rudder that changes shape w/ angle of attack

I was speaking with some local builders and sailors, discussing the merits of twin daggerboards (or centerboards) on small boats, where one is used for each tack. In addition to canting and/or toeing in, one obvious advantage is freedom from symetrical foil sections - i.e. section can be optimized for speed and tack with a flat leeward side and curved weatherside to generate increased lift to weather. So I got to thinking (dangerous, I know) about applying the same priciple to rudder design. Specifically, how to make a rudder generate more lift with less "brakes", i.e. generate the same lift without putting the helm as far over.

Conventional thinking holds that rudders must have symetrical sections so that they can work in both directions - but what if the rudder could change shape? My current (very preliminary) thinking leads me to hang a thin rudder from a skeg with cam shaped sections, then wrap the whole package in an elastic skin. As the wheel/tiller is moved from side to side the cam shape will create a curve on the weather side. The skin on the leeward side will contract to form a relatively flat surface (of course there will be some concavity due pressure on the skin). If properly done, the sections of the rudder will generate more lift, while presenting a smaller face to the flow of water over the foil. Incidentally, at dead center the foil would be symetrical.

I know its a pretty tall order, but certainly not insurmountable. I find the construction and experimentation of such a rudder an endeavor worth undertaking. Construction, mechanical engineering, and materials science aside, where can I find information on asym foil sections moving through the water at 5 - 9 knots, at varying angles of attack (e.g. 0 - 35 degrees)? Specifically, what section shapes generate the most lift in that range of speeds and angles? The most powerful foil section for 6 knots at 20 degrees will have a slightly different shape than one for 6 knots at 10 degrees. My goal is to make the rudder morph from one shape to another based on ideal sections for a particular angle of attack. Any help in what shapes I am trying to obtain would be most valuable.

Respectfully,

Kevin Barry

 

Why not just fix the LE of the rudder to the transom, and have the tiller over-reach to move the TE from side to side with flexible skins?
Not TOTALLY tongue-in-cheek
Wouldn't want to trust it offshore, though, without years of trials.

Steve

 

Thanks for the reply Steve. I am looking to improve on current rudder design - not entirely change it. The alterations you suggest would remove the top plate from the foil (created by the hull). Same penalties with transom hung rudders, or any other case where the rudder breaks the surface of the water.

Here's a quick pic to show more what i had in mind. Pretty much a conventional skeg hung rudder, but the skin from the LE of skeg to TE of rudder forms the flat edge to leeward. The rudder post is set back to help create the bulge with the LE of rudder moving out from behind the skeg. The skin on the low pressure side will be stressed between the LE of the rudder and the LE of the skeg. The skin on the high pressure side will be stressed fro LE of skeg to TE of rudder.

The increasing the fore and aft sectional length of the skeg with respect to the sectional length of the rudder would move the bulge further back. Moving the rudder post further aft increases the speed at which the bulge is created when operating the helm. I am just trying to fing out where this bulge should be.

I will first test on dinghies (lighter, more responsive, cheaper and faster to build prototypes). If the skin were to fail, you would still be left with a largely conventional skeg hung rudder.

Kevin

 

I designed such a foil over thirty years ago but it was for a keel and not a rudder. It had a metal central ballast section covered with a rubber/plastic skin with flexible steel inner skin. It was filled with a fluid of specific gravity equal to water. A central cam could be set port, starboard or centered to flatten one side and extend the other to the desired foil shape. The fluid transfered from one side to the other through holes in the main central keel.

Never got beyond the design stage, like many other of my "bright" ideas but I do think it would work. Whether it is worth the cost and complexity is another matter. Might make a difference to AC boats though.

 

Very nice Idea

I know this is a boat-design site but I can't help feeling that I've seen this idea somewhere before... See below...

3 meter ASW 17 (ish) glider. note that the ailerons, rudder and elevator are all hinged off a flying surface. Obviously, on an aircraft all-moving rudders are difficult, and all-moving ailerons have been tried without much success. All-moving tailplanes however, are used (60inch span Toledo below, red) as they can be balanced and built so that any flex in the tailplane does not cause a catastrophic failure in structure or control.

