Radically Different Yacht Keel - "Loop Keel"

Apart from developing the Monofoil speedsailing craft, Jon Howes and myself have been working on a radical new 'loop keel' for sailing yachts. There are full details of the design and how it works on our website (www.hmtmarine.com) and we would welcome any questions or comments from other members.

As a little 'teaser', the keel allows you to vary the inertia and hence the mass of the yacht, by interacting with the water flowing past the keel. In this way you can increase the mass of your yacht by as much as one third, at the flick of a switch.

The 'Loop Keel' is designed not just to create a faster yacht, but also one that is much safer and more seaworthy. We have spent three years developing the loop keel from the initial concept. The process has involved building seven different prototypes and conducting two sets of tank tests at the Wolfson Unit in Southampton. Our tests have proved that it will outperform a fin keel of the same area and mass, making it a compelling case to be fitted as the standard keel.

Most members know that there have been many claims of new designs that will be the next 'best thing' and we fully expect that it will take a huge amount of time and effort to convince the sailing community at large that this keel is better than the traditional fin.

To quote John Kenneth Galbraith "Faced with the choice between changing one's mind and proving that there is no need to do so, almost everyone gets busy on the proof."

James Macnaghten

 

Very impressive!!!!

Good job!!!

 

Loop Keel

Excellent thinking! Looks to me like the keel would develop enough lift(at higher speeds) to reduce displacement and therefore wetted surface? Did you find any need for additional pitch stabiity at faster speeds?

 

Given the structural problems that convetional fin keels have even now (The amout of mid performance yachts I've come across recently needing work on the keel and surrounding structure is shocking, its amazing more dont fall off) How are you going to make that work? Areas of that look like they will be pretty highly stresssed...

 

Doug,

The keel is not designed to provide lift, but to vary the added mass of the vessel. If you use it in it's normal configuration (forces outwards and higher dynamic pressure within the loop) then the forces all work in the right direction to resist leeway and hold the boat down - generating more power. If you reverse the direction of the forces and lower the pressure within the loop, then it can make you go faster in light conditions. However, the forces generated are in the wrong direction to resist leeway and, in addition, the loop will try to tip the boat over.

The following diagrams show the forces clearly for the normal operation. You can see immediately what will go wrong if you reverse the direction of the forces.

 

So according to your sayings, the advantage of extra righting moment provided by the downward bit of the keel hydro force overcomes the added drag of the hull whose added mass has been increased...?

Can you tell us % of overall speed increase using this new concept?

What about when the yacht reaches planning speeds? The downward force becomes an inconvenient, isn't it?

Finally, the prototype shown on your second rendering (1st post) looks a bit challenging from a structural point of view. What have you build the real prototypes out of?

Cheers,

Charles

 

Loop keel structure

Matt,

The loop keel has a significant structural advantage over fin keels in that it is a triangulated structure in the lateral load sense. The hull sides above the keel attachments form a natural load path for the shroud loads and these are transmittted, almost directly, into the keel ballast.

A normal fin keel on the other hand transmits these loads to the hull and then requires a significant ring frame, bulkhead or similar to take these loads into the root of the fin root which then carries the reacting loads from the keel in the most disadvantageous direction (ie, bending carried by the section's lowest second moment of area).

The loop keel is therefore much stronger and stiffer laterally than an equivalent fin keel. Longitudinally, the highest section modulus becomes available to carry any longitudinal loads. These are trivial in normal use of course, but anything but trivial in the case of a grounding or collision. In this respect it is no worse than a fin keel however and the hydrodynamic need to have a leading edge root extension at the hull attachment point can actually make the attachment longitudinally superior to that of a fin.

Doug,

Thanks for your comment but you have slightly missed the point here. The flow around this device must be considered as a closed ring vortex added to the the normal waterstream passing the yacht. Pressures induced on the hull by the keel flow system provide the continuation of the ring beytween the two attachment points. When not resisting leeway and with the boat upright this is therefore a lossless vortex ring with no vorticity shed into the wake (unlike twin keels). If resisting leeway, this flow system is still present and unchanged, however, there is a thrid flow system present associated with the generation of the leeway resisting force. This does produce a vortex wak just like a fin but much weaker than that of a fin due to the highly non-planar configuration, ie, lower vortex drag.

