such as 63 or 64 series may look attractive, because of the reduced drag in the bucket, however Marchaj identifies the angles of attack that dinghies operate upwind (Int Canoe) which is on the borders of the bucket, offwind the board is up (or partly up) so the reduced drag is less relevant.

I tried discussing sections with Bethwaite, but he couldn't support his sections with hard data.

I would give more consideration to using a gybing board, the effect of this is that you effectively sail 2.5 degrees freer (therefore faster) than those around you, for the same pointing!

All this is meaningless, if you mangle the LE or TE. THe board is also useful to stand on.

DG ex GBR1301,1242,1208

[tspeer]

Originally posted by DavidG
Four digit NACA sections in my experiece are pretty good for sailing dinghies.

Yes, they are. I was surprised to see that at low Reynolds numbers and low angles of attack, the NACA 0012 (basis for all the NACA 4-digit sections) has laminar flow over 70% of the chord! The transition point moves forward smoothly with angle of attack, which is exactly what you want for a robust low-Reynolds number section. It goes a long way toward expaining the well-deserved repuation of the NACA 0012 as a good all-round section.

Bucket sections such as 63 or 64 series may look attractive, because of the reduced drag in the bucket, however Marchaj identifies the angles of attack that dinghies operate upwind (Int Canoe) which is on the borders of the bucket, offwind the board is up (or partly up) so the reduced drag is less relevant.

I agree that the NACA 6-series sections are not a good choice for sailboats. Many of their vices have been associated with laminar flow sections in general, but that need not be the case. However, modern tools, like XFOIL, allow you to make the bucket much wider so that it does encompass the normal range of angles of attack. The 6-series used a constant roof-top pressure distribution right to the leading edge, and a straight-line pressure recovery to the trailing edge at their design condition. So they were pretty much single point-design solutions.

You can make a more robust section by designing each segment for a successively higher angle of attack as you get to the leading edge. This effectively rounds off the pressure distribution compared to the NACA 6-series. It widens the bucket and makes the section much more robust at higher angles of attack, eliminating the leading edge separation of the 6-series sections. This addresses the nasty ventilation characteristics of a 6-series rudder at high speeds, and makes a 6-series board easier to accelerate out of a tack.

You can also make a more rounded transition between the rooftop and the recovery region, and make the recovery itself somewhat concave for better performance. The rounded transition promotes a short laminar separation bubble to control boundary layer transition from laminar to turbulent. Making the pressure recovery somewhat concave allows a greater amount of recovery in a shorter distance. So you can extend the laminar roof-top for a lower minimum drag. Or you can raise the rooftop for the same length of laminar flow, which increases the design angle of attack and makes the bucket wider. But don't make it too concave, or you will lose the progressive trailing edge stall.

Past investigators have tended to use the NACA sections without these adaptations to make them suitable for use in boards and rudders. Or they compared ad hoc changes (sharper or blunter) to leading edge shape without calculating what the flow would really do. As a result, the whole concept of low-drag laminar sections has fallen into disrepute among sailors.

If you don't want to design your own sections, the Eppler and Wortmann sections would be better choices than the NACA 6-series. There are a number of databases where you can get the coordinates for more modern sections. Two good ones are
http://www.aae.uiuc.edu/m-selig/ads/coord_database.html
http://www.nasg.com/afdb/index-e.phtml
If polar data are not available, you can easily generate them using XFOIL.


I tried discussing sections with Bethwaite, but he couldn't support his sections with hard data.


He's definitely an experimentalist, not a theoretician. Plus, the main airfoil design tool available when he was doing his research was the Eppler code, and it's very hard to use as well as not being able to handle things like Bethwaites wingmast sections. However, his data are not to be sneezed at - he's done a great job of basing yacht design on rational observation rather than "just-so-stories".

Bethwaite has sent me some tracings of Taser mast sections for me to compare with tear-drop wingmast sections. Unfortunately the tools I have now can't handle the backward-facing steps between mast and sail. So I need to work on getting a Navier-Stokes code installed so I can generate some CFD data to match his experiments.


I would give more consideration to using a gybing board, the effect of this is that you effectively sail 2.5 degrees freer (therefore faster) than those around you, for the same pointing!
...