Thinking back about 50 years, it was found that there was a loss of elevator control as Spitfires approached the speed of sound. It was finally realised that the way to solve this problem was to use an all-flying tailplane. sure enough, aircraft with all-flying tails exceeded the speed of sound. The reason for the loss of control was that the tailplane was bending due to the loads that the elevator was supplying.... back to wings...

So the obvious question is why use an aileron on a wing if it can induce control problems? well, control reversals can occur when the aileron wags the wing (rather than the other way round) and with centrally led controls, aileron-driven wing oscillation (in bending) can occur. The problems with using all-moving tips are irrelevant here, but vulnerablility is the major issue. Also, it is heavier to design a system to take the bending loads, and so the easy, light solution is ailerons. By using ailerons there is less interference drag between the control and flying surfaces.

This should not surprise anyone,

Tim B.

 

Tom,

Thanks for the feedback. At this early stage I would aim to keep cost and complexity down. With that in mind, initially I would not seal the apparatus; given that the whole thing is underwater, I figured that the voids would be filled with a fluid with specific gravity equal to water - namely water .
Although having a sealed form that is filled with fluid (perhaps under slight pressure) could aid in reducing the tendency of the high pressure side from caving in, one would need to consider that the volume to be occupied by the fluid increases with every degree off center (imagine the area of a very obtuse triangle; as two legs are brought together from say 179 degrees to 150 degrees, the increase in area is readily apparent). Basically you would have to be able to control the amount of fluid in the foil - increasing it as the helm went over, and reducing it as the helm came back to center. My initial approach is simpler (for better or worse) - I was looking at an unsealed (naturally vented) system where the elasticity of the skin is the only thing to resist the the tendency of the high pressure side from becoming too concave. I would start experimenting with plain ole rubber (approx 1/2" thickness to start) because there is about a 20% difference in length between the short side (high pressure) and the long side (low pressure).

In your keel design, how did you determine where to put the hump? How far did it stick out? How far back from the leading edge was the max? Where can I find this information? I am not a foil expert by any means, but as a mechanical engineer who loves sailing, I would appreciate it if my challenge were simply to create something that could morph into these shapes instead of having to find the best shapes through trial and error. I guess what i am looking for is a set of 7 or 8 optimized sections moving through water at approx speed of 7 Knots at angles of attack from 0 to 35 degrees in 5 degree increments. I can interpolate from there.

Regards,

Kevin

 

Articulating the leading edge is not a very effective means of increasing the rudder effectiveness. And it's not necessary to have a flexible skin and smooth variation of the camber. A hinged flap is a much simpler and effective solution.

What you're looking for is called an articulated rudder or a double-hinged rudder. They've been around since about 1896.

Becker Marine Systems sells them for ships (see animation of rudder)
As does Van der Velden Marine Systems as the Barke rudder:


One can, of course put more than one hinge into the rudder. This is known as "discrete variable camber", and was flown on the X-29 aircraft.

Several systems have been developed for continously variable camber. One example is the NASA-Boeing Mission Adaptive Wing that was flown on an F-111. Note the smooth variable camber on both leading and trailing edges. The top and bottom skins were flexible at the leading and trailing edges. Inside the wings were banana-shaped arms that acted as ribs and were rotated by actuators. The inboard part of each arm was oriented fore-aft and restrained in a pair of bearings. When rotated to the side, the camber flattened out. When rotated to the vertical plane, the banana shape formed the camber.

See:
http://www.dfrc.nasa.gov/DTRS/1992/PDF/H-1855.pdf

 

Tom,

Thanks for the pics and your insights. I have learned much from your many posts and hoped that you would respond to this thread. I downloaded xfoil this afternoon and will see what sort of shapes i can come up with.

