The design of soft wing sails for cruising

Before I utter any comments, please note that although I have a few years of aircraft design behind me, and have spent long hours lurking on this forum trying to map my understanding of aerodynamics to sailing, I have only just started sailing (not that 30 mins at a dinghy helm counts for anything!).

Also, aerodynamic understanding varies greatly from one amateur to the next, so if some or all of my comments are old hat, just ignore and read on.

With that in mind, here goes:

I copied your planform to CAD to take some measurements (assuming it is correctly proportioned) and got an aspect ratio of 2.05 (luff squared/area). This greatly limits the maximum lift coefficient practically attainable, regardless of which airfoil you choose.
The second major impact of the low aspect ratio is on induced drag and subsequently the 3D sail's L/D polar. While the 2D airfoil data promises L/D values peaking at High CL values, The real figures look more like this:

CL L/D
0.3 12
0.6 9
0.9 6
1.2 5

This perhaps indicates why light air performance is difficult: you try to maximise drive by increasing the AoA right up to the verge of stall for max CL, but in the process the L/D ratio collapses. In stronger wind, with the sheets let out and the AoA lower, things look better. This is most likely where the wing-sail comes into its own right: the ability to reduce AoA without luffing.

You have probably already experienced the high angles of attack possible with such a short aspect ratio sail without stalling it. This creates unique reaching potential where the drag-lift resultant can be maximised.

Although airfoil selection is still important, its profile drag is a minor contributor in the big picture for a rig like this. The Eppler 422 is probably one of the better options anyway, and fits your method of construction very well.

 

Thank you, Will.

I've decided to bite the bullet and add a metre to the mainmast, so the AR will now be 2.3.

The real question is, taking the planform as drawn as being common to both, will the 3D shape of a wing sail perform better than a single thickness junk sail? The latter with cambered panels could achieve the shape of the camber line of e422, over about 70% of its area, but the luff would be sharp, not rounded, and that, I believe, is where the difference between them will lie. The big advantage of the thickness of e422 is its tolerance to a wide variation of apparent windspeed and AoA: as the boat pitches up in the trough of a wave, the apparent wind speed goes down, and the AoA goes up; as it pitches forward over the crest, the opposite applies.

 

Just to clarify: my mention of "3D" vs "2D" in the previous post refers to the type of "flow" which in turn depends on span, i.e. 3D = sail with finite span, 2D = sail with infinite span. The higher the aspect ratio, the closer the flow behaves like 2D flow. It has nothing to do with "thick" vs "single surface" section shape, but with the tip losses or lack thereof. The airfoil data obtained by xfoil and the like almost always refer to "2D flow". Windtunnel test data typically represent 2D flow as well. This is achieved by making the foil span right across the tunnel with no gaps at the tips. This prevents any spanwise flow and tip losses. Flow around a finite sail has a spanwise component as well (hence "3D"): towards the tip on the windward side, and away from the tips to leeward.

To get back to your question: Had the thin sail been rigid, the main loss would have been on the upper AoA extreme. A sharp leading edge generally stalls earlier, limiting max CL. At least for 2D flow it does. The low aspect ratio might however mitigate that tendency to some extent.
A thin, soft sail will have the same high AoA limit, but also be limited on the lower end of the AoA range because of luffing and the camber collapsing when the leading edge is backwinded.
I suppose it then depends on exactly how much variation in AoA is at stake here, i.e. just how big are the waves and how violent is the pitching? The variation will also vary spanwise, i.e. increase with height above deck. You could use a thinner section lower down (less profile drag), but the gains would be negligible.

I am starting to wonder just how much benefit your reversible camber really has. A thick symmetrical section might give you very comparable results for the range of AoA involved. I will play some more on XFLR5 - it allows 2d Xfoil modelling as well as finite span wing analysis.

 

Ok this is what I get. I kept the planform a simple rectangle (no need to adjust reynolds numbers towards the tip) just to see what trends pop up. Aspect ratio was increased to 2.3.

I compared the e422 (green) with a pure symmetrical naca 0015 (blue) as well as a naca 2415 (2% camber)(red), assuming some camber can be obtained with slightly flexible battens and clever use of the sail's natural tendency to billow.

This is where I cannot contribute further, since I don't know what AoA or CL values are applicable in which windstrength. It is obvious that the less cambered foils have the upper hand below a CL of around 0.5. The trade-off towards the higher CL values (both CL/CD ratio and Max CL) versus construction complexity and weight is something you will have to evaluate.

 

It can be seen from the basic aerodynamic lift equations, that aspect ratio is single most influential parameter when it comes to improving the l/d ratio of a wing or sail. Tom Speer eluded to this earlier in the thread... Chasing the last few percent with different foil sections or plan shapes is almost inconsequential in comparison...

If keeping the rig size less tall is fundamentally important for the boats usage, then why not consider multiple high aspect ratio sails/wings/masts? Peak efficiency per sail area...

