Proposal: Sail Performance Analysis Rig (SPAR)

[Dave Z]

Proposal toward an Inexpensive Sail Performance Analysis Rig (SPAR)
by Dave Zeiger (http://Triloboats.com)

SPAR is an idea I’ve been kicking around for a while. I’ve come to realize that actual development would benefit from the collaborative efforts of a community of potential users, rather than from my own best guess at the needs involved.

The goal is to be able to perform the kind of rig analysis and comparison that typically requires access to a wind tunnel, or to two or more identical hulls. The former is too expensive, and the latter, while loads o’ fun, involves willing participants, considerable resources and introduces many extraneous variables which cloud results.

How about an apparatus home-built to a standard plan, which can mount scaled rigs and measure the angle and intensity of the thrust they generate?

The sketch shows a base of (framed) half sheet of thick, standard plywood. Struts (pipe?) extend from the base corners, converging on a simple gimbaled mount. A mast is mounted with a set length extending below the pin. Four standardized spring tensioners center the mast heel (at rest) over a grid affixed to the base. Windspeeds are measured by a standardized anemometer. The apparatus is fixed in place, oriented into the wind, while a rig is put through its paces.

So long as the system components are standardized, individual comparisons may be conducted under uniform constraints. Test sails of given area, for example. Likely, though, no constraint will be relevant to all investigations, so, as in any experiment, full documentation of purpose, set-up and conditions will be important for 'peer review'.

While on the 'rough and ready' side, such a platform should generate reasonably accurate data when conventions are observed.

Points for equipment standardization:

• Plan of apparatus
  Anemometer
  Windvane (for determining wind/orientation)
  Corner tensioning devices (spring steel? surgical tubing? Bungee? fishing scales?)

Points for comparison standardization:

• Conventions for reporting sheet and sail attachments
  Categories for comparisons (e.g., area, max spar length, location of CE)
  Grid(s) for data readings
  Reporting form(s)
  Guidelines for use
  Guidelines for extending comparisons and/or standards.

I propose a brainstorming forum to run for a set time. That a ‘committee’ work up a first draft which is submitted for forum review. That a working draft be field tested, with evaluation at the end of, say, a year before final drafting. Further, that a clearinghouse could be established to collect and publish results.

I’d think the whole thing could be built and equipped for little financial outlay, making it widely accessible.


Notes:

Smaller scales are possible and may be preferable, using artificial wind. This would require a standard, variable speed fan, and guidelines for scaling.

My guess is that this larger size is sufficient to somewhat smooth 'transients'... small errors from variation of materials or construction, as well as the vagaries of natural wind.

Richard Dryden has suggested an option to record mast base deflections.

My best initial thought is to use thermal paper for the grids and a high IR laser pointer, mounted at the base of the mast, to 'print' a trace of mast base deflections directly onto the paper. This approach avoids any frictional interference from more typical spring-loaded marking devices.

[Dave Z]

SPAR Gimballed Tube vs Pin, and Heel Stop Collar: Addressing Rotating vs Fixed Mast Issues

Richard Dreyden is looking into a rotating mast rig. His solution is to use a gimballed tube, capped at the bottom, in place of the pin through the mast. His mast will rotate within it.

This strikes me as an improved general approach to the basic SPAR plan, with the following advantages:

• Masts can be simply slipped into the SPAR's tube, and may be fixed or not as each subject demands. This avoids driving a pin through the mast at the gimbal/partners. While construction is complicated a bit, set up is easier.
• Any point/marker/recording device can be permanently mounted in the tube, rather than each mast under consideration.

A consideration for equivalent comparison between rotating vs fixed masts is some means to ensure that the length of mast below partners (gimbal pins) can be made equal. The lever arm of a rotating mast with a bearing device at its heel would be shortened, and thus deflections would be greater in comparison to those not stepped on a bearing, skewing data returns.

Perhaps a bearing collar set at a standard height, low in the tube (fixing the mast heel stop height), with some space below it for specialized hardware at the heel? This would determine a standard lever arm regardless details below the collar.

Dave Z

[Robert Biegler]

I can see how knowing the direction of the wind and of the deflection at the base gives you the lift to drag ratio, but only if any vertical sheeting forces are contained within the rig, through a really strong vang, or a wishbone boom. Sheeting as shown in the picture would pull the rig back and mimic aerodynamic drag.

I don't see how this set-up can distinguish between sail force and torque. A sail with a higher centre of effort would always produce more torque for the same total force.

Edmond Bruce suggested measuring aerodynamic drag angle (a transformation of lift to drag ratio) by anchoring a boat on a bridle, in a pond without current, and measuring the angle of the anchor rode to the wind. That measurement would include the aerodynamic drag of the hull, which is good for predicting the performance of a rig on this hull, but not good for measuring the aerodynamics of the rig itself to compare to someone else's measurements using a different hull.