The closed loop vortex is generating an outward force all around the ring, ie, this force is trying to force water inside the loop. If the boat heels then as the weather side of the keel is exposed above the water surface this forcing of the flow tends to raise the water level within the loop, ie, above the surrounding sea level. This effect is proportional to the square of the speed. The raised sea level within the exposed loop may be regarded as a sort of travelling water ballast. Note that the keel actually affects the water ahead of, and behind the keel and so simply calculating the mass of water actually bounded by the exposed loop will give a gross under-estimate of the dynamic righting moment.

What is actually going on is a little more complex.... (you knew I was going to say that!...). The flow disturbance around any body has some momentum associated with it. This momentum is directly proportional to speed exactly as for a simple, dead mass. If the momentum contained within the flow distubance is divided by the speed of travel the result is a virtual mass which acts exactly as if it is a part of the mass of the body. This is the key to what the keel is doing.

The circulation system (closed vortex system) generates an added, virtual mass, this mass is proprtional to the vessel's speed and appears as an increase in the apparent displacement of the vessel. (there is a lot going on here and it would take a small book to get it all down... See our website for more). If this is starting to sound weird, consider this: the circulation system around the loop slows the flow through the centre of the loop and accelerates it around the outside. As there is a slower flow now passing directly beneath the hull the free stream flow has to move further from the centreline of the boat to get past, exactly as if the boat had a bigger underbody than it actually possesses. Think of it as a sort of hydrodynamic magnifying glass applied to the bottom of the boat to make it act like a larger vessel.

This added mass may be varied by changing the strength of the closed loop circulation: here are three ways of doing it;

1. Place flaps on the limbs of the keel, deflecting a flap towards the centre of the loop will increase circulation and the added mass.

2. Apply it to a yacht that tends to drop the bow on heeling (like many modern yachts in fact). This will increase the angle of attack of the leward limb, the weather limb is partly exposed and there will be some daylight within the loop, the effect is that the heap of sea gets higher as the closed circulation has increased along with the added mass.

3. Place a moveable bump uder the hull. If the rising part of the bump is within the loop it will induce an increased incidence on the rest of the loop, up goes the added mass.

These effects can be reversed, ie, the same methods can be used in reverse to reduce the added mass. This allows a light displacemnt craft to behave as a heavy displacement one and vice versa at the will of the crew. Taken to the extreme this would then become the hydrofoil to which you were alluding in your post.

As we know, heavy displacement boats can really pack on the canvas so this is very interesting when applied to lightweight boats in evil conditions. More usefully, it allows a normal cruising boat to become a whole lot more comfortable, shallower draft, often faster and can completely negate any tendency to broach ( due to the sweep applied to the loop keel, heeling takes a forward part of the weather side of the keel out of the water, this moves the lateral centre of pressure aft and can cancel the effect of moving the centre of thrust of the rig outboard).

You will gather that this has been a very involved mental and physical excercise. We are pleased with the outcome and have also bent, broken and damaged a lot of Laser 1 parts (which we used for sailing trials) by only going out if the wind was howling. We have one stalking horse in the form of a Laser with a fin and bulb added and one with a swept, flapped loop keel. The fin keel is particularly frightening downwind in a blow with lots of death roll and broaching. The loop keel you simply point and go, all the while wondering why that idiot in the other boat is making such a meal of it!

Jon.

 

Charlythe wind:

Thank you for for questions, the answers are fundamental to what is going on here:

"So according to your sayings, the advantage of extra righting moment provided by the downward bit of the keel hydro force overcomes the added drag of the hull whose added mass has been increased...?"

Yes, but obviously there will be parts of the sailing envelope where this is not so apparent. From our tank tests we evaluated the ratio of power (increase in righting moment) to increase in drag from generating that righting moment. This becomes favourable at about 15 degrees of heel allowing the boat to be pushed harder than an equivalnet fin keeled boat. Downwind this is surprisingly effective, surprising since downwind is often considered to be upright sailing. What it allows is the ability to really pile on the canvas if you feel so inclined and the tendency for the rig to take charge is significantly diminished since every time it tries boat heel is involved and then the keel can do its stuff and calm things down.

"Can you tell us % of overall speed increase using this new concept?"

