Winglets on sails?

[champ0815]

Wow, thank you very much, Tom! This is an answer I can think about!
If I get you right, it's about the same effect if you add the area of the winglet perpendicular to a given sail or add it vertical? Very interesting! So the fuss with winglets on planes is all about reducing the length? Well, at least on a sail boat this would result in a lower heeling moment from rig weight - center of gravity would be lower.
The effect of the center of effort of the sail remains unclear to me. Stays it on the same height with winglets as with a sail where the area of the winglet is added to the top as an extension?

 

[oldsailor7]

The effect of the center of effort of the sail remains unclear to me. Stays it on the same height with winglets as with a sail where the area of the winglet is added to the top as an extension?

I guess it depends on HOW you get the extra area on the top of the sail.
If you have to extend the top of the mast to achieve it, the extra weight of the mast extension and its parasitic drag may well nullify the benifits.

The modern fathead sail with an elliptical planform would appear to solve this problem and also keep the CE of the sail at the same position.

 

[tspeer]

...
If I get you right, it's about the same effect if you add the area of the winglet perpendicular to a given sail or add it vertical? Very interesting! So the fuss with winglets on planes is all about reducing the length?

And style. Standard Class sailplanes use winglets because their span is limited by the class rules. Airliners are sometimes constrained by the width of the area at terminal gates. But if you look at Boeing's 787, for example, it doesn't have winglets. Winglets do provide the same drag reduction with less bending moment (equivalent to heeling moment), so they can be more attractive as retrofits than simple span extensions.

Here's another way to look at it. The lift is often fixed by the stability of the boat. So there are really only two things the winglet can do. It can provide a reduction in induced drag, which is equivalent to an increase in span, or it can provide the same drag with a reduction in heeling moment. But there are other means to achieve the same end, and it becomes a matter of tradeoffs of other drag sources and practical considerations as to which route to go.

Well, at least on a sail boat this would result in a lower heeling moment from rig weight - center of gravity would be lower.
The effect of the center of effort of the sail remains unclear to me. Stays it on the same height with winglets as with a sail where the area of the winglet is added to the top as an extension?

The winglet would raise the center of effort, if the winglet and sail have the optimum loading. However, it doesn't raise the center of effort as much as a physical increase in span equal to the winglet's equivalent increase in span.

With regard to the center of effort, there are really two centers to consider. The instantaneous center of effort is the heeling moment divided by the lift. Twisting off the head of the sail will lower the instantaneous center of effort.

Then there's the aerodynamic center. This is the location where the change in moment divided by a change in lift is zero. It's where the additional heeling moment from a gust will appear to act if the twist is kept constant.

Here's the exact same comparison as before, but this time the vertical axis is the height of the aerodynamic center instead of the instantaneous center of effort:

It's interesting that designing for a different induced velocity distribution, with a corresponding change in both twist and planform shape, doesn't change the tradeoff between the drag and aerodynamic center. The taller rig may have the same center of effort because it's twisted off, but in a gust, it will still feel like a taller rig.

The square-head sails are effective in two ways. They have a planform shape that is closer to the optimal planform for their span, and they twist off to reduce the height of the center of effort when hit by a gust. So the height of the aerodynamic center is effectively reduced by the feedback of the loads on the sail.

 

[champ0815]

First of all, thank you again for your elaborate answers! Since I am no engineer, it is very likely that some of the implications of your answers get lost on me. So excuse me if I ask again even when to you the answer has been already given somewhere above...
How would a square-head sail with a winglet perform? I imagine a lightweight winglet attached to the top batten of the sail, so that the flexibility of the upper part of the sail is maintained and gust response is not affected.
And how about the lower edge of the sail? In some book the theory of keels was explained, with the statement that the hull acts like a endplate making the keel so efficient. Would a cabin roof swiping boom improve the efficiency of the sail likewise?

 

[tspeer]

...How would a square-head sail with a winglet perform? I imagine a lightweight winglet attached to the top batten of the sail, so that the flexibility of the upper part of the sail is maintained and gust response is not affected.

You would have to do an engineering analysis to answer this question. It would depend on the planform of the sail, the twist, the camber (draft). And the jib and its interaction with the mainsail. The results I've shown above were calculated with a spreadsheet (which apparently doesn't run any more in the latest Windows versions) that could be modified to handle more than one surface. I've meant to do that for some time, but never got around to it.

And how about the lower edge of the sail? In some book the theory of keels was explained, with the statement that the hull acts like a endplate making the keel so efficient. Would a cabin roof swiping boom improve the efficiency of the sail likewise?

