Probably the biggest “fail point” of modern proa designs is inattention to the problem of balancing the forces of the sail with those of the hull and rudders/boards/etc. This keypost is a lightly edited copy of a blog comment I made, which Michael asked me to post here in the forum, regarding this rendering of a proposed proa by James Brett: http://proafile.com/images/article_images/Brett.jpg  Hopefully we can initiate a discussion that will clarify the whole problem; it is after all a matter of physics.

However, the problem isn’t simple, which is why so many people have gotten the answers wrong. This post deals on a fairly superficial level with only one aspect of the issue, which is the effect on the rudders of having the driving sail forces aligned with or to lee of the hull, while the center of hydrodynamic drag is located a good distance to windward, somewhere between the hull and ama.

The most successful modern proas all have been designed with a solution to this problem in mind; Russ Brown’s boats and De Jours Meilleurs, all with masts displaced to windward, come readily to mind. It is possible, despite my doubts, that James Brett has this in hand, as the location of the mast near the windward edge of the hull, plus the apparent windward cant of the mast, move the Center of Pressure of the sail somewhat to windward. Other proa designers are well advised to examine what has worked in the past- and explore the reasons why…

        ____________________________________

This is an intriguing design and a very nice looking proa, but there appears to be a serious problem of misalignment between power and drag.

I think the mast may need to be moved closer to the lee side of the deck house! Exactly how much depends on the characteristics of the sail. Here’s why:

With the sole exception of when you are flying the ama, the lateral center of hydro drag will vary and be somewhere between the two hulls, while the Center of Pressure, specifically the “Center of Forwards Power,”  will usually be to leewards of the mast. This creates torque between the center of drag, the CFP, and the rudders, as follows:

{Drag (in ft/lbs or metric equivalent) X lever arm between CD and C-Fwd-Power} / lever arm to rudders.

This places an extra lifting load on the aft rudder, leading to higher angle of attack, higher drag, earlier stall, and larger (heavier) rudders to compensate.

Since this is a cruising boat it is unreasonable to expect that the ama will be flying constantly, and since most cruising is done on off-wind courses, this constant drag and torque will be maximized as the ama will be in highest-drag mode, supporting quite a bit of weight. There can also exist a pretty drastic positive feedback mechanism between the bow board and the stern board whenever the boat yaws seriously to windward; front board lift will increase at a rapid rate as angle of attack increases- trying to spin the boat even farther to windward- while the aft board stalls while trying to maintain course.

Keeping loads off the aft board is important on off-wind courses. The effects of circular water molecule motion as each wave front passes means the rudder will be located in a cyclical low-lift environment for several seconds at a time (because boat and wave are traveling in similar directions) and can lead to aft rudder stall and a broach.

Back to the equation: it’s obvious that if the CD and CFP are aligned, then torque on the rudders is reduced to zero. This is the point of moving the mast more to windward. This gets you (A) more lightly loaded rudders that will be less likely to stall; (B) improved directional stability, and; (C) better maneuverability in extremis.

This is also why Russ Brown’s proas have worked so well, where many others haven’t. He’s got the forces aligned! Several other proa designers have NOT paid attention to this, and their boats have had continual “rudder issues.” Now you know why… 😊

This problem will NOT show up on a small-scale prototype! Reason: on a day-sailer/camp-cruiser sized boat, mobile crew weight makes up a majority of the windward “ballast,” and when you move off-wind this weight naturally moves closer to the main hull as overturning forces decline. Therefore the center of drag also moves closer to the main hull, and the out-of-alignment effect is mostly eliminated.

On a cruising-sized boat this crew movement, if it happens at all, will account for much less change in the position of CD, so force misalignment is not reduced.

This proa sports an unusual rig and I know the designer has experimented with it on a smaller scale. I’m sure he has a good idea of where the CP is, and will be able to adjust mast position and bring CD and CP into alignment, especially for off-wind courses. It appears the mast is free-standing and so needs strong support, but it is not difficult or heavy to do this. If he can tie mast support into the deck house, this likely will result in both a reduction in overall weight and an improvement in static stability (as well as freeing up prime real estate inside the hull).

Thanks Diazo for your straight forward explanation. It explains a lot, . . .  I think! 😊  Can i check with you, though?

Does the term ‘Center of Forwards Power’ equate to the ‘Center of Effort’ of the sailplan?