This is obviously not a single answer! Upwind, the keel can take the leeway off the hull as the keel generates its lateral force without needing to crab the hull so lower hull drag is achievable. The higher effective aspect ratio can also result in lower vortex drag in this condition and a well designed loop keel should be worth about 10% or so in VMG. Downwind, the keel can be designed to allow the boat to plane and so this comes down to nerve just like any other boat. Due to the calming influence of the keel we have found that nerve lasts a lot longer with the loop keel test boat than the fin-keel stalking horse.

"What about when the yacht reaches planning speeds? The downward force becomes an inconvenient, isn't it?"

You need to consider the closed loop nature of the forces on the keel. This is actually an outward force all around the loop, completed by induced forces on the hull bottom. When the boat is upright these forces are in balance (ie, a no nett external forces either up or down). If designed with the form of the hull afterbody in mind this device does not therefore inhibit planing.

"Finally, the prototype shown on your second rendering (1st post) looks a bit challenging from a structural point of view. What have you build the real prototypes out of?"


See my post above regarding structure. The actual sailing prototypes were initially formed of skinned MDF (horrible stuff) around a welded steel core or 25mm x 50mm with a welded steel lump at the bottom for ballast. The final swept test keel was made from solid carbon fibre over a former which was then removed. Full scale the easiest way for cheap production would undoubtedly be to cast the unit in steel as one piece or form it from a specially rolled section. For one offs, the composite approach is probably best although one-off large steel castings are surprisingly affordable.

Jon.

 

One specifically for Doug!

Doug, One possible configuration of loop keel is to make it as a true circle when viewed from the front, or at least the top part including the bit through the hull, as a true arc. The keel can then pass through the hull in an arched trunking allowing the whole unit to be rolled relative to the hull. This can obviously then be used to move the ballast to weather without damaging the hydrodynamic properties of the keel.

This is obviously going to be far stronger than any current canting keel. If the keel rolling sustem breaks then the result will not be an uncontrolled hammering as occurs with current designs as the keel will still be fully constrained in its trunking and a simple locking emergency brake could clamp it in position without danger to the crew, allowing much more opportunity for repair.

Effectively, a canting loop keel.

Jon.

 

Interesting Work

Jon,

If you might, could you speak to the process of inspiration in studying the work of Lanchester with James and extending the thinking in the direction you guys have to this point?

Chris

 

It seems that you are talking of a keel system with movable parts, but the drawings don’t show them.

So, I guess that the button(s) are for doing this, but you have not posted the drawings with the movable parts. I would like to see them.

If half of what you predict "works" accordingly, this keel would be a major breakthrough in sailing technology. I hope you are right

 

Chris,

Lanchester was responsible for the circulation theory of lift. Apparent mass was something that was appreciated by others in the 19th century. As far as I can tell, Lanchester never involved himself in the added mass due to circulation systems although he was quick to realise that a vortex could not end in thin air and either needed to be infinite, closed loop, or bounded by a surface. The infinite case is the trailing vortex that any leeway resisting keel creates and has drag, the bounded or closed loop systems are effectively lossless.

The article in Yachting World (April 2007) is pretty accurate in most areas but ascribed ring wing/boxwing/bibplane theory to Lanchester. Lanchester developed circulation and vortex lift theories, Max Munk and Prandtl were responsible for the various incarnations of biplane and boxwing theory.

Fred Lanchester's work was widely panned in academic circles when he developed it in the 1890's, the last laugh is squarely with him however as his theories are not only correct, they are the basis of the computational methods now used universally throughout the aircraft industry for aerodynamic design.

Jon.

 

Vega,

"It seems that you are talking of a keel system with movable parts, but the drawings don’t show them."

Not necessarily, my earlier post (Option 2 in your post above) points out that you can get the same variable displacement effect as a function of heel angle by matching hull-form to keel design, ie, if the head drops with heel the lower keel limb will experience increased incidence and therefore circulation and induce this on the other limb. Our final Laser test boat does have flaps on the trailing edge and these are visible in the pictures.

"So, I guess that the button(s) are for doing this, but you have not posted the drawings with the movable parts. I would like to see them."

How this is implemented on someone elses boat is a mechanical design issue for the designer of that boat. On our final Laser prototype we did this with cables in conduits from the flaps to levers in the cockpit. There will be other methods.