It seems to me that a jibing board will operate at the same angle of attack as a conventional board. The reason is that the leeway will increase until the board produces enough lift to match the loads applied by the rig, and the contribution of the hull at low leeway angles is probably small compared to the board and rudder. The board is going through the water with essentially the same orientation whether it is jibing or not, because it still needs the same angle of attack to produce a similar force.

So what you're really doing is rotating the hull relative to the board. This has implications above and below the waterline.

If the hull is generating a significant amount of lift, this load will be shifted to the board, which should be able to carry it with less drag than the hull, due to the much greater span of the board. However, if all the hull lift is eliminated, this reduces the effective span of the board to being the same as its physical span. So there's a benefit to having the hull carry some of the load, as long as it doesn't cause flow separation or shift the lateral balance and increase the rudder trim drag. From a lift-induced drag point of view, the optimum hull load would result in the wake coming off the hull being deflected sideways at the same angle as the wake coming off the board. You want the whole vortex wake shed into the water to behave as if it were a rigid sheet along the whole span from waterline to board tip.

Obviously, it's hard to determine just what amount of lift this is without a very sophisticated calculation or test program that takes into account the board, hull, and free surface.

I suspect the greater effect of a jibing board comes from the rig rather than hydrodynamic effects. Rotating the hull makes it act like a ballestron boom. This opens up the angle between foretriangle and the apparent wind, effectively moving the forestay to leeward. The main should not be affected as much because it can be sheeted to the same angle of attack in either case.

[JWC]

Why not make a daggerboard that has a compound foil?

What I'm thinking about is a NACA 0010 or 0012 for the upper 2/3 of the board and a 63-010 or 63-012 for the lower 1/3. This way you get mostly the characteristics of the 00 series when the board is fully down sailing upwind, and the lower drag 63 series section when the board is raised sailing offwind.

Perhaps there is not much point to going with the 63 section for the lower 1/3 since the area is reduced so much downwind that it doesn't matter much?

I have an NS14 which is a box rule boat similar to an I14, but with a lot less sail area. I'm planning to make a new board soon and am searching for ideas.

Thanks for the info on the earlier postings. Lots to consider.

:confused:

[ggggGuest]

0063 is pretty conventional for high performance dinghy sections. I understand that increasing the chord has the effect widening (and shallowing) the drag bucket, the twelve foot skiffs for instance carry relatively fat foils. Better a fatter foil operating in the bucket than a thinner one operating out.

Rudders are another game entirely and need to be very good at high angles of attack with excellent stall properties. I don't know of a conventional section.

[Tim B]

JWC,

A very good idea indeed, the problem is, though, that most of us (LARK Sailors) sail downwind with no board at all, just relying on the hull area if we possibly can. However, your idea may come into it's own using a dagger-board, where it is very difficult to remove enough board without fouling the vang. Remember, though, that as a centerboard is lifted, it rotates, and therefore, the effective chord (parallel to the flow) increases. Therefore, if you sail with a 12% foil fully down, you sail with a 6% section when the board is raised 60 degrees. Tspeer mentioned X-Foil, it is one of the best freely available codes, it does not consider cavitation, but it is pretty good for the rest of the time.

Cheers,

Tim B.

[SailDesign]

Rudder foils are strange - they do not actually operate at high angles of attack in lighter, easier-to-turn boats, as the stern starts swinging quickly, altering the vector to a much smaller angle than it would at first appear. The rudder on American Express (Steve Black's tri) was a 10" chord 0020 foil, and worked extremely well. When you are likely to be moving fast, reduced area is the key. you don't need area, but you do need strength, hence fatter foils with smaller chords. This is also good for spade rudders that are simpler to build if the stock can be enclosed within the blades at least at the root.
Steve

[Guest]

I have a 14' dinghy designed by John Spencer which is similar in many ways to an NS 14. I have built a rudder using a foil designed by Martin Hepperle. It seemed to me that the reynolds numbers would be similar to racing type model aircraft wings, so I decided to give it a go. I can't really be sure yet because of limited testing against other boats but it hasn't slowed me down any.
Does anyone have an opinion on this reynold number similarity. Also another possible foil series worth looking at are those designed by Harry Riblett for light aircraft. They specifically address some of the the leading edge problems of NACA sections and are optimised for RN 1000000.
Both are laminar flow sections and Martin Hepperles web site is at: http://www.mh-aerotools.de/airfoils/index.htm

[nico]

Xfoil - Ncrit

What ncrit should be used for analysis of underwater sections?

