sail aerodynamics

[markdrela]

Originally Posted by RHough

If the performance is related to projected chord, I would guess that the overlapped area is next to useless?

Multielement airfoils on aircraft have nearly zero overlap --- maybe 1% chord at most. It doesn't help to increase lift.

One possible good reason for using a healthy overlap on a jib/mainsail is to push the mast farther forward into the low-velocity region on the windward side of the jib. The profile-drag penalty of the mast is proportional to the local velocity cubed, so even a modest velocity decrease at the mast can be very beneficial. I can see how this might easily overcome the skin-friction penalty of the overlap itself.

[markdrela]

Originally Posted by Paul Scott

It seems to me this argues for overlap?

No. The point I'm trying to make is that designing a slot geometry based on inviscid lift arguments is utterly pointless. The slot should be designed by viscous flow considerations. Specifically:

1) To relieve the adverse pressure gradients on the jib and mainsail boundary layers. This will increase CLmax, but only if you also crank up the AoA.

2) To reduce the velocity at the mast. This will decrease profile drag.

[RHough]

Originally Posted by markdrela

No. The point I'm trying to make is that designing a slot geometry based on inviscid lift arguments is utterly pointless. The slot should be designed by viscous flow considerations. Specifically:

1) To relieve the adverse pressure gradients on the jib and mainsail boundary layers. This will increase CLmax, but only if you also crank up the AoA.

2) To reduce the velocity at the mast. This will decrease profile drag.

That would fit with area limited classes ending up with small, non overlapping jibs.

Does it also argue for rotating the mast to reduce profile drag rather than going for overlap and hoping for a net gain?

Once the boat is powered up and max C/L is not needed, are the considerations still the same? ie Overlap is good only if the reduction in profile drag is greater than the increase in skin friction?

[Paul Scott]

Originally Posted by markdrela

No. The point I'm trying to make is that designing a slot geometry based on inviscid lift arguments is utterly pointless. The slot should be designed by viscous flow considerations. Specifically:

1) To relieve the adverse pressure gradients on the jib and mainsail boundary layers. This will increase CLmax, but only if you also crank up the AoA.

2) To reduce the velocity at the mast. This will decrease profile drag.

Mark, do your points at least partly explain why Dennis Conner's AC cat had the jib (and it's leach) so far ahead of it's wing sail (vs it's conventional rig?)? And if so, the jib would tend to migrate closer to the main (even to overlap) the more lumpy things get at the leading edge of the main? If this is true, does the importance of inflencing Kutta condition at the leach of the jib become more or less important on a sliding scale depending on the profile drag around the leading edge of the main? Or do you consider the Kutta condition/ jib leach arguement too much on the inviscid end of things?

[markdrela]

On an airplane, the shaping of the slot geometry --- slot width, overlap, contraction angle, etc --- is entirely driven by the requirement of having "nice" (non-spiky) pressure distributions on the leading edge of the rear element, and also to prevent the merging of the slat wake and rear-element BL farther downstream. In other words, it's a BL management problem. It's not a lift manipulation problem, via Kutta arguments or whatever, mainly because the slot geometry doesn't affect overall lift at a given AoA.

On a jib/mainsail, there is an additional consideration of reducing the overall velocity over the leading edge, i.e. the mast, not just in reducing localized Cp spikes. And any overlap rules in sail rating may also come into play. The AC cat had the obvious requirement that the rigid jib's leech had to clear the mainsail's leading edge when changing tacks. That would produce a very open-looking slot.

[RHough]

I'm looking at a mutlihull design ...

For a given RM limit, what would the design steps be to arrive at the optimum rig?

Is there a sail area/righting moment ratio sort of formula that is used to find a starting point? SA/D upwind of 40+ is not unusual and SA/D downwind of 70+ is also not unusual from some of the boats I've looked at.

I know that tspeer has done a lot of work to optimize planforms for high L/D while limiting heeling moment.

Before I try to hammer this out from a clean sheet of paper, I thought I'd ask if there are some rules of thumb to get the starting point close.

It seems to me that a flexible sailplan that allows the crew to change area to use the available RM to best effect is the goal. This would lead me to look at multiple head sail choices on a bowsprit of some sort (not a new idea). At some point there must be diminishing returns when adding area this way. How would you deal with large changes in the longitudinal location of the centre of effort?

Can anyone point me in the right direction here?

[Erwan]

Hi

Look at this link:http://www.tspeer.com/
If you look at "Optimum planform" I guess you will get all your answers

Best regards

Erwan

[RHough]

Originally Posted by Erwan

Hi

Look at this link:http://www.tspeer.com/
If you look at "Optimum planform" I guess you will get all your answers

Best regards

Erwan

Thanks, I'm as familiar with Tom's work as my level of understanding allows. I'm looking for a little simpler explanation perhaps ... to reduce the amount of time I have to spend relearning long forgotten math and perhaps having to struggle through all the theory that I once used more regularly when I was designing RC sailplanes ...

The problem of root gaps did not have to be considered since in aircraft design you can just eliminate the gap. It would be easier for me to relate to and understand Tom's work if there were some examples to relate real world sail plans to his theory. I may have missed those.

[tspeer]

Originally Posted by RHough

...Is there a sail area/righting moment ratio sort of formula that is used to find a starting point? SA/D upwind of 40+ is not unusual and SA/D downwind of 70+ is also not unusual from some of the boats I've looked at....

Shuttleworth has stability indices for tri's and cat's. He has indices for sideways capsize, pitchpoling and diagonal capsize.

Another good source of starting points are catalogs by noted designers, such as Kurt Hughes and John Shuttleworth. Kurt's catalog, in particular, has a wide range of designs from small day-sailers to large yachts. You're sure to find a design that's close to the sort of boat you want, and go from there.