"If half of what you predict "works" accordingly, this keel would be a major breakthrough in sailing technology. I hope you are right"

Thanks, we are pleased with the results and have actual sailing prototypes to prove it!

Jon.

 

Hi Jon,

Thanks for the response. I'm very interested in the creative process that evolved for the two of you as you hatched this idea as it applies to boating.

Could you speak to that part of the energy? Was it collaborative in nature? Do one of you have more invested in the process than the other, or did it fly out of a mutual discovery process, bouncing ideas back and forth?

Chris

 

Chris,

It was a bit of both. The original idea was mine, from before I met my business partner, James Macnaghten.

After Tony Bullimore went for an extended swim in 1996/1997 there was much talk of flutter and dynamic problems associated with high aspect ratio fins. As I had been invovled in non-planar wing design in the past and written various optimisation programs for this purpose I started to contemplate closed loop keel systems (very like a Prandtl box wing). The problem with any closed loop foil is that it is quite a difficult task to get it to work well over a range of incidence values but this can be addressed by a good optimisation code and is not therefore insurmountable. My thinking was that a closed loop keel config could be structurally vastly superior and could at least offer the same performance as a fin and should really be superior if properly designed and matched to the hull. Stalling of high aspect ratio surfaces also occurs at lower angles than for low aspect or non-planar surfaces and so the primary motivations were strength, integration with rig loads, stall resistance and potentially better performance through effective aspect ratio enhancement.

When sketching a suitable basic config in front view I noticed that the centres of the keel limbs were such that when the boat was heeled, if a constant outward force system was superimposed on the basic leeway resisting force sytem, then this constant outward force also delivered a righting moment about the centre of bouyancy of the vessel... This is when I began to get excited.

It went into one of my notebooks for a year or two until I mentioned it to James and collectively we thought it worth a go. (those blasted notebooks.... I think James is beginning to wish that I would keep them under lock and key!)

We obtained acouple of old Laser hulls and fitted one with a very crude loop keel with a ballast bulb (Clark Y section... more on this very shortly!) and a fin and bulb on the other with similar area. When we sailed them two things were apparent, the loop keel boat was incredibly stiff and so the dynamic righting moment was very evident also, the fin keel boat whistled straight past the loop keeled boat so this was only a partial success! A secondary observation was that inspite of the poor speed of the loop keel boat it did point higher and was able to keep sailing when pressed hard whereas the fin keel boat rapidly descended into a broaching display. Downwind things were even more dramatic with the loop keel being remarkably easy to sail even in a force 7 while the fin keel by then spent most of its time only marginally under control.

Back to the Clark Y section: The poor performance bothered me a lot. One thing that was clear was that the bluff leading edge of the Clark Y section was a fantastic spray producer making a nice vertical fountain when the boat was powered up. The second surprise was that the wave system of the loop keel boat looked a little extreme for the speed... First prototypes and all that! However, given that any experiment that delivers exactly the expected result is, by definition, a dumb experiment, there was enough there to make further mental effort worthwhile.

The Clark Y section, at zero geometric incidence produces a CL of around 0.5. I had incresed this further by adding about 3 degrees of incidence bringing the CL to about 0.8. When considering the momentum contained in this circulation system it was clear that I had perhaps got close to a 50% increase in apparent displacement. The foils themselves also produced a major wave system and so this needed work.

As the duct formed by the loop was effectively a convergent passage the water was having to change momentum from entry to exit. Although in an infinite continuum this would not matter, close to a water surface this could ony result in a surface wave system that we did not want.

The answer was therefore to force equal loop exit and entry areas which meant that the lift on the foils, in the closed loop sense, had to come only from camber and chord and not from additional incidence. The only way of forcing the inlet and exit areas to be equal was to use a sharp leading edge, ie, to force the stagnation point at the LE to sit in more or less the same point at all times. I designed a family of sections with tiny leading edge radii and an elliptical chordwise load distribution to reduce the tendency to produce infinite coefficients of pressure at the leading edge. As the root of each limb is the first part to leave the water I also introduced a leading edge root extension in this region as a swept leading edge tends to generate a local vortex in lieu of complete flow separation, this assisted with the difficult air-water interface region where ventilation can triger early stalling. The sharp leading edge is also essential for a surface piercing foil as it does not create spray drag in the same way as a bluff shape.