Nico

[tspeer]

I've heard ncrit = 1 because transition happens sooner in water than air - all those microscopic critters, bubbles, and particles. I haven't made any runs to compare ncrit = 1 with the default (ncrit = 9, IIRC), though. Look through some of Drela's posts in the XFOIL mailing list archives.

[Matt Lingley]

I'm playing with a NACA 63-009 section for rudders on a fast Cat, is this a good idea? I've been told it may be prone to stalling?

[Andy P]

I make lots of foils for many classes, and normally use the NACA 00 series, but use a NACA 63 variant for faster boats that make less leeway eg int Moth, Cherub.
The disadvantage of the 63 is after a bad tack, when it takes a while to get going again.
I sometimes also use the 63 section for rudders - they appear to ventilate more easliy than the 00, but by quickly moving the tiller and unloading the foil, the flow reattaches much faster than for 00, where you need to slow down as well ( if you can before spinning out)

Gybing boards:

These work best in flatter bottom hulls without pointy stems. ie they don't work well in boats that use the narrow bow sections to help with lift.
So not used in cats, moths, UK National 12, Merlin rocket, but work OK in 5o5, Fireball.

[Unregistered]

hi;
my name is emad and i am m.s in naval architect and work in arvandan ship building co. in iran.
i want naca series for rudders , if you have please give me copy.
thank you,
best regards.
bye.

[Unregistered]

hi;
my name is emad and i am m.s in naval architect and work in arvandan ship building co. in iran.
i want naca series for rudders , if you have please give me copy.
thank you,
best regards.
bye.
my email address is :"emad_rabeaee@yahoo.com"

[nico]

Have a look at http://www.aae.uiuc.edu/m-selig/ads/coord_database.html

some 4digit and 6 series are there

Nico

[tspeer]

...i want naca series for rudders , if you have please give me copy....

http://naca.larc.nasa.gov/reports/1945/naca-report-824/

[JWC]

I am presently fairing a new NACA0010 daggerboard and I am curious about the trailing edge. I haven't been able to find much info on this topic, but most boards and rudders I have seen are squared off at the trailing edge. I assume this is more for practical purposes considering the lack of strength for a fine edge. I'd appreciate any insight on how the trailing edge affects drag.

Thanks,
John

[SailDesign]

"as skinny as possible, but as fat as it needs to be" seem to be the words of wisdom concerning foils.
However, If it is a high-speed boat, consider knocking the "square" edge into a 45-degree angle to stop the little trailing vortices from humming as you sail. They make a terrible noise sometimes, and the energy to make that is - you guessed it! - drag.
Steve

[Unregistered]

Fastacraft manufacture two sizes of moulded dinghy foils, one is NACA0012 and the other NACA0011. see www.fastacraft.com

[kenwstr]

Hi

I have a small slow sailing tender, I did say slow (savage Sprat).

It has what looks like a MJ rudder fitted which is totally ineffective
with a somewhat sharp LE.

I am making a somewhat larger deeper rudder and was considering
using the Eppler 472 section. This is a low RN aeorbatic section
large LE radius with max thickness at 18% of chord from the LE.
I am slimming the section down so the rudder will be 20mm thick
so it still fits existing fittings.
I chose this section because the boat has a strong tendancy to swing
upwind in a gybe, or gust even though in steady wind, the helm
is slightly to the lee. I am hopeing to prevent separation that is currently making the boat uncontrolable.

Any Thoughts?

Regards,
Ken

[SailDesign]

Ken,
It sounds as though the problem is more with the rig than the rudder :)
An Eppler section sounds like overkill for a boat like that - most are adequately controlled by a metal plate. Try using a good ol' NACA 00-series, maybe 12-15% max.
Steve

[ken wstr]

Ordinarily I would have thought so too.

However I am a very experienced sailer and have never come across as severe a problem as this. I have tried to balance the rig and it is just not possible.