Following on from your explanation, I’m imagining an end elevation of, say, James Brett’s proa with an equally imaginary line drawn vertically from the ‘Center of Drag’ located between the hulls. My understanding from what you are saying is that in an ideal situation for the rudders, the CoFP would be directly above on this vertical line and the rudders would be located directly below on this same vertical line.

But suppose it were not possible to move the mast which left the CofFP located to lee of this line as it is at present, could you improve the situation for the rudders by moving them to windward of this line a similar amount?

In other words how successfully could you balance one against the other across the line of the Center of Drag? Would the two opposing forces cancel each other out or would they add together?

cheers, James

Thinking more about my comment and question above, i have realised that the rudders will form part of the Total Drag and therefore moving them would also move the CD.
Aligning the CP over the CD makes perfect intuitive sense, Diazo. Just pushing a model across a table top with your finger would demonstrate that.
With the CP to the side of the CD, it creates a turning moment or torque, as you say. Having the rudders counter this from an equal distance (or any position for that matter) from the CD is going to counter the torque at the cost of increased drag and experienced by the helmsman as either weather or lee helm. That’s my estimation, anyway. Any thoughts? from anyone? 😊

I’ve made a handy diagram to illustrate the point being made. Lateral resultants from the sail and foils cancel one another out, what remains are the forward driving force (lift) and aft drag force, which, if not aligned, can cause a turning moment, in this case, to port (weather helm).

While I appreciate the theory and the thinking that has gone into this, there are a couple of other things that I think are worth considering… and I believe that the shortcoming is in thinking of the “aerodynamic force” generated by a sailplan as a resultant, dead-ahead vector; sailboats rarely travel in exactly the same direction as they are pointing.  Where have the aerodynamic drag and hydrodynamic lift vectors gone?

“Conventional” sailboats (by today’s norm, at least) set their centres of effort measurably aft of their centres of lateral resistance to purposely create an imbalance (weather helm) which then needs to be countered and corrected by adjustment of the rudder to maintain directional stability.  This added angle of attack of the rudder (and the inherent drag associated) creates lift to windward to help minimize leeway.  Proa’s have the interesting advantage of having one side of the craft always designated as being “to windward”, so asymmetrical foils and hull shapes can be brought more easily into play; the Polynesians have been doing it for ages…

Vector me this: how does it all apply to the Vestas Sailrocket designs?

@Michael; Thank you for that crystal clear illustration! 😊

@James;  My awkward term (used because I seem to be shorter on memory than this old computer…) “Center of Forwards Power” can be approximately equated with “Center of Effort.” As far as an illustration of basic forces such as we are doing, you can draw them from the same spot and most certainly you can design a well-balanced sailboat that way. In reality they are not located exactly in the same place, since the part of the sail immediately behind the leading edge provides comparatively more net forwards drive than the middle of the sail, which provides more net lateral force (but you need the middle of the sail- and the aft end- to get the power out of the front). You can get a better idea of what is going on from the attached image of an airfoil pressure diagram, from dauntless-soft.com You can see that the “local resultants” (if there is such a term) immediately near the leading edge point more decidedly forwards than those on the rest of the wing.

In other words how successfully could you balance one against the other across the line of the Center of Drag? Would the two opposing forces cancel each other out or would they add together?
...
Thinking more about my comment and question above, i have realised that the rudders will form part of the Total Drag and therefore moving them would also move the CD.

Yes. 😊 Actually, if the rudders were originally on the windward side of the hull as in Michael’s drawing, and you moved them to the leeward side of the hull, that would help by partially balancing the drag and moving CD closer inline with the CE.

You can certainly balance these forces against each other. For example, to counteract the offset CD you could move the sail forwards relative to the boards so that the lateral component of sail effort would tend to “spin” the boat out to leeward. At a range of sailing angles and speeds, in certain conditions, this could exactly counterbalance the “spin” to windward caused by the offset CD. But not always, and eventually you reach a point where your compromise breaks down.