With James encouragement we decided that some test tank measured data, while probably not producing a perfect keel, would at least allow us to firstly see if calculated dynamic righting moments were about right, and secondly, to observe the keel under controlled conditions.

We were loaned a test hull to use as a base for which the fin keel performance was known and I set to and designed a loop keel for this with my new sections. As added mass was clearly playing a big part in the behaviour I also made a second loop keel for the model with 3 degrees more incidence. The results were beginning to get interesting: Dynamic righting was now very evident as was the total lack of spray from the new foils. Also, many photographs were taken which clearly showed the higher sea level within the exposed side of the loop when running. We still did not have the area and camber distributions correct as the induced drag was higher than we would have liked and I had not focussed on this at the design stage, but other benefits started to become apparent, for example, although the area of the loop keel was more than that of the fin, when working back through the results the loop appeared to have about twice as much area as it actually possessed, a very weird result indeed! (I am still working on why this may be the case and it seems to be related to the enclosed flow path created by the loop). The keel with increased incidence also had higher drag in proportion to the incidence increase and I initially put this down to the higher apparent displacement, this was not the reason and later trials showed it to be directly related to the convergent flow passage that it created within the loop.

We now went for a full understanding of the added mass effects. The Laser keel was ripped apart and replaced by one along the lines of the test tank model with a parallel flow path inside the loop. James had the idea of towing these around the local reservoir with some improvised load measurement devices (towing to one side of a RIB via various gantries etc). We modded another laser with a similar loop keel, this time with the camber reversed to create a closed loop system with the forces in the opposite direction, I was beginning to wonder along the lines of "if we can increase apparent mass can we also reduce it?" (shades of anti-matter here!). The towing at Graffam Water gave some rough indications that the drag of the loop keel was now simialr to that of the fin, further, that the reversed camber loop really did have slightly reduced displacement drag, we were beginning to smell the home straight...

Sailing trials also revealed similar drag performance from both loop and fin keels with dynamic righting and lower leeway on the loop. The handling was also transformed with the loop keel boat becoming extremely docile on all points of sailing. One surprise was that if caught in irons the loop keel was very hard to break out, further consideration of where the keel thought it was longitudinally led me to project lines from the keel limb centres of pressure normal to the zero-lift surfaces onto the centreline of the boat (you will need to read that at least three times.... sorry). This point was about 200mm ahead of the actual location of the keel so we moved the keel back by this amount and this solved the problem fully.

We went, after a period of agonising, (this is not cheap) for another set of tank trials, this time in the much larger Gosport facility. To nail the apparent mass issue once and for all we ran a series of acceleration tests to identify the variation of mass. For this series of tests we also thought it worth putting all that we had discovered into a final test design. This was the swept version, I was keen on sweep as it increases stall resistance (already high due to the non-planar configuration) and, since this keel relies on lifting a weather limb partially out of the water and out of play, allows the lateral centre of pressure to be varied with heel angle by design, this was intended to allow elimination of broaching.

The test results were fascinating. The acceleration trials, after elimination of some towing rig frequencies, showed a clear variation in mass from the lowest (reverse camber) to the next lowest (fin keel) to the flapped, swept keel with flaps deflected outwards and the highest with flaps deflected inwards. The overall variation being of the order of 50kilos in a total mass of around 300 kilos (apparent plus boat mass). The heeled towing trials showed, as expected, a range of dynamic heeling moments, very high effective area of around twice the geometric area, some drag increase associated with dynamic righting and a clear benefit in terms of power to drag ratio above 15 degrees of heel.

Sailing the swept version confirmed the test results but also showed hugely improved downwind handling to the point where mast-bending conditions actually felt quite effortless. Cranking the flaps on for upwind courses produced a noticable tendency to grip the water, bite in and go.

The dynamic of how this has developed from my original idea was very much think it, test it, talk about it, sleep on it develop it, analyse it. I find after an initial concept the talking around always helps to order ideas, not always the case... Monofoil was more or less a worked up concept prior to outside involvment but this is a very tough call indeed (and very lonely at times being so far from the beaten path, in spite of comments in various places Monofoil actually has very little in common with the Bernard Smith concept other than a canted wing... which Percy Pilcher used in the 1890s). Like many end products, the loop keel has become an amalgam of the efforts of several people, in addition to my own work, mostly James but others as well to varying degrees.

Jon