I can rake the mast forward to the point of definite lee helm, 10 deg or more on the tiller in steady wind. However a gust will produce severe weather helm and in some cased completely overpower the rudder. I suspect the hull shape is the basic cause so my only option is to compensate with a more effective rudder.

I was given this boat, and the rudder is obviously not designed for it and is way too small. I scaled a new rudder to the same proportion of the centre board as on a Europe class. Even so, I think these helm variations combined with the boats low speed might be too severe which is the reason for considering the E472 section.

There are so many problems with this boat, I want to get rid of it but feel it would be irresponsible in its current state. It's a real handful to keep control.

The other reason for suggesting the E472 is that I am very impressed with it though I have lost access to the online polars. If I remember correctly, the Cl/Cd and Cl crit are quite impressive and would allow smaller rudder and boards to be used on well designed boats. Alternatively a thinner version might be a better optimisation between profile and induced drag.

D = 1/2 p V^2 S Cd is the general form for gas dynamics, I read the 1/2 is used because of the compressibility of gas. As water is incompressible, should this value = 1 in hydrodynamics?

Regards,
Ken

[SailDesign]

Ken,
Good luck with this one - sounds like you're going to need it.
Two things -
1, Keep the 1/2 in the formula. It is there by definition. :)
2. Remember that we are not, in aerodynamicist's terms, speaking of particularly low Reynolds' Numbers here. What may be great on a slow plane is not goingot fare so well on a slow boat, given the relative "nu" values. (Yeah, I know there's a way to produce the little widgets, but I can never remember it)
Steve

[DavidG]

Having a new centreboard and looking for a good section (National 12 slightly slower than Merlin Rocket!)

My constraints are, centreboard case 30mm thick,
Maximum Chord 360mm
Mean Chord 280mm
Span 1070mm
Maximum RN 6.6*10^5

Does anyone have any thought of what angle of attack sailing dinghies really are operating at upwind, Marchaj quoted @ 7 degrees for International Canoes, which is out on the boundaries of most bucket sections?

Is it realistic to think that a foilmaker can handcraft an Eppler or Wortman section well enough (sorry Andy P I'm sure you can) to make the effort worthwhile?

What happens when my crew treads on the trailing edge of a laminar foil?

If the answers to these questions are what I think they will be, I suspect that a trusty NACA 0008 fattening out to a NACA 0011 will be the reliable, but somewhat unimaginative solution.

Any thoughts welcomed!

David

[brian eiland]

Tha latest issue of Profession BoatBuilder had a very revealing photo on the cover. (#90 Aug/Sept 2004). Its a picture of the stern of the big cat Club Med flying a hull. But most amazing is the rudder profile. http://www.boatdesign.net/gallery/showphoto.php/photo/2170

I have a very difficult time imagining that this rudder profle would not be SOOO sensitive to stalling, particularly on a hi speed run downwind and down the face of those waves in the southern ocean.

And it doesn't leave a lot of margin for error in the balancing of the rig over the underwater foils....seems as though it could develop a nasty helm real easy.

And to go all the way around the world, in a VERY fast mode, with all of the obstacles out there in our oceans, with only two slivers of steering devices below you like this, Balls!

[tspeer]

Having a new centreboard and looking for a good section ...Maximum RN 6.6*10^5
...Marchaj quoted @ 7 degrees for International Canoes, which is out on the boundaries of most bucket sections?

Yes, that would be well outside the bucket of a NACA 6-series section.

If the angle of attack is 7 degrees, it's because the board is well sized to the load. You could get the angle of attack down to less than 3 degrees by more than doubling the area. But then you're more than doubling the parasite drag, too.

As long as the flow is attached, differences in section design are going to make fairly minor differences in the drag of the boat. The key is avoiding separation. A section that has a higher maximum lift can let you reduce the chord to make up for some increase in the minimum drag.



Is it realistic to think that a foilmaker can handcraft an Eppler or Wortman section well enough (sorry Andy P I'm sure you can) to make the effort worthwhile?

Yes, especially these days with numerically controlled machining of plugs, cores and templates.

What happens when my crew treads on the trailing edge of a laminar foil?