Traditional proas draw an elegant balance between all these forces; the sail is well forwards, so when beating it is drawing through of the hull CLF (no boards!). The crew adjusts the CD close to the hull by moving in or out along the platform, keeping the ama just barely skimming (little drag, little effect on the boat). When the boat bears off onto a reach, the sheets are started, the sail moves to leeward, and the sail is allowed more curvature. This also moves the CE forwards of the CLF, creating a leeward-turning counterforce to counteract the driving force, which has moved farther to lee and is trying to spin the boat to windward. The crew now moves close to or onto the main hull, again keeping the CD nearly in line with the CE. If the crew then moves aft (moving CLF aft), they can push this balance all the way into a broad reach, never touching the rudder or steering paddle.

But that’s where it breaks down; as you move into a run, the sail is so far out to leeward- similar to a monohull cat-boat- that it is no longer possible to finesse the balance. There’s essentially no overturning force any more, and the ama cannot be lifted free of the water, so the CD moves out towards it, away from the hull. So the crew now has to take turns manhandling a ginormous, drag-tastic steering paddle/oar/quarter-rudder to maintain course.

The “better” solution- available to us but not to the Pacific Islanders- would be to move the sail more directly in front of the boat, again bringing the CE and CD back into alignment.

Another illustration of this balancing is implicit in Michael’s drawing. Assume for a moment that the two proas are both flying amas. This moves both CD’s over to the hull. However the heeling that flew the ama also moves the CE farther to leeward. Proa~1- the unbalanced proa- is now in much better balance. Proa~2, the balanced proa, keeps most of its original state of balance; as the heel angle moves both CD and sail force to leeward, the two forces remain in close alignment. Verdict: Proa~2 has a much wider range where the forces will remain in balance.

@MTP;  Thanks for posting! You are of course correct that there will be a leeway angle, so technically you would want to draw the power and drag vectors along that angle. However, for a modern, board-equipped proa, this angle for best speed should be kept to about 4 degrees when beating, perhaps 2 degrees reaching, and 0-1 degree running. This small angular difference doesn’t have much effect on the amount of torque created by the out-of-balance condition. For the sake of simple illustration I think we can ignore it, but for design purposes of course it is very important to consider.

The aerodynamic drag and hydrodynamic lift vectors haven’t gone anywhere; we’ve just ignored them for sake of illustrating a single point. 😊 In truth, aero drag is a lot more important on multihulls than on monos, and the lighter and cleaner the multi, the more it matters. Best practice would be to factor these into a real design. This is usually a “seat of the pants” exercise, but most/all of the fast racing cats nowadays have been designed using a more scientific approach to minimizing aero drag.

Aero drag is less important for the current illustration because you are usually going to run into your worst out-of-balance problems when you are deep-reaching or off-wind. Think again about the heeling example above; if you could keep Proa~1’s ama flying, you could sail to windward all day and never have to deal with the torque. Reaching, aero drag is much less an issue, and of course downwind it can actually increase your speed (spinnakers. etc.).

Hydro lift has two components- lifting the boat bodily, and lateral lift to resist leeway. Again relative to true course, lateral resistance always = lateral sail force, or the boat will make the adjustment for you by sliding sideways one direction or the other until the forces cancel out. However, I left it out of the keypost for a reason: while you definitely have a resultant of {leeway resistance + the drag of getting that resistance}, you mostly only have that while beating. In fact drag from leeway resistance approaches zero as you move from a broad reach to a run. So this must be looked at separately from the straight-line drag of the boat which has nothing to do with leeway resistance. The alignment-of-forces principle remains the same, but it has to be taken out of the equation a little differently in each case.

Vertical hydro lift is a tricky subject and a lot more important to proa design than most people realize. But it’s hard to calculate and computers do it a lot better than humans. Maybe 13 years ago I was helping a British fellow who was trying to write performance prediction software that would run on then-current PC’s. For testing I sent him DWG’s of a very narrow trad proa that was based on Waan Aelon Kein.  http://marshall.csu.edu.au/Marshalls/html/culture/Enewetak_Walap.html

He wrote me back; the computer said that the boat wouldn’t sail at all; the sail would drive the bow under the water instantly. Of course, Waan Aelon Kein sails perfectly well… He wrote me back a few days later, saying there was a bug and hydro lift was not being calculated properly, and that it was enough to make the boat work after all. I think he abandoned the project soon after- given the capabilities of 1998 computers it isn’t hard to see why, but since then I’ve often wondered about his result.