There is a huge vortex generated at the foot of the sail, and its effect is shown on the graphs above. The simple theory shows a big increase in drag for the smallest gaps, but some experimental data indicates the effect of small gaps may not be so bad:

However, for gaps bigger than about 5% of the span, the theory is not too bad.

Technically, what really matters is how the lift is distributed along the span, and that depends on how much lift is carried by the hull as well as the rig. That requires a really sophisticated computational fluid dynamics program to capture the relevant physics.

I think the practical message is the optimum planform looks a lot like a sailboard rig. We can get more out of the same sail area by raising the clew to maybe a third of the luff length, instead of having the maximum chord at the foot.

 

[oldsailor7]

Back in the day--(Early 1960s) when the only way to get a multihull was to build it yourself, we had wooden booms.

In the interest of stiffness with lightness we built them of 1" x ?" stock. (usually Douglas Fir for stiffness).

We made them in the form of a "T" with typically the vertical 1"x 6", and the horizontal top 1"x 10".

With the sail track screwed down the centreline of the flat top we had a Virtual "End Plate" which we hoped would help to reduce the foot vortex.

Of course we had no evidence of how successful it was. But it made us FEEL better about it.

BTW having been an enthusiastic Windsurfer for 10 years, I heartily agree withthe idea of raising the clew--particularly on a Catamaran with a loose footed main. Efficiency + more headroom when jibing.

 

[PI Design]

According to Fluid Dynamic Drag by Hoerner, the junction of two foils (such as a horizontal winglet atop a squre head sail) creates an interference drag that is typically equal in magnitude to the drag created from a foil 10 times longer than the length of the connection. In other words, if the square head was 1m in chord and an endplate was placed on top to prevent tip loss, the interferece drag between the sail head and the endplate would be equal to the drag one would expect from a 10m wing. So whilst you will reduce tip loss drag, you will gain an interefence drag. It will depend on the particular geometry of the design as to whether the gains outweigh the losses.

 

[Chris Ostlind]

A couple of years ago, I fiddled with this concept, but came at it from a totally different conceptual angle. I was working on the designs for a family of smallish (14-22') trimarans for speedy recreational sailing.

With capsizing and righting at the forefront of objections to multihulls, I wanted to develop a masthead flotation system that would keep the boats from turtling, allowing them to be righted quickly by the average guy with little hassle.

There have been several strategies over the years for this very issue. The Hobie "Bob" is one and also a system of installing an inflatbale bag in the masthead itself should the boat get tossed. Both of them have their own particular issues associated, so I looked for another solution.

I came up with a concept for building square topped sails that had foam pieces installed at the top of the sail providing the necessary flotation for the intended work. That the design allowed the foam to be shaped to act as an end plate device, also came into play, perhaps allowing one design concept to comfortably work for two beneficial solutions to multihull sailing issues.

There have been no further developments in the concept to date and I'm not even sure that it would work well in the endplate application when one applies the previous comments from Tom and PI. Still, it's worth a look by a savvy sailmaker who would be willing to experiment with the concept to conclusion, or discard it all together.

If anyone here has any suggestions, one way or another, it would be fun to hear what you have to say about the potential.

 

[yipster]

studying all things concerned i still say that interesting and thanks for showing
than there are inflatable wings and came acros this '66 future rig again

 

[oldsailor7]

According to Fluid Dynamic Drag by Hoerner, the junction of two foils (such as a horizontal winglet atop a squre head sail) creates an interference drag that is typically equal in magnitude to the drag created from a foil 10 times longer than the length of the connection. In other words, if the square head was 1m in chord and an endplate was placed on top to prevent tip loss, the interferece drag between the sail head and the endplate would be equal to the drag one would expect from a 10m wing. So whilst you will reduce tip loss drag, you will gain an interefence drag. It will depend on the particular geometry of the design as to whether the gains outweigh the losses.

I feel this must be an exaggeration, otherwise you wouldn't see the normal tailplane and fin arrangements on aircraft.

 

[markdrela]

According to Fluid Dynamic Drag by Hoerner, the junction of two foils (such as a horizontal winglet atop a squre head sail) creates an interference drag that is typically equal in magnitude to the drag created from a foil 10 times longer than the length of the connection.

The interference drag that Hoerner shows is indeed huge, but it's for 30% thick struts. Not surprisingly, there's massive separation at the intersection. So it certainly does not follow that you'll get similar drag penalties from the wing/winglet junction.