 

There is the problem of interaction between multiple masts to take into account. The rear mast experiences an apparent wind that is further forward than the front mast, due to the deflection of the wind by the front mast. If the front mast is trimmed optimally, all the air will be deflected directly aft, leaving the rear mast with no usable AoA to produce thrust with. Any "Lift" produced will be acting perpendicular to the boat centerline, adding to heel. A cat-yawl with its mast spaced as far apart as possible should suffer the least from this effect - I think!

It does however make one wonder about the best camber for the aft sail - I suspect a slightly flatter camber than the front sail will give better results than identical foils.

Also, for a given pair of mast heights and total sail area, would it not be better to rather transfer some mizzen area to the main, leaving the mizzen with a higher aspect ratio and in turn a steeper Cl vs alpha slope. I am sure it will help the mizzen produce more drive per unit area, but I am not sure if it will be enough to offset the increased induced drag on the main. Sorry, just thinking out loud here. It might be possible to model it as a "wing" and "elevator" in XFLR5...

 

This is by far the most baffling thing to me who is new to the practical side of sailing (I have yet to bang my head on a boom of any sort...): Why waste the precious few square feet between the lowest batten/boom and the deck? Besides the fact that it adds weight and heeling force where it is of least concern, the far more important benefit is the effect on aspect ratio, which for all practical purposes doubles, depending on how well you can seal off the sail on the deck. It need not even extend all the way to the clew, just a simple flat piece of see-through vinyl or polycarb sheet from the leading edge as far back as the life-line clearance will allow with the sail out over the side.

The mizzen could similarly be sealed off against the bridge-deck roof.

I have searched this forum in the past regarding this and came up with precious little - I can remember only one reference to someone who tried it and was most impressed with the result. Tom Speer has produced many graphs and figures illustrating the effect on span-wise lift distribution against deck-gap/mast-height ratios, clearly showing the big jump in efficiency once those last few inches are closed off.

Sure, there will be practical issues to deal with, such as crew moving to the peak (use a front hatch?) and clearing deck clutter (it was only creating drag and hampering windward ability anyway!). Correct me if I'm wrong here, but for going to windward the sail operates in a fairly narrow range of angles, so effective sail-deck sealing need not extend to the boom's full range of travel.

Here is the polar for the e422 foil with aspect ratio 4.6 (yellow). Not only does it more than double the CL/CD ratio at high CL values, it also significantly increases the CLmax attainable.

A secondary effect that I suspect might happen is that the leeward low pressure will "transfer" onto the hull topsides (as it does on an aircraft fuselage section adjacent to the wing), right down to the water. On a cat rig, the topsides adjacent to the front sail will experience some thrust and a force to leeward, while the windward topsides will experience drag and leeward forces. The low pressure transfer to leeward topsides will be more pronounced because of the sails proximity to the leeward rail, but the reduced area (heel) of the leeward topsides might cancel any benefit from this.

 

There have been a number of rigs built with symmetrical wings, either soft or hard, over the years. I can think of some of my early attempts, decades ago; the Gallant rig, the Walker Wingsail, Bertrand Fercot's current sails on Grand PHA, Gabriel Elkaim's small catamaran rig ...
The soft symmetrical wings that I know about have performed rather poorly, though they are all very docile and easy to use.
At one time, I was able to use Designfoil, and made a section similar to the Elkaim section, by taking the 'chen' foil and mirroring its top surface. It would indeed do pretty well on the computer screen, though the Cl curve had a very sharp peak, and it would be as difficult to keep "in the groove" as a thin, single surface sail, I suspect.

 

This is in fact what I am doing. My previous wing sail ketch rig had too much area in the mizzen, and the helm was badly balanced. The first reefs had to be taken in the mizzen too early, to avoid excessive weather helm. So my proposed rig has as much area in the main as I can get.
Yes, there is an adverse interaction between main and mizzen. They do not act as two parts of a single system, as a bermudan main and headsail do; they act more like two boats racing together and spoiling each other's wind, so the further apart they are, the better. The main is very much the driving sail, and the mizzen is the trimming and balancing sail, and contributes little drive to windward, as it has to be sheeted in so far. I have made the main as high as the boat can carry, at an aspect ratio of 2.3 . The mizzen could be made to a higher aspect ratio, but there seems little point as it is backwinded by the main when going to windward. Making the mizzen to a thick symmetrical section could certainly be considered, as it would be easier to sheet, and off the wind it doesn't matter what section is has.
Also to be considered is that going to windward is of prime importance to a race boat, but is only one part of the package for an offshore cruiser. A boat with a tall rig and a deep keel is uncomfortable at best, dangerous at worst, when running before a strong wind and a high sea. In the real world of offshore cruising, the motto is "moderation in all things" - no extremes in the hull or the rig.

The design at its current state of development is attached (more work on the battens and yard has been done).