That depends on how strong the trailing edge is! The trailing edge doesn't have much to do with maintaining laminar flow. The boundary layer will be turbulent by the time it gets there. Scratches and small dirt deposits will be buried in the boundary layer and not cause a lot of drag.

A sharp-cornered square trailing edge of modest thickness is just as effective as a knife-edged trailing edge, and a whole lot more robust.

If the answers to these questions are what I think they will be, I suspect that a trusty NACA 0008 fattening out to a NACA 0011 will be the reliable, but somewhat unimaginative solution.

At your Reynolds numbers, the problem isn't maintaining laminar flow, the problem is how to get rid of it gracefully! Laminar flow has less skin friction, but it also separates much more easily than a turbulent boundary layer. Below a Reynolds number of 250K or so, it's not possible to achieve transition naturally without separation within the chord length of the foil. So there will be a laminar separation bubble and the section needs to be designed to make the bubble short and to move smoothly from well back on the chord at low angles of attack to near the leading edge at high angles of attack.

The NACA 4-digit sections have exactly these characteristics. So, yeah, I think I'd go with the tried and true.

I'd look to reduce drag some other way, such as by making the board longer.

Has anyone tried a Bieker-style pitch trim foil on a National 12 rudder?

(p.s. I used to sail Merlins at Cookham Reach)

[DavidG]

Has anyone tried a Bieker-style pitch trim foil on a National 12 rudder?

Three people have tried varients of the theme, one with an adjustable tab on the bottom of a rudder daggerbox cassette, one with tilting pintles with a rudder with fixed tabs and one similar to the Beiker using push cables and windsurfer fins, in all cases there is little change in performance.

I did a spreadsheet analysis of the reduction in hull wetted area using different foil areas and angles of attack and concluded that trim foils work in 14's because the V^2 function really starts working at planing speeds, and 14'ers ar not interested in lightwind sailing.

Would be interesting to know of any CFD or Tank work that can analyse the apparent energy recovery from the upwash off the hull? What do you think Tom?



(p.s. I used to sail Merlins at Cookham Reach)

I see Pat Blake around, in fact many years ago I owned Mythelated Spirit.

Anyway I am off to Salcombe and then the Championships at Weymouth.

[Milton Thrasher]

SCALLOPED WHALE FLIPPERS – DESIGN FOR FOILS?

In the August 2004 issue of Scientific American there is an article that discussed the investigation of the hydrodynamic performance of humpback whale flippers. Dr. Frank E. Fish, a biology professor at Pennsylvania’s West Chester University and a specialist in the hydrodynamics of vertebrate swimming noticed that the humpback whale pectoral flippers had evenly spaced bumps along the leading edge.

He worked with fluid dynamics engineer Laurens E. Howle of Duke University and David S. Miklosovic and Mark M. Murray of the US Naval Academy. Howle built 22 inch long plastic fins, one with the scallops and one with a smooth leading edge. In a wind tunnel at the Naval Academy testing the scalloped fins showed advantages, especially at higher angles of attack. The test results were reported in the May issue of Physics of Fluids, Leading-edge tubercles delay stall on humpback whale (Megaptera novaeangliae) flippers). The scalloped fins generated 8 percent more lift and as much as 32 percent less drag than the same size smooth fin, (probably at high attack angles). The scalloped fin resisted stall 40 percent steeper than the smooth fin.

The reported reason for the performance improvement was that the flow coming off the bumps formed pairs of counter rotating swirls or eddies coming off each side of the bumps. The bumps act as vortex generators that keep the general flow attached to the suction side of the fin when there is an angle of attack that normally would cause the flow to separate from the leading edge.

There may well be an application for r/c keels and rudders since they operate in the transient Reynolds number range between laminar and turbulent flow. The current trend for keels is to make them very thin with sharp leading edges, which makes them liable for flow separation at low angles of attack. The scalloped leading edge delays separation and thus improves turning ability without causing excessive drag. The drag of the scalloped fin was similar to the straight fin at zero angle of attack. A scalloped rudder would work the same going down wind, but when the rudder is kicked over for a tack the rudder would create a faster turn while losing less speed through the turn. It’s possible that a million years of whale evolution may have developed a better rudder for us.

Dr. Frank Fish has patented this concept.

Scalloped edges could be added to existing blades for
testing against standard blades on matched pairs of one-design sailboats such as Lasers or Sunfish to prove or disprove the theory.