The articulated rudder that you speak of looks a lot like the trim tab on the rudder of a Giles Vertue I sailed. In San Diego I also sailed a norsea 27 with a trim tab. Both boats had transom hung rudders with a control post coming up to the after end of the tiller. The tab was installed to assist with self steering gear. When sailing the Norsea 27 it was amazing to steer the boat with an 18" tiller attached to the control post for the trim tab. It was that experience years ago that really got me interested in rudder design. At the time i thought "this is great! why do we waste our time with long tillers?". After a great afternoon I learned that trim tabs can ease 90% of steering - but you need need direct control of the rudder when going hard over and to tack efficiently.

Works great on a transom hung rudder, but i prefer an inboard rudder for many reasons. Incidentally, I have some thoughts on controlling a trim tab on an inboard rudder - basically a large diameter hollow rudder stock to house an inner shaft that can control the trim tab. The tiller is also hollow, and the trim tab control rod has a universal joint at the tiller to rudder post connection that allows trim tab control rod to continue through the tiller to a rotating handle on the end of the tiller. You can use the tiller as normal, or rotate the handle on the end to control the trim tab and the tiller basically moves by itself. Its power steering for a tiller. I think its a pretty cool way to steer a boat, but I must admit to really liking tiller steering.

If i understand your explanation of "discrete variable camber" correcty, it is the concept that best describes what i am trying to do; there would be varying concavity to the high pressure side of the foil due to deformation of the elastic skin, a very real change in shape on the low pressure side due to the cam, and if a trim tab were thrown in the mix - there would definately be multiple hinges.

I am not an aeronautical engineer, and certainly not a foil expert - hence this post and request for information. I am looking for ideal section shapes (intuitively asymetrical) for one tack only for 7 Knots at varying angles of attack from 0 to 35 degrees. Can anyone recommend foil types or shapes? this should be a brainstorming session - please don't let questions on how to construct such a foil, nor penalties on the other tack, nor how to morph it from one shape to another restrict thought. Imagining that such a structure could be created, what are the varying sectional shapes? I will learn to use Xfoil, but can anyone get me in the ball park for optimized shapes?

Most repectfully,

Kevin Barry

 

In aeronautical parlance, there are a number of different types of flaps or tabs. A tab is basically a small flap whose purpose is to alter the hinge moments of a control surface, rather than to change the effectiveness of the surface.

The trim tab rudder you experienced is an example of a servo tab (if the tiller is directly connected to the tab and the tab moves the rudder) or balance tab (tiller is connected to the rudder, and the tab is slaved off of the rudder motion). The tab is linked to move in the opposite direction of the rudder. This actually reduces the rudder effectiveness a little bit, but the reduction in hinge moment makes it possible to move the rudder farther for a given load on the tiller.

An anti-servo tab is linked to move in the same direction as the rudder. This increases the effectiveness of the rudder - the effect you want. It also significantly increases the hinge moments the control system has to overcome. Notice where the rudder pivot is in the Becker rudder sketch above - it's nearly at the center of the rudder chord. Such a rudder would be grossly over-balanced and unstable if weren't for the movement of the flap/tab.

The discrete variable camber approach of the X-29 basically has multiple hinge lines - all of them linked to deflect in the same direction. Not quite the same as a smooth variation in the camber, but it's a good engineering compromise between the difficulties of a truly smooth change in contour and either a single flap or no flap at all.

The bane of tab sytems is any freeplay or backlash in the hinges and linkages. This can lead to flutter.

As you discovered with the Norsea 27, once the tab stalls, that's the limit of your control effectiveness. Low aspect rudders are still effective when stalled, so tab control may not be as effective at low speeds as direct control of the rudder. There's a variation on the servo tab called the spring tab. This has the driving linkage connected to the main surface with a spring as well as being directly linked to the tab.

At low speeds, where the hinge moments are small, the spring does the work of moving the surface, making it work more like the tiller is hooked directly to the rudder. At higher speeds, the loads on the rudder make the spring "give", resulting in more of the tiller's input being transfered to the tab. This adds a power-steering effect at high speeds, reducing the force on the tiller.

