The design of soft wing sails for cruising

Very, very interesting.

Thanks for your work.

 

Is this the reefing alternative with all the trailing lines ?

 

The 110 came from post 31 of this thread. The post includes the XFLR5 output showing the L/D. At the time it was generated, we were only comparing sections. As such it was a simple 2D evaluation, not a full 3D evaluation.

The actual wing in the photos has lots of "construction compromises" that will limit the achievable performance. However as you read all of his posts, it is clear that he understood this before he started.

David put a lot of effort into his prototype. I really hope he is able to come back and say it was well worth it.

I still worry that getting the cloth to be smooth, wrinkle free, and have the desired contours while in service may be both very important and very difficult to achieve.

Even if his attempt does not do as good as he hoped, I see real potential for soft wings.

I am really hoping he will at least make some attempt at on the water performance assessment.

 

If your meaning is "does the chinese sheet take care of all the reefing requirement at the leech of the sail, and do I get seven reefs without having to have seven clew pennants?", then the answer is "yes"!

 

Fair enough.

Having seven levels of reefing is not all that big a feature imho.

In high winds, three or even two reefing points is plenty.

Fine tuning a reefed sail in high winds is a bit pointless. By then, any lift component is swamped by the amount of sheer pressure available.

 

Many years ago I worked in a boatshop where the forerunner of this wingsail was being developed.
http://www.advancedwingsystems.com
I understand it's limitations and am fascinated by the other soft wings by David Tyler, Omar, Beneteau et al.
There are a couple of things I am not getting and so have a big ask.
Could someone please do an overlay graphic that shows the ideal wing shape (for multihull type speeds) hard on the wind, broad reaching and downwind.
From what I am reading the current crop of soft wings aren't able to get sufficient camber for ideal off wind shape so I would like to see what that looks like, so I, and others can put our thinking caps on.
Second point. Headsails and slots, what I am getting is that headsails are almost redundant on a soft sail, is this so ?
A slot on the back 1/3 of a wing has a lot to offer does anyone have a useful link that explains how this works and the relationship between the two parts of the sail ? Sorry I don't have the resources and the time to accumulate text books on aerodynamics ! A simple explanation/graphic will suffice.
Thanks in advance,
RR

 

Don't have time for any graphics, best I can offer is an explaination as I see it.

For downwind, maximum lift to drag ratio is not as important as outright maximum lift. This is because the drag vector is favourable toward the downwind VMG. Also apparent wind speeds are lower, so minimising drag is not as important as it is at high apparent speeds such as upwind.

In this downwind case, maximum lift coefficient can best be obtained via increasing camber. There is a limit where trailing edge stall occurs as the camber becomes too great for the given operating condition. That is, The camber creates a lower pressure on the lee side until the flow seperates and the stall begins. Thus camber is limited in a single element for a given condition. Now slots or multi elements can allow greater effective camber and thus greater lift coefficient, by reliving the pressure differential, preventing the trailing edge stall, by allowing the flow from the lee side to merge with the windward side in th slot area. Thus multi elements can acheive higher lift coefficient than single elements, and are best suited to downwind sailing for their high lift coefficient. Note the heavily cambered wings of the americas cup boats when foiling downwind.

Now the upwind case...

Here lift to drag ratio is most important. That is, maximum efficiency. Maximising lift should now be offset against the induced drag rather than simply maximising it outright as in the downwind case. This is because the drag is now acting in a vector that is unfavourable to the upwind VMG. so the more drag you create, the more it pulls you downwind and reduces your pointing angle. Also we now have very high apparent wind. Windage and induced drag are proportional to velocity squared, so minimising drag can go along way at these higher apparent speeds.

Back to the first question, as far as I can figure, the ideal plan shape is largely irrelevant. Instead whats more pertinent, is the span wise loading on the wing. Planform and twist both effect it. If the span wise loading is optimal, you will be able to minimize induced drag , mostly tip losses, maximize lift for that foil area and operating condition and acheive a more favourable lift drag ratio. Thus the great effort put into the twist control ability of the americas cup wings...

 

Well done David, and thanks for the pictures. We are looking forward to hearing how the sail measures up to your expectations.

Well said groper. For a given total area, a multi-element sail will give higher maximum lift. The beauty of the junk rig and junk-type wings is that you can always add more sail area to make up for the lack of a high lift coefficient, all because of the easy reefing ability.

I have spent the last month or so toying with Tom Speer's Vortex95 spreadsheet (found on one of the sail aerodynamics threads - it is used to calculate lift and drag for a given planform and twist). I have added profile/form-drag parameters (his spreadsheet only calculates induced drag) and a number of other essentials such as true wind speed, point of sail, hull drag (hydro) and keel variables. I could then split the forces into their drive- and leeway components and calculate boat speed and heeling moments etc.

It soon became apparent what a major limiting factor heeling is, as one cannot simply go higher and skinnier on the rig without increasing the heeling moments beyond what the hull can support.

