The design of soft wing sails for cruising

[P Flados]

Thank you. As I said above, I think I'm now in favour of e422 until something better comes along. Cl and Cd seem to be about the same as for the two sections you evaluated, but max Cl/Cd seems to go up from about 90 (at alpha = 10), to 109 (at alpha = 8), at Re = 500,000. Actually Re= 1,000,000 is about the lowest I'm interested in, that being 7 knots of wind with a chord of 4.44 metres.

Re of 1E6 is no problem.

You did not complain about the shape so I left is as is.

I did run a range of cambers for you and did pick out a sweet spot (L/D 110 ) to run with pressure distribution.

The performance is pretty good if you really can get a shape close to this. As you flatten it out for upwind work, drag is low while keeping good lift. Dial in moderate camber and L/D is great for reaching. Go for even more camber and you max lift is strong for downwind work.

 

[David Tyler]

Speaking of squalls, how have the sails & frames held up in storms? What amount of sail maintenance is required as compared to the standard set of sails (eg sloop main, jib)?

In all the miles I sailed with this rig, I never encountered Beaufort F10. Enough F8 though, and some screaming white squalls around Tasmania. Most of the mileage was with battens constructed mostly of pultruded GRP. I'll have to draw a veil over the shape of those, as I didn't know enough about aerodynamics. There were a few problems due to fasteners coming undone, and ultimately some wear and tear at the hinge point. It's easy enough to make a rig that's fit for coastal cruising, more difficult to make one that can be guaranteed for tens of thousands of ocean miles.
In New Zealand, I made replacement battens to a better profile, but only T5 alloy was available. I got trouble with fatigue, due to all the rivets and screws and nuts being assembled into soft alloy tubes. The bent tubes of the forward sections stood up to the service OK, but the single, straight after tubes deformed, and I had to increase the diameter. Again, if T6 alloy had been available, there would have been no problem.
This experience has persuaded me that future battens should be assembled from carbon braided-tube/epoxy tubes, with no mechanical fasteners other than the two essential (horizontal and vertical) hinge points. My drawing here (and attached) shows only three main components: the forward curved nose, or wishbone (cyan); a hinge component (green) and an after triangular batten (red). The nose would be made of two mouldings bonded together. The after batten would be assembled from two tubes, a flat piece at the forward end and a central bracing tube. The hinge component would be assembled from two flat pieces and a tube.
The sails don't suffer any more than with other rigs. There is no flogging. With one or two reefs down, the leech of the lower panels flaps lazily, and over time, this result in breakdown of the cloth just inside the leech (this will occur on a slab-reefed bermudan mainsail, as well). The topping lifts tend to rub against the sail, producing some shiny patches, but no holes. Perhaps the most difficult thing to achieve is an attachment of the sheet spans to the sail and/or batten that is long-lasting. They have to move through a large horizontal angle, as well as a vertical angle; probably the best thing is to make them from Dyneema/Spectra, and tie them to very small hoops on the ends of the battens.

 

[David Tyler]

I did run a range of cambers for you and did pick out a sweet spot (L/D 110 ) to run with pressure distribution.

The performance is pretty good if you really can get a shape close to this. As you flatten it out for upwind work, drag is low while keeping good lift. Dial in moderate camber and L/D is great for reaching. Go for even more camber and your max lift is strong for downwind work.

Thanks again. The conclusion that I'm reaching is that a cruising sailor such as myself won't be able to tell the difference between the performances of your original section, my adaptation of fx77w153 and e422. They all give a very acceptable Cl max around 1.6; a very acceptable pressure distribution; a very acceptable Cl/Cd.
Can I get a shape close to these?
For the all-important section from the stagnation point, at a few % aft on the windward side, around the luff to 40% aft on the leeward side, the answer is yes. The cloth tension is working in our favour.
For the convex section from the stagnation point to 40% aft on the weather side, the answer is no; the cloth is deflected inwards here, as there's nothing to hold it out, in between the battens. This is OK, because the necessarily mirrored shape of the batten nose stands too proud on the weather side, and so the sail is actually lying fairly close to the designed foil section shape(I've drawn a chain-dotted line at the approximate position of the sail here).
For the after 60%, the answer is no, if the batten is straight, as there must be some billowing of the cloth in between the battens. This biasses me towards e422, where a virtue is made of necessity: the windward side is allowed to billow inwards until it touches the concave curve of the batten, and then the tension in the cloth will hold it in shape; and the leeward side is allowed to billow out away from the batten, again until the cloth tension limits the movement.
Your final point, on adjusting the camber, I can't achieve, if I stick to my KISS principles. Remember that I'm not a racing sailor looking for the last tenth of a knot. My target is a rig that:
1. will match the performance of a very well set up cruising bermudan sloop
2. can do that when sailed by a single-handed, or elderly, or disabled, or very lazy sailor
3. has a low initial cost, has low stresses and is low-tech
4. is easy to build and easy to assemble
5. has the docility and easy handling of junk rig.

 

[JosephT]

Yes, you can't go wrong with carbon fiber...fewer parts and lighter. The problem with metal obviously manifests itself over time. The fatigue rate, especially in stronger winds, simply introduces failure.