Alternatively, the spring can be comparatively weak, resulting in primarily using the tab for small deflections, but for large deflections the linkage hits a stop and the tiller moves the rudder directly. So you get tab control most of the time, but can still use large rudder movements at low speeds. It sounds like this would be useful for the Norsea 27.

 

Underwater Variable camber foils

Underwater Variable camber foils have been tried since more than 60 years, specially in sailing.

Theorically there are many advantages: low Cx, high Cz and big angles before stalling. On planes, submarines, some keels (America's Cup) and rigid wings (look a the Class C catamarans, variable camber foils work nicely. On rudders the results are very mitigated. The military navies and big commercial boats design groups have spent millions whitout tangible results.

The reason is the technical difficulties and problems overcome any theoretical advantage. Soft skins will vibrate badly, and will have deformations. Slots, tabs and other hinges, are mechanisms which are at best difficult to mantain in salted water.

Water is 800 times denser than air and the efforts on a rudder are higher than on a wing's plane, with generally a lot of vibrations.

As I have seen twice in my career, the whole rudder ends after the sea trials in the garbage. The real advantages are nihil.

There are more interesting paths to follow:

-Twin rudders. Both in sail and motor. For powerboats the Eppler 193 (an old simple foil easy to make) works well until 25-30 knots, add an ackerman differential steering (a very simple device) and it works well with no grat difficulties of tuning.

- On power boats I'll be more radical; suppression of the rudder and make move the propeller as it's done with an outboard or some surface drives. On big ships the justified fashion is for Azipods rotating over 360 degrees.

I think that the pod with electric engines is the best solution: good angle hull/propeller as you have not to worry about shaft. Propeller running in "clean water". Diesel engine running as generator at its best consumption RPM. No more gear box, no shaft to align, and other items. The architectural advantages are numerous. It would be interesting to try it on a small boat.

 

The trim tab on the Norsea 27 would be classed a servo tab by the definition Tom gave. However, the tab can also act as an anti servo tab as well. The only connection between the rudder and the tab is a set of pintles and gudgeons, so the tab can go both ways. On the after end of the main tiller is quarter circular bar with holes to receive a small belaying pin. Underway (especially on long tacks), instead of slightly pulling the tiller weather to compensate for helm, the main tiller can be centered once the 18" tab tiller is belayed to weather one or two notches (hence hinged in the same direction).

With regards to the wing on the X-29, there is not exactly a smooth variation in camber – would not a skin help smooth the transition? Ilan Voyager rightly points out problems with vibrations and deformation. These are very real challenges to overcome. Of course problems are exacerbated in large, and/or high speed boats. Could there not be a workable solution for a sailboat traveling up to 10 knots? What are the two boats that tried these rudders that ended up in the garbage?

My thanks to everyone contributing to this thread. Although surrounded by boatbuilders I simply would not be able to have this kind of discussion without the benefit of this forum. Still looking for refs on sectional shapes.....

Kevin Barry

 

Here are the standard references for sections:
Abbott & Von Doenhoff, "Theory of Wing Sections," Dover Books, 1959.
Eppler, R., "Airfoil Design and Data", Springer Verlag, 1990 (out of print).
Wortmann, F. X., "Stutgarter Profil Katalog Vol I and Vol II" (in Deutsch, also out of print).

You can find many sections at http://www.aae.uiuc.edu/m-selig/ads/ and http://www.nasg.com/afdb/index-e.phtml.

There are many references to section data in the literature - see the Chesapeake Sailing Yacht Symposium for papers relative to boating.

With XFOIL (http://raphael.mit.edu/xfoil/) you can generate the data you need for any single element section. For hdrofoils, it's probably better to set ncrit = 3 from it's default value of ncrit = 9. This will make the transition point more representative of test data taken in water. XFOIL can also generate data for plain flaps (no slots). So if you can define the shape, you can evaluate it quantitatively.