Vestas Sailrocket; probably the best way to think about this is to imagine that the connection between the daggerboard and the wing is a solid kite string. The way Sailrocket is set up, the daggerboard is under the hull and the wing is set up on a small ama some distance from the daggerboard (the ama flies when the boat is at speed). The angles of the two match so that the resultant of the wing draws directly through the daggerboard, which is mounted at an angle so it’s force is exactly in the opposite direction of the wing. In other words, sail forces can only be transmitted to the daggerboard directly along the lines of CE resultant, without torque. Therefore, side forces cancel out, and vertical forces cancel out; forwards thrust is delivered to the boat from the daggerboard, directly in line with CD. The rest of the boat has nothing to do with handling sail forces, until something goes wrong of course.

It is very clever and elegant, and was thought up back in the Sixties by Bernard Smith, who illustrated it in his book “The 40 Knot Sailboat.”

http://www.sailrocket.com/node/287
http://www.sailrocket.com/blogs
http://www.sailrocket.com/node/259
http://70point8percent.blogspot.com/2009/05/proa-11-to-sublime-bernard-smith-and.html
http://www.geocities.com/aerohydro/designframeset.htm

I suppose my point (as tangential as it might have been) was that it isn’t fair to take a snapshot of only the two vectors that you want to compare.  A bird’s eye view of those two vectors on either of the Sailrocket designs will demonstrate an enormous offset. The offset is in fact what keeps her stable, granted only on her optimum course.  In my school years I often proved dramatic fallacies by distracting everyone from seeing the big-picture for just long enough to make the grade….  If we pencil in the lateral hydrodynamic and aerodynamic forces on the close-hauled proa doodled above they would dwarf the longitudinal vectors to the point where their offset is virtually irrelevant.

I like the thought process here… and I’m not at all trying to be confrontational, but real-life has smacked me in the face enough times to make me always step back and ask more questions.

A new tangent:  I used to [until very recently] race quarter-tonners, with all of their IOR-inspired flaws.  I’d suggest that the only time that “those two vectors” were ever anywhere near aligned was when we had the chute poled-out on a deep reach or run… where they were so balanced that the boat was essentially rendered unstable and always flirting with a death-roll disaster.  Balance and stability (in roll or yaw) are not necessarily the same thing…

My thought?:  Let the perceived drag to forward motion encourage the craft to turn to weather… and shape the hull asymmetrically [the main hull at least, but perhaps the ama as well] to create a natural tendency to turn to the lee, while creating lift to weather.  There might be more drag, but that in turn will create more lift.  Finding the sweet-spot of efficiency is the challenge.

Diazo, thank you very much for this detailed reply. It was much more than i was expecting! Your descriptions made it very easy for me to visualize the forces affecting each other and their effect on the boat’s course.

I have a much greater understanding (and appreciation) of not only how these forces that you are talking about interact but also how ingenious the traditional proas were. Thanks 😊

This whole issue of the Ama drag is one reason I like the idea of a free standing junk sail that can rotate thru 360 degrees.  I would have a sheet to each bow of the boat for traditional shunting.the sheeting would be run to the very aft of the sail to allow the boat (1) tack in the traditional fashion in certain situations and (2) allow the sail center to be somewhere between the Vaka and The Ama on a dead run, something traditional proas don’t usually do.

The drag of the two hulls could be somewhat balanced by fine tuning the sheeting.  if the sail is not allowed to come to full parallel with the “beam” of the boat, the tendency will be to depress the Ama further into the water.  If the sail is sheeted PAST the “beam”, the resulting vector would actually tend to lift the ama.

If the sail were to leeward, this would of course be reversed.

There would of course be some major study on what percentage of the sail to have fore and aft of the mast, the actual mast location from side to side as well as the aspect ratio of the sail to fine tune all of this.

I tend to favor a lower aspect ratio for a Proa as this more closely mimics the overturning force of the traditional crabclaw, but this of course exascerbates (did I spell that right?) the movement of the center of thrust with different sheeting angles.  of of course if the mast is very near center of the sail, this reduces the variance.  Take a look at how Terho arranged the Junk on Ping Pong.  He’s much farther back with the mast than the 10% most people are using now.  the maximum camber can also be in a wide range of places depending on how the panels are cut, so this must also be considered in determining exactly where all of the thrust is, and how it’s affected on the “bad tack”.

I’m wondering just how far Brett has gotten down this road with his rig on the 53 footer…...

Tom