I am considering trying it on my Irwin 23 with wing keel and high aspect
rudder but without a trial horse it would be hard to judge the advantage.
I would be interested in hearing the results if someone else tries it.

I will post a graphic image showing a pair of plastic fins, one curved edged and one scalloped to be sure that the concept is conveyed correctly.

[tspeer]

...Would be interesting to know of any CFD or Tank work that can analyse the apparent energy recovery from the upwash off the hull? What do you think Tom?

I agree, but I don't know of any CFD code that would be capable of handling that kind of near-field free surface hydrodynamics and be affordable to the amateur designer.

Sounds like it could be a good topic for a university student, if there was someone to fund the project.

[Milton Thrasher]

Searching for and testing of CFD code or other simulation software to prove the advantages of scalloped edges on sailboat foils would be a waste of time even if you stumbled onto some such code before sea trials are done. Until the advantages of scalloped edges are proven in actual use, you could not rely on computer simulation results.

No funding is needed for trials of scalloped edges on Laser or Sunfish sailboats. It would be very simple to cut scalloped edges into piecies of plastic and affix them to the leading edges of the foils. The cost would be very minor. There are lots of college sailing programs that have the necessary fleets of sailboats already and students with enough initiative to conduct the tests.

Another test opportunity is with radio controlled model sailboats. There are many clubs that race one-design models such as the American Model Yacht Club sanctioned Soling One Meter Class marketed by Victor Model Products of Downing, CA which can be found on the http://www.AMYA.com web pages with a link to suppliers.

We have many Soling One Design sailors here in Sarasota, FL that race the
Soling models on four different sites several days each week. I plan to contact some of the more venturesome and ask them to try the idea out.

The idea of seeking sponsorship or funding smacks of trying to get government grants for foolish projects. There are too many of those already such as studying the sex lives of African Red Squirrels which the US Government has funded recently.

[tspeer]

There's definitely a role for CFD as well as classical cut-and-try methods. Simple geometries can have very complex fluid dynamics - look at the flow around a sphere - and it takes sophisticated shapes to have simple fluid dynamics - such as laminar flow sections. I think you'd be hard pressed to stumble on a modern laminar flow section design by hand carving boards and rudders. They can really only be designed using computational methods.

The same thing could hold for tubercles. They may have evolved in whales for reasons that have little to do with hydrodynamics. Or it could be that to make them work well requires subtle shaping that would be hard to guess if you didn't understand the real flow mechanisms behind them. Fluid dynamics in general, and it seems to me sailing in particular, is filled with what I call (after Kipling) "Just So Stories" - "explanations" that get handed down and widely quoted yet are completely wrong. A classic example is, "Lift is produced because the air flows a longer distance over the lee side than the windward side."

CFD has the advantage that it gives a very detailed picture of the "why", whereas experiments often give the end result with no visibility as to the causes. The problem, of course, is that CFD always leaves out some of the physics in order to make the problem computationally tractable. If you leave out important physics, then the results won't match experiment.

The best way to make progress is a combination of computation and experiment. The computation can guide the experiment and eliminate a lot of unpromising cases. And the experiment is always needed to validate the computation.

Sure, you could hack away at Laser boards. If the mods are successful, fine, but you won't know if you're getting the most out of the tubercles or not. And if they aren't successful, is it because tubercles don't really work, or is it because you just didn't implement them correctly?

And there are lots of pitfalls with small scale models not being representative of the real thing - remember Mariner?

[sorenfdk]

I agree with Tom. It took millions of years to come up with the tubercle design of today, and who knows if it has reached the optimum point yet?

Better to try different shapes in the computer to try and discover the secrets behind the tubercles and then use this knowledge to try and design a board or a rudder.

[tspeer]

There's a role for both. If the effectiveness of the tubercles is not highly sensitive to their shape, then simple experiments would show them to be beneficial. And, unless you have the software already, that's the cheapest way to go.

But research using CFD is definitely not a boondogle.

[Frank Fish]

SCALLOPED WHALE FLIPPERS – DESIGN FOR FOILS?