Another trend that I noticed is that performance for a given heeling moment is generally better for small sail areas operating at higher CL values than the other way around. I used a Laser dinghy as a "calibration" exercise, so all my sail variables such as CLmax is limited to a typical single sail.

"Best" upwind performance is generally found at the rig's maximum L/D operating point, which tends to be found towards the upper lift coefficient range and thus explains the observation in the previous paragraph. The lower the basic drag of the rig, the lower the value of CL at which max L/D will be found. A clean, unstayed rig such as David's soft wing might very well still be able to operate at max L/D even though the CLmax is not as great as on a sloop.

Another very important variable linked to the heeling limit is wind speed. There is only one wind speed for which max L/D sail trim corresponds to the maximum heeling loads that the hull can support. For winds lighter than that, it still pays to maximise lift beyond the best L/D value, but you will find that heeling loads start to increase much quicker for a given increase in drive.

In stronger winds the best performance will be for the reefed planform and sheeting combination that gives best L/D at maximum heel. For a triangular sail which keeps its aspect ratio constant despite being reefed, the sheeting angle that produces max L/D will also not change and neither will the max L/D value itself. The aim would be to find just the right amount of sail area to produce maximum heel when using the max L/D sheeting angle. In reality the section of the mast without any sail behind it will increase in drag and lower the max possible L/D ratio.

For rectangular sails the aspect ratio decreases when reefed, and so will the max possible L/D.

 

So for a multihull with minimal heeling that means...............?

 

It could mean any or all of the following:
Taller, higher aspect ratio rigs,
Higher CL for a given sail area and wind strength,
More sail area,

all of which means increased drive for a given boat length. You effectively take the rig of much larger mono-hull and plant it on a very light, easily driven platform...which then leads to rather humiliating scenarios to even larger mono's as in this video:

https://www.youtube.com/watch?v=k90uvFENVcY

The limit on the heeling load is still there, it is just much higher. A multi-hull can still be pushed until it capsizes or buries an ama, as the case may be.

Just note that the sail rig only provides half the forces that are at play here. The other half comes from the foils underwater, and need to be sized to match the sail rig.
As heeling loads increase, the leeway angle automatically increases until the keel experiences an angle of attack that combines with water speed to provide enough lift to balance the rig loads. The keel also has a maximum L/D operating point, and to push it beyond that point will lead to a sharp rise in underwater drag.

 

The new Boeing 787 dreamliner has a new wing, it has improved the l/d ratio for a pretty big increase in fuel efficiency. From the pics, it looks as though it has a little more washout and overcome construction difficulties to produce a lovely twisted wing... no doubt the engineers have designed for optimal span wise wing loading .

If you want to read more good info on multi element wings, lift drag and yada yada, google "high lift device" and associated terms for passenger jet aircraft flaps. This is where 99% of all the research has been done, yacht sails are only a spin off in comparison. ..

 

According to Abbott & Von Doenhoff's NACA report 824, the roughness aft of 20% chord has very little effect on either CLmax or lift curve slope.

The roughness on the leading edge generally reduces CLmax by about 0.4 on foils with thickness of 9% or more. Only a symmetrical thin foil was evaluated (NACA 0006) which showed a slight increase in CLmax, both with and without a split flap. My guess is that the roughened leading edge and subsequent turbulent boundary layer allowed a higher angle of attack before stall. Such a thin airfoil normally stalls on the leading edge like a flat plate, which limits the max AoA and hence CLmax. The actual value of CLmax for the thin rough foil was still noticeably less than that of the thicker rough foils (0.9 vs 1.4 for a NACA 4412).

Unfortunately no cambered foils were tested with such low thickness, but the data of CLmax for foils with split flaps still gives a good indication of what is possible:
The rough naca 1412 has a CLmax of 2.1 compared to 2.5 for the smooth version. Still a drop of 0.4, but percentage wise the drop is only 15% compared to 30% for the unflapped foil. Camber therefore reduces the harmful effect of roughness.

 

I think I had had seen something about aft of 20% being less important.

The getting a soft wing to be at the desired section shape was what I meant by "contour". Getting the rear of the wing right should be less of a challenge.

If the OP's wing holds shape like he wants, the forward portion looks good based on XFLR5. However, that is a big "If". If it does not, it may still be good, or it may be less good.

That is why I am so interested in actual on the water observations.

 

Great insights Will. Is there any way I could get a copy of your spreadsheet?

I have an old laser and would love to build a a high L/D soft wing for it. I am encouraged that it relies on batten building skills I have as opposed to sewing skills I do not have.

 

Thanks Skyak. I am still very new to sailing and have very little practical experience, so take anything I say with a pinch or two of salt.

I have, for instance, realised since my last post that the reason for the "boat" to reach its max L/D only at high CL values could rather be as a result of the high "base" drag of the hull and sailor. The sail itself reaches its own max L/D at much lower CL values.

The spreadsheet still needs some refinement, especially since I had to use fudge factors for it to match the real Laser polars. It is important to give every variable the proper weighting if any reliable results are to be obtained from scaling or deviation from the standard rig design. I will make it available eventually.