Regarding Flados' suggestion for a low drag sweet spot (L/D 110), if you had camber adjuster with a gage on it (much like a flap gage in the cockpit of an aircraft), that could be a useful tool. It could help you achieve max performance in addition to finding a good extreme angle that may act like a real set of flaps (e.g. in case you needed to de-power quickly and reduce speed).

Here's an interesting mast config that includes adjustable spreaders to achieve desired airfoil shapes. Looks quite promising and it would be interesting to learn if there are any working prototypes out there.

http://www.strutpatent.com/patent/08281727/wind-propelled-vehicle-including-wing-sail

 

[David Tyler]

Here's an interesting mast config that includes adjustable spreaders to achieve desired airfoil shapes. Looks quite promising and it would be interesting to learn if there are any working prototypes out there.

http://www.strutpatent.com/patent/08281727/wind-propelled-vehicle-including-wing-sail

This is the Omer wingsail, designed by Ilan Gonen, and there are working rigs on several boats. It's far too complex and high-tech for my taste, though.

Have a look at
http://www.youtube.com/watch?v=1Qu_ZgQcr8k

 

[P Flados]

If you want to simply limiting the range of travel on a hinge to achieve a design camber, you are correct that on the fly camber adjustment is not available.

I can understand the desire to keep it simple, and I agree you are getting a reasonable level of performance.

I also agree that the deflection in between the ribs will not hurt much of anything once you get back away from the stagnation point. You have figured out that a good rounded transition from this region up around the stagnation point is probably needed. I am not sure how much you can control this with vertical tension on sailcloth, but this is the obvious approach.

In my opinion, the foil shape is actually more "very good" as compared to "good enough". Any required compromises with planform or in the shape transitions near the top and bottom have more potential for adverse impact than any less than idea foil shape across the middle.

Overall I think you have made good choices thus far . The rest is execution . Good luck and let us all know how it works out

 

[David Tyler]

Having settled on Eppler e422 as a suitable foil section that can be achieved in practice, the next thing to consider is the plan-form.

The lower part of the sail does not present any design difficulties: the use of chinese sheeting requires that the luff and leech shall be parallel from the clew to the top sheeted batten, to avoid difficulties in handling.

There are really only two options for the head of the sail. They are illustrated on the attached drawing.

Option 1. There can be a yard that is parallel to the battens, and long enough that its after end remains aft of the topping lift at all times.

Option 2. There can be a headboard that has its after end lower than its forward end, and is in line with the upper part of the leech, so that it will pass up between the topping lifts.

I used option 2 on my previous wing sails, but as far as I understand it, option 1 should be better aerodynamically.

Are there any comments?

 

[David Tyler]

... the next thing to consider is the plan-form.

Are there any comments?

Nothing received, so I'll settle on a short yard, parallel to the battens.

 

[Petros]

the flat top sail is far more efficent that the one with raked top. In fact you get double loss on the raked top one, a large drag area and the lose of efficiency due to the poor (pointed) top.

I suspect the efficiency would be improved futher if you make it more elliptical shaped, the upper battens a little shorter, and the longest ones at half height, and the last one or two batten/boomlets a bit shorter.

I have made a few junk rigs out of tarps for a dingy in this configuration and it is able to make much better head way, even with a flat sail, than other similar sized boats with lug or triangular sails that were much larger area. It was also a remarkably good light air performer as well. Not a completely adequate comparison, but the performance difference was noticeable despite the small size of the sail I was using.

In fact, I suspect the plan form shape may be more important, or perhaps at least as important, as the camber on the sail. I know people might want to argue with this, but I say this as a former aerospace engineer with several years in aerodynamics. Go read up on plan form theory from Abott and Von Donhoff. The closer to an ellips you can make the sail, the more efficient it will be, better L/D, less drag means less heel as well (futher improving the sails efficiency).

Something to consider.

 

[David Tyler]

And yet, I read things such as this, in Wikipedia, by Googling "elliptical wing":

The elliptical wing was first used on aircraft in the 1920s, but has only seen limited use, for a number of reasons:
- the compound curves involved are difficult and costly to manufacture,
- the pure elliptical shape as a superior planform may be a myth.[citation needed] A truncated ellipse, same span, same area, has, for all practical purposes, the same induced drag. Trapeze planforms with 0.4 or 0.5 taper ratio are induced drag equivalent, too.
- furthermore, the wing's uniform lift distribution causes the entire span of the wing to stall simultaneously, potentially causing loss of control with little warning. To compensate, aircraft such as the Supermarine Spitfire used a modified elliptical wing with washout, though such compromises increase induced drag and reduce a wing’s efficiency. Note, though, that the typical tapered wing planform has to employ more washout than the elliptical planform wing to see similar stall performance, which puts the tapered wing at a disadvantage.
Few aircraft have used elliptical wings, and even fewer have seen mass production, most in the 1930s and 40s."