 

Well, yes, for a race boat it's of prime importance. That's why we see deck-sweeping genoas. But for a cruiser, 360 visibility and ease of handling by a small crew are more important. True, the junk rig has one advantage here over western rigs - it was common practice to take the first reef upwards, by hauling on running topping lifts and mast lifts, and so a deck-sweeping wing sail in light airs is a possibility. You might want to rig a boat for single-handed ocean racing this way. However, I repeat that a cruiser needs to be more moderately rigged,with a package that is as good on all points of sail, and in all wind strengths and conditions, as can be devised, whereas a race boat goes for windward performance first and foremost.

For information, my current sail is three feet off the deck at the mast, and that's as low as it will go, consistent with clearing the pulpit, keeping the clew out of the seas on a wild reach or run, and maintaining visibility underneath it on a close reach.

 

The benefit I imagined the zero-gap sail to have for cruising was to use the improved efficiency to offset against sail area. In other words, you can get away with less sail area (with its cost and weight and handling benefits) for the same initial moderate performance.
Assuming that sail area is reduced until initial upwind performance is achieved, would that mean the other points of sail will be under-canvassed?

Alternatively-
What would be the pros/cons in keeping more or less the same sail area but in a much more squat planform (lower mast, longer chord), assuming once again that aspect ratio is only reduced until initial upwind performance is equalled? I am wondering in particular about the effect it might have on running downwind? I would imagine higher sheet-loads as one con at least...just wondering if there might be some unique advantage that could benefit a particular type of cruising.

 

If all things are otherwise equal, the lower aspect ratio ssil results in a lower L/D for the same sail area rig. higher L/D is a benefit when reaching because the boat will heel less, leaving more sail available to drive the boat forward, and when beating into the wind. The extra drag of a low aspect ratio sail is beneficial when going down wind, but the total available drive would be somewhat lower with the lower efficiency of the lower aspect ratio rig. of course, a lower aspect ratio rig also means the center of effort is lower on the mast, meaning less heel in a reach as well. hard to say. I think the higher the efficiency of the rig, the better the drive in all conditions for the same size sail.

 

There's plenty of evidence, both theory and practical, that a low quadrilateral sail (square sail, lug sail, gaff sail, junk sail) is both more efficient and easier to live with, downwind, than a tall, narrow triangle. C A Marchaj did wind tunnel testing on this many years ago. Some weekend cruisers never sail upwind, so they ought to be happy with such a rig, if they could break the triangular habit.

On the other hand, there's plenty of evidence, both theory and practical, that a tall narrow quadrilateral sail (AC wing and others) is more efficient upwind. Fast multis never go downwind, the apparent wind is always forward of the beam, so they're happy with the tall narrow quadrilateral.

In between, there all the folks like me who want to sail across oceans, up and down narrow rivers, through the difficult channels between islands where the wind comes at you from all directions, seemingly at once. To make us happy, the rig has to perform well and comfortably under many different conditions - which sometimes equates to fast, but not all the time. There's nothing worse in a cruising boat than having a rig which won't slow down. You wouldn't accept a car that only had two speeds - top speed and stop - would you? In the same way, a cruising boat must be able to sail at the right speed. That's why I'm interested to hear what the theorists have to say, but in the end, I'll be going for the moderate rig, capable in all conditions. There's a well-known phrase that applies here: "Good Enough".

 

That is why I think the same-area squat-planform rig might have some uses. Take for example your existing rig design with a AR of 2.3 and lets say you reduce the "conventionally measured" AR to 1.6 (that is roughly 19% more chord and 19-25% mast height reduction, depending on how big the initial gap was). And lets say for some practical deck layout reason that not all but a significant portion of the foot of the sail can be sealed such that an "effective AR" of 3 is achieved with the sail sheeted in.
Now when the sail is eased to a point where some of the seal is broken by the foot swinging over the rail, your effective aspect ratio is reduced by an amount that will depend on the ratio of deck beam to foot length.
Performance to windward will still be better than originally, and downwind performance also shows promise. There might therefore still remain the option to reduce sail area for a rig like this should the original performance already be good enough.

Ok, hijack off. I will stop rambling about this zero-gap idea. Thanks for the tolerance and the input. I understand the value you assign to 360 visibility. It is clearly a matter that will need attention (see-through material or whatever) if a rig like this is to be successful for cruising, and might be exactly why it has never been adopted.

 

I was about to suggest the same, but thought I should look at some comparative curves before suggesting something based on a general "rule-of-thumb".
I compared the E422 with the NACA 6414 (which has a more rounded leading edge) using designfoil. Surprisingly, the E422 has much less of a low pressure peak at the leading edge than the NACA foil at higher angles of attack, so it should handle the changes in wind direction better.

I also suspected that the good Cl/Cd values for the E422 stem from the laminar flow over the first 20% of the top/leeward surface, something which would be near impossible to achieve with soft fabric. So I checked the option for "adding insects to leading edge" to force turbulent flow over the entire foil, but the e422 still equalled or outperformed the NACA foil and showed a good advantage at Cl-max.
Interestingly, the NACA 6314 (max camber at 30%) gave almost identical results as the e422 - perhaps indicating that camber distribution is more important than leading edge roundness alone. The e422 has its max camber at 35.5%. It must have something to do with the entry angle at a given AoA.