"What caught my eye this time was the cambers that were used in what was intended to be a matched comparison. If we regard the SJR as two separate sails, each of them with 8% camber – then we might think it is a fair comparison with a contiguous sail which has 8% camber.

(In my SJR I put 10% camber into the jibs and 8% into the mains (more or less) – this might be too much for a sail which is tended to sail very fast, but it seems to suit my boat well enough.)

However, I am not convinced that Karl’s assumption is valid. If we see the SJR as a composite single sail, which has a slot in it (which is the way I see it) then the SJR model looks to have less camber than the contiguous model."

I had to think hard about your comments/observations Graeme!
To be brief, and coming from a Bermudan sail background, the two sails of the SJR are just that, two sails that are locked into a relative position to each other with an optimized mast position. What they are, and never will never be, what a Contiguous sail is to my thinking because they do what a two sail example does, like  Bermudan and what a single element sail never can in the way of air flow effects on each other as we read in the article David T suggested we read and I quote from it

"6.1 Effect of Jib on the Mainsail
The major effects of the jib on the mainsail are as
follows:
1. The jib causes the stagnation point on the mainsail to
shift around toward the leading edge of the main (the
header effect).
2. As a result of the above, the peak suction velocities on
the lee side of the main from the main leading edge to
the area of the jib trailing edge are greatly reduced.
Since the peak suction velocities are reduced, the
recovery adverse pressures are also lower.
3. Because of the reduced pressure gradients on the
mainsail, the chances of the boundary layer separating
and the airfoil stalling are reduced.
4. The mainsail can be operated efficiently at higher
angles of attack without flow separation and stalling
than would be the case with the mainsail alone."

The next point is, your suggested camber change for the SJR to "make it more fair" does not do the test justice I think. Why, because I see the test as simply being the best each sail format can do for equal conditions, wind strength and direction. Lets say you put both models at 45 degrees to the same wind strength, say the equivalent of 10knts.
As a bench mark, the Contiguous sail is set at 8% camber, what ever we think best for the condition but say 8 in this case because we assume it's best. It's true, that one of the models might work better with a different camber but that's OK because it's the target of the model maker to make each model perform the best it can  for the equal given conditions. But to some degree I rely on 8% at any scale working the same. If you changed one, you'd have to try changing the other and that was beyond the scope of my experiment.
Simply, the object was not to de-tune one model for the conditions to make them perform the same. Actually, going by two sail theory, the SJR may have even performed a little better if the mainsail (see, you can't talk about it as being one sail!) was maybe 2% less in daft and if so, I would not consider that unfair if it further boosted it's performance! You can never change the fact that both SJR sails operate differently because there's two of them. The fact that Slieve suggests 8% as being optimum for the jib, is why I used it and what would be unfair is to change that. It's my understanding that 8% for a single sail is very common for a Junk Rig  and so that is why I chose it for the Contiguous sail. What would be unfair would be to try and use the sails out of their range and I do admit, one sail or the other might do better in extremely light winds and one maybe better in the stronger conditions, bit I wasn't testing for that but my intuition (which could be very wrong!) tells me the Contiguous might be the better of the two in very light conditions, it would be interesting to know!

Anonymous wrote:

Hi Karl,

Congratulations to you for publishing your video and showing what can be done with simple tools and some fore thought. It it this sort of effort that encourages others to look more deeply at their rigs and which move things forward. I particularly like the way you explain your reasoning and do not quote it as firm measurements and firm conclusions as readers can be misled if given performance figures to the 0.1% accuracy when we can really only make guestimates.

Well done.

But please do not take what I write next as a criticism, but more of an explanation. Karl’s experiments are using panels with are not based on straight battens but on airfoil shaped ribs so really are testing the airfoil shape at the mid height of the panel, and not the probably reduced effect over the whole or average height of the panel. This is not a bad thing, but must be understood when trying to evaluate the effect. Another consideration is that the flow in a rig is not always horizontal, and is much effected by the tip losses, under the bottom batten and over the top of the rig. Where ‘under the bottom’ may be fairly constant for most rigs the ‘over the top’ effect can be very great.

Tests on any rig cannot be considered complete until tip loss is included. I can quote personal experience of this from the ‘50’s to the ‘70’s, and in his excellent writings Tony Marchaj also indicates the importance of tip losses and induced drag. From this background I believe a long yard at an angle of 30deg to the horizontal is a ‘best guess’ starting point, and works well with a high balanced rig. From feedback it appears that those who deviate from this are less impressed with the performance.

I keep trying to produce a new up to date write up of the SJR, but it keeps getting pushed to the back burner when other things get in the way. As for rig balance, the recommendation for Centre of Area hasn’t changed, but mast position of 27 to 35% jib luff to main leech for the parallelogram panels is now suggested to suit the accommodation. 33% is still preferred, but not considered to be that critical.