In the August 2004 issue of Scientific American there is an article that discussed the investigation of the hydrodynamic performance of humpback whale flippers. Dr. Frank E. Fish, a biology professor at Pennsylvania’s West Chester University and a specialist in the hydrodynamics of vertebrate swimming noticed that the humpback whale pectoral flippers had evenly spaced bumps along the leading edge.

He worked with fluid dynamics engineer Laurens E. Howle of Duke University and David S. Miklosovic and Mark M. Murray of the US Naval Academy. Howle built 22 inch long plastic fins, one with the scallops and one with a smooth leading edge. In a wind tunnel at the Naval Academy testing the scalloped fins showed advantages, especially at higher angles of attack. The test results were reported in the May issue of Physics of Fluids, Leading-edge tubercles delay stall on humpback whale (Megaptera novaeangliae) flippers). The scalloped fins generated 8 percent more lift and as much as 32 percent less drag than the same size smooth fin, (probably at high attack angles). The scalloped fin resisted stall 40 percent steeper than the smooth fin.

The reported reason for the performance improvement was that the flow coming off the bumps formed pairs of counter rotating swirls or eddies coming off each side of the bumps. The bumps act as vortex generators that keep the general flow attached to the suction side of the fin when there is an angle of attack that normally would cause the flow to separate from the leading edge.

There may well be an application for r/c keels and rudders since they operate in the transient Reynolds number range between laminar and turbulent flow. The current trend for keels is to make them very thin with sharp leading edges, which makes them liable for flow separation at low angles of attack. The scalloped leading edge delays separation and thus improves turning ability without causing excessive drag. The drag of the scalloped fin was similar to the straight fin at zero angle of attack. A scalloped rudder would work the same going down wind, but when the rudder is kicked over for a tack the rudder would create a faster turn while losing less speed through the turn. It’s possible that a million years of whale evolution may have developed a better rudder for us.

Dr. Frank Fish has patented this concept.

Scalloped edges could be added to existing blades for
testing against standard blades on matched pairs of one-design sailboats such as Lasers or Sunfish to prove or disprove the theory.

I am considering trying it on my Irwin 23 with wing keel and high aspect
rudder but without a trial horse it would be hard to judge the advantage.
I would be interested in hearing the results if someone else tries it.



I will post a graphic image showing a pair of plastic fins, one curved edged and one scalloped to be sure that the concept is conveyed correctly.

If you would like to try the design I can supply the humpback tubercle geometry. F. Fish

[Steen Worsøe]

If you would like to try the design I can supply the humpback tubercle geometry. F. Fish

Dear Dr. Fish.

I have read you article on your work regarding the Humpback Whale flippers, which I thinks sounds very interesting.

Can you supply me with further information on your work?
- lift and drag coeffieicnt etc,
I would be very interested to see the humpback tubercle geometry.

Best regards

Steen Worsøe
Technical Manager of the MacArtney Group

[Frank Fish]

Steen, please contact me through my email at ffish@wcupa.edu and I can supply you with the information on the humpback flipper. I would be interested to hear of a possible application.

[Ilan Voyager]

An apparté from an avid reader of this thread:

For those too young, Mariner was a 12 meter JI designed by Britton Chance in the beginning of the seventies.

After a very intensive campaign on models at the test towing tank, it results a very curious shape, like a the belly of a hydropic whale pregnant at least of gemels, and the stern brutally cut. The theory behind the chainsaw cut of the stern was that some water will stay stuck to the stern, thus reconstuting a "phantom" smooth flowing stern.

12 meters JI were very heavy and slow boats with a plumb mine inside the keel. The size of the wave made at hull's limit speed was astonishing, you could see the keelson of the rear part of the boat. It's amazing to see how much money has been spent on boats so inefficient that could be beaten by a couple of kids on a Hobie 16 with worn sails.

The cut stern worked in towing tank but it seems that someone forgot or understimated the factor of viscosity of water, "more" viscous on a scale model that on the full size boat (similar to the bumblebee which can fly with its small wings because air,at its size and reynolds number, is syrup alike).

Scale models are very tricky. That explain why on ships the hydro guys work on very big models, often more than 8 meters long, and may spend weeks correcting the results with the scale factor, born of two spoons of physics, an enourmous bulk of empirism and a pinch of wild guessing.