A shallow floating platform, barely breaking the surface, might solve the problem. So an alternative suggestion is to attach the rig to a platform consisting of a connecting structure at or below the water surface, and three or four floats as far from the rig as possible. Partially fill with water to get the floats as low as possible. Then proceed as Bruce described. Measuring the force on the anchor rode would measure aerodynamic force, and combining that with a wind speed measurement would allow calculation of a coefficient of lift.

Keeping the angle of attack constant as the platform swings in an arc requires two parallel anchor rodes. I think that is a necessary complication. And keeping the platform upright may require setting up the forward rode as a bridle that leads part way up the rig.

[Dave Z]

Robert Biegler writes:

I can see how knowing the direction of the wind and of the deflection at the base gives you the lift to drag ratio, but only if any vertical sheeting forces are contained within the rig, through a really strong vang, or a wishbone boom. Sheeting as shown in the picture would pull the rig back and mimic aerodynamic drag.

I don't see how this set-up can distinguish between sail force and torque. A sail with a higher centre of effort would always produce more torque for the same total force.

Hi Robert.

I'm already over my head! I'll leave answers to wiser heads. Please forgive my ill-informed questions...

As an amateur, I'm pretty much solely interested in comparing competing equivalent sail forms in terms of forward thrust generated by the sail, relative to the wind across a range of sheeting angles, and the cost in heeling force. Both of these seem to be indicated by mast heel deflection in the SPAR set-up.

Q: Do you agree? If not, can you explain (in simple terms) why not?

Q: I'm not sure what you mean by sail torque... is that what I'm thinking of as heeling force?

I agree that a higher CE has predictable effects. But part of what SPAR helps analyze is actual CE (as opposed to 2D geometric center). It also helps analyze trade-offs between, say, amount and incidence of windward force vs amount of heel.

Q: Does this make sense, and/or satisfy?

Upward vectors do not seem to be directly measured by SPAR... that's another aspect that could possibly be measured in a floating rig.

Q: Are the aspects not measured by the SPAR (as drawn) worth the complications of a floating rig?

Q: If so, might that be specified as an option for those with higher levels of interest and resources?

Overall, I think that the SPAR approach can be standardized at several levels of precision and range of factors to which it is responsive. Any individual variation or extension seems possible - even one off - so long as it is well described when reporting results for 'peer review'.

My hope is that the less rigorous levels would remain within easy reach of the backyard researcher, while passing useful results up the chain of dedication.

Dave Z

[Robert Biegler]

As an amateur, I'm pretty much solely interested in comparing competing equivalent sail forms in terms of forward thrust generated by the sail, relative to the wind across a range of sheeting angles, and the cost in heeling force. Both of these seem to be indicated by mast heel deflection in the SPAR set-up.

Yes, and that is exactly the problem. Say you tighten a vang on the sail in your drawing, and that increases lateral deflection. Was that increased deflection caused by the sail having a better shape and producing more force, or has the increased deflection been caused by the sail's centre of effort rising, while force remained the same or even decreased? If you have a single measurement for both force and lever arm, you can't know how each changed when you did something to the sail.

Q: I'm not sure what you mean by sail torque... is that what I'm thinking of as heeling force?

Well, heeling moment, not heeling force, because the heeling moment is a torque, which is force times lever arm. The distinction is central to what I think is a problem with the proposed set-up, so this is not just nitpicking.

I agree that a higher CE has predictable effects. But part of what SPAR helps analyze is actual CE (as opposed to 2D geometric center).

I don't see how. I think this setup can't distinguish between the effects of a change in force and the effects of a change in lever arm.

It also helps analyze trade-offs between, say, amount and incidence of windward force vs amount of heel.

My turn to ask for clarification. What do you mean by windward force?

Upward vectors do not seem to be directly measured by SPAR... that's another aspect that could possibly be measured in a floating rig.

What I have in mind would only measure horizontal components.

Q: Are the aspects not measured by the SPAR (as drawn) worth the complications of a floating rig?

Depends on how important it is to you to distinguish between sail force and heeling moment. I don't think SPAR can do that. I think SPAR can only give you the lift/drag ratio, and that only if you sheet the sail so that the sheet is perfectly horizontal, so that it does not pull the mast aft. The moment you allow the sheet to pull the mast aft, you are not even measuring lift/drag any longer.

With those caveats, lift/drag curves as a function of angle of attack are useful. But if that's all you have, high aspect ratio rigs will always look better than otherwise similar lower aspect ratio rigs, and I'm sure you have experience telling you that is not always so. Your triloboats don't use very high aspect ratio, for what I expect are very good reasons.

Regards

Robert

[Dave Z]

Hi Robert,

Okay, I've thought about your comments and think I have it clear, more or less, what your mean. In response, I'll refer to the following diagram:



To start by answering your question, my ill-chosen "windward thrust" is labelled "Thrust Vector" in this diagram, which is the total sail vector as transmitted to the mast. I've also heard it call 'drive'.