I find the multihull footprint plot to be very useful. The idea is to compute a virtual c.g. location that provides the same heeling and pitching moment as the applied loads (black lines). The virtual c.g. will correspond to the combined center of buoyancy of the hulls (blue lines). By plotting the c.b. for different combinations of heel and trim, you can quickly see what the solution will be for any wind condition or point of sail.

Shuttleworth's stability indices can be presented in the same manner. The black lines in the previous plot were based on a modified Hazen sail model, and Shuttleworth's indices are quite comparable.

[CT 249]

Fascinating thread. Particular thanks to Mark and Tom.

Mark, re the fact that that the profile-drag penalty of the mast is proportional to the local velocity cubed. Does this mean that the comparative advantage of a lower-drag mast is reduced in slower boats, where the apparent wind is lower and therefore the local velocity is less?

Does this explain the (comparative) lack of advantage of wing masts in medium-performance boats where the apparent wind is lower?

In craft like the Tasar and NS14 and the other dinghies and skiffs that have tried wing masts, experience has not shown the sort of speed increase that many theories would indicate. This is despite development involving 1900+ boats (in the NS14s alone) over several decades.

This fact (and gust response which is also related to the inherent speed and stability of the craft) neatly ties in with the observed fact that wingmasts have a great advantage in fast una-rig cats, but no huge advantage in many other craft.

It's nice to see that the theory we get from such impeccable sources is always in accord with practise, whereas the theory we get from lesser sources is often wildly at odds with what is proven to work on the water.

[RHough]

Originally Posted by CT 249

Fascinating thread. Particular thanks to Mark and Tom.

Mark, re the fact that that the profile-drag penalty of the mast is proportional to the local velocity cubed. Does this mean that the comparative advantage of a lower-drag mast is reduced in slower boats, where the apparent wind is lower and therefore the local velocity is less?

Does this explain the (comparative) lack of advantage of wing masts in medium-performance boats where the apparent wind is lower?

In craft like the Tasar and NS14 and the other dinghies and skiffs that have tried wing masts, experience has not shown the sort of speed increase that many theories would indicate. This is despite development involving 1900+ boats (in the NS14s alone) over several decades.

This fact (and gust response which is also related to the inherent speed and stability of the craft) neatly ties in with the observed fact that wingmasts have a great advantage in fast una-rig cats, but no huge advantage in many other craft.

It's nice to see that the theory we get from such impeccable sources is always in accord with practise, whereas the theory we get from lesser sources is often wildly at odds with what is proven to work on the water.

I would think you have to know how much of the total drag is the mast. On rigs where the mast is a large percentage of the total, a wing mast would probably have greater effect. The effect of putting wing masts on a schooner might be hard to measure.

[Paul Scott]

Although Mark's statement that slot geometry doesn't change lift at a given AoA has me doodling some funky ideas...........

Paul

Don't Panic.

[Petros]

slots allow higher Cl at higher angle of attack (it delays the stall to a higher AOA), at the expense of higher drag. It seems to me what is missing in this discussion is a consideration of total drag in the rig design. Would not the best performance come from the best L/D, not merely from a high Cl? The sail/wing configuration that gives the best L/D is very different than one that gives the best Cl max. ISTM that when the whole rig has poor L/D, you would get higher healing moment, and higher leeway, both affecting performance.

[RHough]

Originally Posted by Petros

slots allow higher Cl at higher angle of attack (it delays the stall to a higher AOA), at the expense of higher drag. It seems to me what is missing in this discussion is a consideration of total drag in the rig design. Would not the best performance come from the best L/D, not merely from a high Cl? The sail/wing configuration that gives the best L/D is very different than one that gives the best Cl max. ISTM that when the whole rig has poor L/D, you would get higher healing moment, and higher leeway, both affecting performance.

I think it is wind speed dependant. On a displacement mono, below hull speed, the greatest lift that the rig can produce, is not enough to bring the boat to it's limits. There is some fixed amount of power (lift) needed to drive the boat. As soon as the rig can produce that power at less than maximum C/L, then high L/D trim provides the best performance?

[CT 249]

RE "Would not the best performance come from the best L/D, not merely from a high Cl? The sail/wing configuration that gives the best L/D is very different than one that gives the best Cl max."

I'm no expert, but Mr Drela certainly is. He has pointed out here something that accords about 100% with reality (as I have seen it) - that it is not the L/D of the rig itself that is the important thing, but the L/D of the rig and drag package.

As I understand it, if you are sailing an A Class cat, a very low-drag platform and a high-powered comparatively stable package, then increasing the L/D of the rig will create a performance advantage. The drag and required power of the whole kit 'n kaboodle is so low that reducing the rig's L/D is vital.

If, on the other hand,you are sailing an IRC racer/cuiser like a Benny 40.7, the hull drag is so high that maximising L is the vital thing; improving the L/D of the rig itself will result in a highly "efficient" rig (in L/D terms) that is so feeble that the boat will not be able to sail to the end of its slip.

If I understand Messr Drela correctly, it's not L over D of the rig; it's L of the rig over D of the hull/rig package; or is that L of the package over drag of the package?

This approach, which seems to be an elegant statement of common sense and experience, seems to be borne out in reality. It's easily seen on windsurfers, where you can shift high L rigs onto boards designed for high L/D and vice versa. After all, for all the stuff spoken about improving L/D how many of the "improved" rigs for the typical boat have ever proven to be vastly superior on the water? Very, very few - or is it none?

PS good point RHoough, I was just thinking in the limited terms of 'performance' rigs.