The absence of test data for sails

[amolitor]

Perhaps we're seeing some sort of language difficulty here, Ivor, but saying things like:

"I can conceive of no way that an un-battened sail can ever have attached flow on its leeward side."

doesn't look very good. The next paragraph of your remarks seems to be aimed at dismissing people who disagree with your claim that sails are always stalled. This is also an unpromising attitude, being just the sort of thing crazy people say. But, again, perhaps we're just having trouble with terminology! It's accepted as fact, and there are many pictures to go with it, that the normal state of affairs for a working sail is attached turbulent flow on the leeward side.

Perhaps you're considering 'attached turbulent flow' or 'turbulent flow attached for a while, which then separates' as 'stalled?'

[amolitor]

Another point to mention, your memo on the deplorable lack of test data for sails (linked in your first posting) mentions that one could test solid curved sheets. Indeed, you allow as how it has probably been done.

Again, Marchaj. He absolutely did this, and has lots of lovely pictures, graphs, and data from those tests. It's a good idea, and it has been done! I am surprised that you didn't know it, however. Marchaj's book Aero-Hydrodynamics of Sailing covers, as far as I can see, all this material and more. It is backed up by decades of experimental work, with charts, graphs and figures.

[tom28571]

Originally Posted by ivor Bittle

I cannot take on board all the comments to this thread. Suffice to say that they have added to my working knowledge and I am grateful.

Ivor Bittle

In his first note Tom 28571 makes several pertinent points and then says that he was off to do something useful. His second is a powerful observation of the current scene but made me wonder where he placed the astonishing performance of windjammers.

.

There was nothing profound in the first sentence. Just my wife calling me to do something that I had promised earlier. At least she considered it more useful than my spending time on the internet

I am not well versed in the performance of windjammers or even sure exactly what defines a "windjammer". If we are talking about clipper ships then I'd guess the fine speed records might illustrate that it is power to weight ratio, a decent hull and well chosen sailing routes that avoid windward courses when possible. Others that call themselves windjammers today don't seem all that spectacular to me but I admit to a lack of direct knowledge. The record passage of "Atlantic" having been burned by smaller boats should temper astonishment a bit although it was astonishing for its time

I do consider that analog thinking can be just a powerful a tool in understanding as laboratory data in phenomena with as many (often unknown) variables as sailing. More learning should probably lead to greater humility as we find more and more that we do not know.

Ivor makes a good observation when he says that it is very useful to go back to basic knowns and avoid presumed "advances" based on faulty assumptions.

Now, I have no idea what Ivor means by no attached flow, ever, on the lee side of a soft sail. Most probably we have, to paraphrase Strother Martin in "Cool Hand Luke", a failure to communicate.

[tspeer]

Many of your comments are true. I liken many explanations of sail aerodynamics to "Just So Stories". They make no more sense than "How the Leopard Got His Spots". And so many of them can be seen to be false if people simply applied them to their every day experience. As to why they continue to be spread, the answer is because people really do not pay any attention to them when actually sailing.

At the same time, those who do have the data do not make them available for proprietary and parochial reasons. For example, all those who are members of US Sailing contribute to funding technical activities, such as wind tunnel tests of sails. I once asked the gentleman in charge of US Sailing's technical activities why the data were not made publicly available, or at least available to members of US Sailing since we were the ones that funded it. The answer I got was, "We do the tests to improve our handicapping systems. If we made the data available, designers would use the data to defeat the handicapping." I was stunned. Would it not be far better to allow people to create faster and safer boats? Which would provide a greater benefit to US Sailing's members? (Of course, recent events have only made even more abundantly clear that US Sailing does not exist for the benefit of its members.)

Getting reliable data is not easy and is tedious. Most people are simply not interested in doing it. The advent of modern digital instrumentation is making it a lot easier. One can now buy whole systems off the shelf that can transmit their data to a laptop computer, making for a turn-key data acquisition system. Previously, the ability to record yacht performance data in this manner was previously only available using custom-built electronics.

It is not enough to measure what happens with a single configuration, although that is typically where most people's interest begins and ends. In order to be truly useful, parametric studies should be done, covering a wide range of design options so that the data may be more generally applied. There are a few examples where that has been done - the Delft series of hull shapes being an outstanding example. But more needs to be done.

One possible avenue is to work closely with universities that have engineering departments. Undergraduate students in particular have very tight constraints with regard to the time to do a project and the resources that are available. However, it would be possible to lay out a long-term plan of research in a particular area, and have students peck away at small pieces of it. That way, each piece is part of a coherent whole, and over time it adds up to a significant accomplishment.

[CT 249]

There are many classes where sail shape is very little constrained by racing rules. These include;

International Canoes;
NS14s
MG14s
12, 14, 16 and 18 Foot Skiffs
Moths
Windsurfers
ORMA 60s
Open 60s

It could also be said that

[ivor Bittle]

This thread has provoked a greater response than I thought it would. I wondered how to respond to the recent comments.