 

Variable camber rudders

The thread is becoming very interesting...

The 2 rudders thrown in the garbage after sea trials were for small (50 meters) warships going at about 26 knots at the beginning of the 80's. The minimal gain was worthless compared to the complication and reliability problems. Warships (and all working boats) must be very very rugged and reliable.

A good principle in engineering is MISS (make it simple and stupid) as the Murphy's Law applies harshly on all mechanical and electronical devices. It's very easy easy to make complicated devices, it demands a lot of brain juice to design simple and effective.

Before I forget: mechanically for the commands it's better to have a concentric tube around the rudder shaft than a hollow shaft. Rudder shafts have a very hard life and a hole or slot in the shaft will compromise the reliability. Also a hollow shaft will have a too big diameter and the final foil will be too thick. The price of a tappered hollow shaft in a marine metal is...let's say fairly high.

Trim tabs are a very useful feature when the the rudder needs a lot of efforts; on a small boat where the stresses are low that is totally useless and the penalty in drag is enormous.

Flaps are more effective and can be "automatic" or manual. In this case you'll start from a symetric foil which will bcome asymetric with the flap. In low reynolds is useless to spend hours on the computer to modify a laminar foil, the gain is minimal compared to a classic Naca or Selig or Eppler 12 % thick at 7 knots. The laminars are very tricky to design and need superior craftmanship to make them because of the very little tolerances, Mr Speer will confirm you. The results of the small foil design software are far from being fool proof in water.

On a small sailing boat, moving in all directions and continualy changing its speed, laminar flow is a view of mind, reality is very different. Planes fly generally at constant speed in a "smooth" medium, with no "parasitic" moves.

The best is a tolerant foil that will be working in conditions only known on the most aerobatic planes. A flap of about 25 % of the chord would be a good start.

There is also a problem of cost... common hinges won´t work as the tolerances are too big, that lends to vibrations. A vibrating flap (even on 1 degree) drags a lot and may destroy itself in the worst case. So you'll have to make special hinges with no play. They have to be small and strong, you'll be badly surprised by their price unless you have a shop. Same for the commands; no play. And so on for all the custom made pieces you'll need. I guess you won't be far of the price of a used beach cat at the end.

With the price of one hinge of the X29 you can make a school for 300 alumns in Central America and have the expenses paid for 5 years. Military have plenty of money and do not regard cost. Imitation of the solutions used on military planes is impossible: simply too high tech and expensive.

Since the Cup of America in the 1880's, very big money has been spent for improving sailboats. Almost everything, even the fooliest, has been tried. If someone would have found the miracle soft skin for rudders, all the rudders of the racing sailboats would be soft skin variable camber and the guy sleeping on a bed made of 1000 USD bank notes glued with epox...

When you see the stratospheric budgets, with big staffs of the cream of the cream of architects, engineers, aero and hydrodynamicians working with very big technical means, do not hope to make alone a variable camber rudder with soft skins, it is like burning green portraits of Franklin under a cold shower.

Look at all the true and good racing sail boats, ranging from small dinghy monohulls to big multihulls able to round the world at a mean speed of 18 knots and cross the Atlantic at 25 knots. Have they rudders with flaps? No. On the big boys the shafts are in titanium, with very special bearings, the skins are in high modulus prepeg carbon fiber cooked in oven under vacuum. Each rudder costs the price of a big car.

But after thousands of hours of engineering with many trials, the rudders remain simple high aspect foils. The foils profiles are very special (not always...) but everybody has abandonned the idea of variable camber rudders...

A very high master of aero-hydrodynamics is Lindsay Cunningham. He designed the fastest sail boat of the world. He designed also the Class C catamaran Yellow Pages which has a more than sophisticated wing with 3 slots. This boat is able to go upwind at three times the true speed of the wind, and reach the 30 knots easily. A part another Class C, nothing, absolutely nothing with sails can beat it in a triangular race with winds beetween 0 and 20 knots.