Much more relevant is Hoerner's figure 26. This shows the drag added by the strut intersection as a function of t/c. For t/c<10%, the added drag is practically nil. So the drag penalty of a typical airplane T-tail intersection is expected to be nil.

For a winglet/wing intersection, a more relevant parameter would be the local Cl rather than t/c, since the Cl is in the lifting case a better indicator of adverse pressure gradients. It's pretty safe to say that the wing and winglet away from the intersection must be reasonably far from local Clmax to prevent any separation problems at the intersection.

 

[Doug Lord]

I used the "wingtips" illustrated here to prevent these rc tri's from turtling. Used the same thing on a 16 footer in 1975. Both worked well for that purpose and their endplates may have worked-I'm not sure. To get an idea of the size of these rc boats click on the left picture and notice the man standing in the upper left corner!(Dr. Sam Bradfield)

 

[Petros]

If designed properly they may help, because the principles are the same for both wing, keel and sail. But as pointed out it is way more efficenct to just increase the span or aspect ratio. However in racing classes where the size of the sail is rule limited, it would be a way of "cheating" the rules. It has the effect of making the sail act bigger with a higher AR. This might work for a season or two, until they out law winglets too (or include their area in the sail area).

I have been thinking of try it out myself, just for fun.

 

[PI Design]

The interference drag that Hoerner shows is indeed huge, but it's for 30% thick struts. Not surprisingly, there's massive separation at the intersection. So it certainly does not follow that you'll get similar drag penalties from the wing/winglet junction.

Much more relevant is Hoerner's figure 26. This shows the drag added by the strut intersection as a function of t/c. For t/c<10%, the added drag is practically nil. So the drag penalty of a typical airplane T-tail intersection is expected to be nil.

For a winglet/wing intersection, a more relevant parameter would be the local Cl rather than t/c, since the Cl is in the lifting case a better indicator of adverse pressure gradients. It's pretty safe to say that the wing and winglet away from the intersection must be reasonably far from local Clmax to prevent any separation problems at the intersection.

That'll teach me to to read properly! You're quite right, the equation related to Fig 26 is what I should have used.
Please ignore my previous comments.

 

[peterraymond]

The bottom of the sail

This thread is almost 2 years old, but I couldn't find any newer ones that were as close to what I've been thinking about. I've drawn some lines and guessed some numbers from Tom Speer's post 15 and you can see the result attached below.

My red lines compare two designs with optimum planform, one with a 5% gap at the bottom and one sealed. Of course there are all the complications of flow around hulls and variation of wind strength with height above the water, but bear with me on this.

The design that is sealed at the base has around 0.68 normalized induced drag, while the second design, with the same lift and same heeling moment, has around 1.82 normalized induced drag. The design with the gap at the bottom has more than twice the induced drag. There are several reasons for this, including the fact that with a gap at the bottom you want to taper the sail at the base similar to a windsurfing sail and the fact that to maintain the same overturning moment the sail can't be as tall.

With this potential for improvement, I think that in classes where ISAF sail measurements are used something can be done. I say this because they do not count the area of nominally horizontal elements so long as they are no larger than 10% of the normal measured sail area. It seems like the only questions are how much can be done and how do you do it.

In a very simplistic example, there is a vortex trailing off the lower edge of the sail. If you put a propeller into this vortex it would slow the vortex down and also create forward thrust. This is similar to what you get from a winglet. A winglet will reduce induced drag at the base and that will let you increase the width, or cord, of the sail there. Both lower the center of effort of the sail. This lowered center of effort lets you make the mast taller, for a further reduction in induced drag. essentially you are moving to the left along the horizontal red line I added to Tom's graph.

In a source I saw recently and can no longer find, it stated that the position of a winglet front-to-back doesn't change it's effectiveness, although I assume it does change the optimum design. I had thought that the best place for a winglet would be at the trailing edge, since that is where you have the most span-wise flow. I think that there is obviously some room for design exploration and optimization.

There is another advantage too, which is that the windward winglet would produce a force downward and the leeward a force upward, so the winglets directly provide a little righting moment. Although, I may really be counting the same benefit twice here.

I picture two tapered winglets with symmetric profiles, one on each side of the sail, that counter-rotate so that one is pitched up while the other down. More complicated would of course be if each had to be designed as a two element wing and I suppose symmetric profiles would be less effective, but I think we are looking at low hanging fruit here.

At the limits, you might be able to put foils around the main beam at the mast, or mount foils on struts above the rear cross beam, but attached somewhere to the sail, or wind-sail, seems cleaner/better.