And I look at the output of Boeing and Airbus, and indeed practically all modern aircraft, and find only trapezium planforms with 0.4 or 0.5 taper ratios. I look at the America's Cup sails, and not an ellipse in sight ( though there is a slight roach on the mainsail). I look at the international Moths, at the other end of the scale, and I see a trapezium planform. Why, I wonder, is it, when these guys have a mega $ budget and do a lot of research, there's not an ellipse to be seen? (I see the Airbus' Sharklets, and the Boeing version - wingtip devices to counter induced drag - though. Would the AC guys would do that, if they could?)

Is the answer that trapezium planforms with 0.4 or 0.5 taper ratios get so close to the ideal that it's not worth going for the last 1%?

It's certainly the answer as far as my project is concerned. But for eye-appeal, the attached sketch (showing a leech drawn as a part of an ellipse,alongside the straight leech) has merit ...

 

[tspeer]

The elliptical wing is only a special case, applicable to an isolated lifting surface without interference from any other source. That's not a good description of a sail. And for a sail, what you really want, except in light wind, is minimum induced drag for a given heeling moment. These planforms are more tapered than the elliptical planform, and look much like the planform of a sailboard rig. (Coincidence? I think not.)

There's little difference in drag between trapezoidal planform and an elliptical planform, and even less between a double-taper planform and elliptical. What matters much more is the twist distribution. You can take any planform and twist it to have the same spanwise load distribution and induced drag as an elliptical wing.

I think the planform shape of your sails is just fine. If you really wanted to reduce drag, you'd make them taller. But I suspect you've picked the span that works for the boat you have in the waters in which you sail.

 

[David Tyler]

Thank you, Tom.

Yes, there is a constraint on the AR - I have to fit the sail to my existing mast. There is also a project to fit this rig, as two near-identical sails (their relative areas depending on where the centre of the sail area needs to be), on to a 40ft schooner, and in that case the existing masts are taller, and the AR can be greater. We will do this by simply spacing the battens further apart.

 

[Petros]

reducing the cord at the base reduces the loss below the lower boom. it is not practial to build either sails or wings with elliptiacal tips, and there are some technical reasons why the squared off end can be better in real air in real conditions.

but I suspect with a curved trailing edge as you have drawn, approximating an ellipse, there is benefit. And it is attractive.

 

[ImaginaryNumber]

Are you making the sails of "typical" stiff sail cloth, or are you using a stretchy "alternative" fabric like some of the junk sail makers are using?

Are you using a lighter material for your double-sided sails than you would for a single-sided junk sail?

What is the weight of your soft-wing rig compared to a similar-sized straight-batten junk rig?

Do you have an opinion of Bertrand Fercot's single-wishbone soft-wing sail compared to a straight-batten junk rig? I know his design is much less aerodynamically efficient than your wing design, but it is also a much simpler KISS construction.

http://wharrambuilders.ning.com/profiles/blogs/sailing-with-grand-pha-tiki46-n-2

 

[David Tyler]

Are you making the sails of "typical" stiff sail cloth, or are you using a stretchy "alternative" fabric like some of the junk sail makers are using?

Are you using a lighter material for your double-sided sails than you would for a single-sided junk sail?

What is the weight of your soft-wing rig compared to a similar-sized straight-batten junk rig?

Do you have an opinion of Bertrand Fercot's single-wishbone soft-wing sail compared to a straight-batten junk rig? I know his design is much less aerodynamically efficient than your wing design, but it is also a much simpler KISS construction.

http://wharrambuilders.ning.com/profiles/blogs/sailing-with-grand-pha-tiki46-n-2

I used Hayward's "Sun Cruise" UV-protected cruising sailcloth for my last two suits of sails. It's softer than other brands of regular sailcloth, and lasted well. Sadly, it's no longer on the market, but there may still be some stock at Kayospruce. However, for a 3D sail, some bias stretch is actually a good thing. If you look closely at the video of the Omer wing sail, made of some hi-tech stuff, there are diagonal creases around the luff at times, as the sail twists. Also, a junk sail furls more neatly if the cloth is soft. I intend to use one of the Marlen Textiles fabrics - Odyssey is favoured for smaller sails, Mustang for medium sized sails, Top Gun for very large sails. Since I am building for ocean-going use, I'll use Mustang, or the new Top Notch 9. Yes, Odyssey could be used, but there is a small question mark hanging over its UV performance - it is only coated on one side, the coated side doesn't look good, so I wouldn't want to put it on the outside, and the uncoated side may have less resistance to UV.

The weight of the complete rig depends very much on materials and skill in construction. My estimate is that I can build the wing sail battens in carbon as lightly as would be the case for a single round tube, but that is subject to trials and confirmation.

Bertrand tried to make articulating battens first, but they failed; I think due to not have put in a horizontal axis at which the after batten can rotate upwards during gybes and at other times when the sail twists. The single wishbone battens were made to allow him to get away cruising without further development time, but I believe they're putting the clock back to about 1980, and my first experiments with soft wings. A straight batten/cambered panel sail should perform at least as well as an uncambered wing, if not better. But it's a trade-off - wing sails are easier to make, even if doubled, as they are cut flat, but the battens are more complex. Pure junk sails with cambered panels are more complex to make, but the battens are simple.