The rudder has the hard task to control an overpowered sail boat which accelerates like a motorcycle from 0 to 35 knots. The rudder must not stall in a close matching race with a lot of tactical jibes in any strenght of wind or any speed. Look at it closely: a simple foil...

Occam's razor is applicable to engineering. The simplest solution is the better.

I would spend my money on a good simple very well made rudder and maybe good sails... I'm a bit sad of being so deceptive, the idea of the variable camber rudder looks so nice but it doesnt keep any of its promises in the real ugly world of naval engineering.

 

Voyager,

Your post reaches the same conclusion that I did about diminishing returns for variable geometry foils. You spoke only of rudders though. A keel lives in a much simpler world than a rudder and if there is to be any benefit of variable geometry, it should be with a keel when sailing to windward. lots of expert sailors used to use jibing centerboards for the same reason. These could not be as effective as variable geometry though.

As for the complexity issue, I thought that a system that works automatically for each tack might solve that. After tacking, the keel would make leeway until the pressure on the leeward side forced fluid to the lower pressure windward side. The resulting foil shape should be more effective in resisting leeway than a symetrical foil. Mechanical problems aside, that was my thought when I worked on it before other more important things took over.

 

For most boats, especially given the way most boats are maintained, the section design of the rudder is not going to noticeably affect the performance. And the NACA 0012 is still a good all-round choice - after all, it was designed to be representative of sections that had proven themselves empirically to be good performers at low speeds. So while it's possible to encounter problems due to a poor choice of section, the difference between good choices is going to be very small.

What most people don't appreciate is that at low Reynolds numbers, the NACA 4-digit sections can have significant amounts of laminar flow, and the problem isn't so much maintaining laminar flow as it is getting rid of it without triggering excessive separation. The long, gentle adverse pressure gradient of the NACA 4-digit sections is very tolerant in that regard.

And as you point out, the difference between a good section and a bad one can be very subtle. For example, the NACA 6-series "laminar flow" sections have a bad reputation at low speeds. This is because they form a leading edge suction peak that leads to laminar separation and a massive leading edge stall. It's simple enough to create a similar section whose leading edge is designed for a higher angle of attack and alleviate this problem. Eppler goes through this exercise in detail in his book, "Airfoil Design and Data". One has to look very closely to see the difference between the two, but it could make a big difference in whether or not the section was acceptable for a specific purpose. Today, with the widespread availability of numerically controlled tools, fabricating a foil to close tolerances is much more practical. So if you do know what you're after, tools like XFOIL can tell you what the tradeoffs are and let you design to precise specifications.

There are a number of influences that size a rudder. If it's sized to provided the necessary yaw stability for the boat, then there's no point in articulating the rudder because it already has sufficient control power. If the rudder is sized for low speed maneuvering, then articulating it might make sense if it allowed one to reduce the rudder size and thereby reduce the drag. And of, course, tabs have proven to be very useful offshore for self-steering gear and to reduce the power requirements for autopilots. So the decision to go to a more complex steering arrangement can only be made in the context of the total system impact on the boat.

The reason one doesn't see articulated rudders on America's Cup boats is because their class rules only allow two moveable appendages, and an articulated rudder would count as both. Which begs the question as to why they settled on two and not three or some other number. While they don't want to give too much scope to radical designs that could make a significant difference in speed between the boats, I believe it's because the keel flaps have shown themselves to be effective but additional complexity has not.

I'm skeptical of flexible skins, too. That's why I brought up articulating the rudder as a way to change the camber without the difficulty of trying to do it smoothly. With a properly designed leading edge, there's little reason to change the camber there and the real gains will be made by cambering the trailing edge - which is what a conventional flap does.

And I'm in complete agreement about a single-piece rudder being the best choice for nearly all applications. You'd have to have really well-defined engineering requirements to drive you to a different solution - in which case you're not likely to be looking for advice on this forum! Much better to pick a decent section and concentrate on the planform design and sizing of the rudder than to try to build something that's going to have ill-defined lumps and bumps in the contour.