Robert Biegler wrote:

Say you tighten a vang on the sail in your drawing, and that increases lateral deflection. Was that increased deflection caused by the sail having a better shape and producing more force, or has the increased deflection been caused by the sail's centre of effort rising, while force remained the same or even decreased? If you have a single measurement for both force and lever arm, you can't know how each changed when you did something to the sail.

Hmm... I would see lateral deflection increasing the angle A+B, but not the magnitude (as indicated by length of vector, or distance of heel from center). If more force is produced, the magnitude would increase. A combination would do both.

In other words, distance of the mast heel from center indicates summed magnitude of forces transmitted to the mast, while angle from the wind indicates the summed direction of forces. So I don't see that there is a single measurement... there is a single point (the mast heel, but it is 'plotted' against centering spring tensions (magnitude) and on a radial graph relative to the wind (angular direction).

[Unfortunately, the equal deflection contour won't be a circular radius, but rather a rounded square?]

Imposing some arbitrary 'vessel' CL seems unnecessary but often helpful. Sail angle needs to be set and recorded for any experiment, and a CL makes a convenient reference. Furthermore, it's often more useful for the sailor to diagram vectors relative to the CL than the wind.

RE Lift/Drag Ratio: As I understand it, sail L/D is an aspect that doesn't get measured by SPAR, except indirectly as it impacts magnitude and direction. But I'll try to read up on it.

RE floating platform: Any forces directed up or down would affect displacement (draft) of a floating platform, so could (in theory) be measured. Likely impractical, however, I'm guessing.

Does that satisfy, or am I still missing something?

Dave Z

[Robert Biegler]

Your diagram describes something not physically possible. You have negative drag. If that existed, we could build a perpetuum mobile. Imagine your sail were mounted on a wind turbine hub, in a dead calm. Assume that the vertical line that shows wind direction is the plane of rotation. Push the rig with your finger to set it moving even a little bit. The moment that movement produces a little apparent wind, a rig with the characteristics you have drawn would have a force vector with a forward component that would make it move further into the wind. There would be a sideways component as well, but the forward component would move the rotor, in a dead calm, only because of the apparent wind created by rotation, even if you take some power off with a generator.

Or you could put the rig on a boat. Draw the centreline parallel to the wind direction to represent what happens if, in a dead calm, you push the boat a little to create apparent wind. The negative drag, the windward component of the sail thrust you have drawn, would keep the boat moving forward forever simply from the apparent wind.

You need to redraw that so that angle A + B is larger than 90 degrees.

To start by answering your question, my ill-chosen "windward thrust" is labelled "Thrust Vector" in this diagram, which is the total sail vector as transmitted to the mast. I've also heard it call 'drive'.

It is called drive because the angle between the sail's thrust and the boat's direction of motion is less than 90 degrees, not the angle between the sail's thrust and where the wind comes from.

Hmm... I would see lateral deflection increasing the angle A+B

By "lateral" I meant the direction of the sail's thrust stays the same. The angle A+B would not change. "Lateral" was meant to distinguish from the rig being pulled back if you attached the sheet to the plywood base. Forget "lateral". It doesn't help.

but not the magnitude (as indicated by length of vector, or distance of heel from center). If more force is produced, the magnitude would increase. A combination would do both.

If the sail produces exactly the same force with the two vang settings, but one vang setting raises the centre of effort and so the lever arm, then the torque (heeling moment) increases because it is the product of force and lever arm. And how far the heel is from the centre measures torque, not force. Imagine you made this rig 2 metres tall, and that the gimbal is at a height of 50cm above the base. If you push at the mast top, 1.5 metres above the gimbal, then you will deflect the rig twice as far as if you had pushed with the same force at a point 75cm above the gimbal.

In other words, distance of the mast heel from center indicates summed magnitude of forces transmitted to the mast,

Torque, not force, and torque is the product of force and lever arm, which is why you don't know force without knowing lever arm. And you don't know the lever arm.

while angle from the wind indicates the summed direction of forces.

Yes. And the angle between sail's thrust and a direction 90 degrees to the wind direction is the drag angle. The cotangent of the drag angle is the lift/drag ratio, so you can calculate the lift/drag ratio from the angle of deflection relative to the wind.

RE Lift/Drag Ratio: As I understand it, sail L/D is an aspect that doesn't get measured by SPAR, except indirectly as it impacts magnitude and direction. But I'll try to read up on it.

See above.

[Unfortunately, the equal deflection contour won't be a circular radius, but rather a rounded square?]

An equal deflection (from the centre, presumably) contour would be by definition circular. Do you mean the deflection contour for equal force in different directions? Yes, you would need to know that.

Regards

Robert

[Dave Z]

Hi Robert,

Thanks for taking me to school... I've now done some actual homework, and see and agree with all your points. I'll mull this over to see if anything useful is left.

Thanks, too, for your patience and clear explanations. I appreciate the thought and careful analysis you've brought to this proposal!

Dave Z