Karl, your feeling that the lift/drag ratio is high with the SJR ties in with the impression I have from sailing the full size rig. In particular the windward heeling situation seems to indicate this to be the case.

One point I would like to stress and that is that the SJR is an attempt to make the simplest and cheapest rig with good performance that can be easily maintained. Straight battens, minimum spaghetti and tweeking lines make a simple to repair rig if in foreign parts.

I don’t want to get drawn into a lot of discussion here as I should be working on the write up, but I must repeat how good and useful experiments and reports like Karl’s video are. Well done.

Cheers, Slieve.

It's quite appropriate and welcome to hear from you on this matter Slieve!
No problem at all with you pointing out the shotcomings of the experiment because of what is excluding and that you find in real world rigs and there certainly was an aspect of testing the basics and treating all else as being equal when at the same time knowing it isn't which may either increase or decrease what I think I have observed. The only certain thing is I'm very interested in understand more about the differences in Junk Rigs and how they might be improved comparatively while still being called Junk Rigs and that's of course a very different line for different people.

I look forward to your write up you are working on, and also any suggestions as to how I might improve my tests, given the models don't become too complicated!

Anonymous wrote:

David T.:

I think it’s worth looking again at the aerodynamicist Arvel Gentry’s work, back in the  1970s, in which he lays out for the first time the aerodynamics involved in the interaction between sails. I’ve attached a copy of one of his papers.

Did anyone read the paper David attached? If not, do it!

It is written for sailors, not engineers, and explains the slot effect really well. Thanks for posting it, David!

Cheers, Paul

I just finished reading it and am satisfied the explanation is there for the results I may have found, and that need further tests and real world experience to verify. The Jib has a beneficial effect on the Main, it helps delay the separation of air from the Main and it allows the Main to "be operated efficiently at higher angles of attack without flow separation and stalling than would be the case with the mainsail alone" I think this directly relates to the SJR.

I saw less heel and very slightly more drive forward on the CL vector. I also thought I saw the flow on the Lee of the Main stay better attached which correlates, to my thinking, with the results of Arvel's summary of his work. If it all stacks up with further tests, the theory verifies what what works in practice and it's a step closer to a version of the Junk Rig being closer to the windward performance of a two sail Bermudan Rig. There are limitations though, modern Bermudan Rigs have lower drag because of lower drag masts and a 3D foil shape that is arguably also lower drag and far greater % of the sail is actually on the deepest draft camber and that last aspect makes me wonder what is lost because of that. The Aero Rig of course gets closer to the Bermudan in that respect but there is the added drag of all the wishbone battens.

But in the overall comparison, the downwind capability, it's lower tech/cost, the ease of use by a short handed crew(less fatigue), the lower loaded sheets etc of the Junk Rig certainly go a long way  make up for it's shortcomings to windward, which seems to be diminishing year by year. What's not to like about that given how much easier it is to sail a Junk Rigged boat!

I welcome feedback on my assumptions.

Anonymous wrote:

Karl, I’m not sure whether you know, but I’m 80, with Parkinson’s and other health issues that make it impossible for me to make such models and repeat your experiment. Much as I might wish it to be otherwise, all I can do is attempt to dredge up out of my memory what I’ve learned about aerodynamics, and suggest that though your experiment has demonstrated that the SJR has a little more drive than an unsplit JR, I don’t think that it’s for the reasons that you believe.

I think I demonstrated on Weaverbird that if you are prepared to put in the extra effort to make a wingsail, you get yet more drive; and with the current mast position, it’s what I would be doing - if I was 20 years younger and was building a rig for your boat. The mould that I made for the forward ends of the battens may still be in the custody of the current owner of Weaverbird, should you wish to give it a go.

Thanks for getting me up to speed with your situation David, for a person that has been so active and creative that can't be easy.
Let me ask a different question so we can try to get to the bottom of results I think I've found and although as you know I have some ideas of explanation, I don't have your experience, so what would you have me do for my next experiments using the same format so that we might further understand what's going on, if there's more to be found?

I've considered a couple other other iterations for what I've done, one of them being trying one or two hinged batten cambers and it would be fun to try and replicate your Weaver Bird Rig but that would be hard to do with rigid panels in the way of replicating both a hinged batten and sewn in camber combination but I could try and replicate the final camber you achieve as in the CAD drawing Paul has shown and sent to me.
I have another idea for the SJR test, and that is to add another split and see what that tells us. I might call it the 'Double Slit/Split' Experiment, what do you think and do you have any suggestions for the split percentages, I was thinking 33% 33% 33% or 25% 25% 50%. The main purpose would be to see if it tells anything more about what might being going on, it might reinforce one theory or another.

The aspect of Karl’s experiment that most caught my eye was the difference in heel angle

between the two models.

Surely if one rig is heeling less than another it is generating less weather helm, meaning

less correction with her rudder, meaning MORE SPEED ?