Amolitor is obviously cross with me for not liking Marchaj. I do not suppose that he is alone in this but I am unrepentant. Some of our difficulty lies in words but it also lies in the diagrams that I mentioned in my last memo. There we have three diagrams that purport to be a representation of the flow over two sails as in a Bermudian rig. They are out-and-out guesses just as someone who does not understand flow patterns might scribble on a blackboard. This is typical of those who cannot be bothered to do things properly or think that any old sketch will do. Whatever the message is in these sketches it will have to be undone and replaced at some time. Why start with it all wrong? Do not risk looking at the text that goes with it.

The question of terminology is also very lax. I know that sails were there before wings but wings had the scientific treatment before sails. Osborne Reynolds first ran into this problem of descriptive words with his classical demonstration of the two modes of flow of water in a pipe. He showed that water could move with laminar flow and that this flow could break down into a flow with tiny disturbances moving randomly throughout the water. He called this turbulent flow.

Subsequently it was found that the flow in rivers is always turbulent, in Reynolds’ sense of the word, so we have to get used to the idea that even when the surface is glass-like there is internal random motion. That same river might flow under a bridge and, in doing so, have to flow round piers. If you go and look at the flow it becomes clear that eddies are shed from the piers and that surface waves show much larger eddies with these small eddies in them and then there is turbulent flow within that. The word turbulent becomes meaningless if it is used for all three. We are short of descriptive words.

When it comes to sails the flow is like that round a bridge pier and not like that over an aerofoil that is working in its proper range of angle of attack. Turbulent is not adequate.

The explanation of the behaviour of sails requires a terminology and it is a matter of interest to think what that might be. As a coherent terminology is available for aerofoils and the sail is just an aerofoil that is working when fully stalled like a rogallo wing it is an obvious source of useful and well-established terminology.

In that terminology the stall occurs when the flow over the upper surface of an aerofoil breaks away from the smooth flow in contact with the upper surface and one, or more, large eddies form between the separated flow and the upper surface of the wing. When the angle of attack is reduced to allow the flow to return to “normal” it is said to go through re-attachment.

If, like Stanford, your opening diagram shows attached flow on the leeward side of the sail life is going to be difficult.

I found the contribution from Tom Speer most illuminating especially the Sail America story. It goes to show how lucky we were that NACA published its data for all of us to use.

I tried to contact a university but I could not understand the complex titles that they all had. I emailed directly to one to suggest a student project but got no reply. When I look at their publications my heart falls.

I was interested in your comments on analogue and computer methods. I am all for the use of computer modelling but somewhere it must be linked with reality. I was impressed when I was reading about using a computer programme as a wind tunnel to narrow down the work required in a real wind tunnel to solve a troublesome problem. Eventually the outcome will appear on an aeroplane and the circle will be closed and the computer modelling will become even better. Is that possible for sails? I do not think so and Petros points to the money to confirm this view.

This thread has been useful to me to feed into my original paper and I want to see whether I can articulate the idea of a framework into which data might be placed in line with Tom Speer’s penultimate paragraph. It will be difficult.

Ivor Bittle

[amolitor]

I'm not cross with you, Ivor. I just think you may be seeing untrammeled wilderness when there is a moderately well worn path slightly to the left and not quite in sight.

[ancient kayaker]

While not an experienced sailing man a few ideas popped into my mind while catching up on this excellent thread which might prompt a thought or two.

Sail design must take into account issues such as available crew size, type of seas and winds in intended sailing area, reefing, maybe trailering, that do not beset aircraft designers. A sail's operating conditions cannot be predetermined precisely unlike a wing, where speed, loading and angle of attack are set by the designer. Therefore, I have some doubts that wind tunnel testing and theory will translate into better sails, there are so many variables, but more strength to you if you can tame this monster!

A lot of stuff like the end plates or wing tips now seen on "advanced" aircraft appeared in model aircraft about fifty years ago. Some of those ideas were never used on full size planes, like turbulator wires strung ahead of the leading edge which was a fad at about the same time. Maybe we need to take a look at what the model sail boat designers are doing these days for inspiration. Problem is, with sails, to a greater degree than wings, size influences design since wind speed and direction refuses to adapt to boat size and course.

Some of the "old stuff" was pretty good. The lateen sails on arab dhows which were reputed to be fast. My next sail set is to be based on the lateen concept, as theoretically I can put up more sail on less boat. Sure, I use a bit of math to predict sail force and heeling moment and I often build a model and test it. My wind tunnel consists of an electric fan and a long length of eavestrough or a large baking pan, depending on what I am testing for. Ultimately though, I have to (and want to) sail it and see.

[ivor Bittle]

I forgot about windjammers. I read a book on them same time ago and they were made of steel and 300 feet long. This would give a hull speed in the low 20's knots. They mainly operated between Europe and S America and logs of these boats showed daily runs of 400 miles. The sails were trimmed with steam winches. It seems likely that they were the largest working boats driven by the wind. There were lots of them. They seem to me to be the upper limit of sail driven boats as things stand now.

Ivor Bittle