I think David is 100% correct in pointing to the Weaverbird soft wing junk sail. And the SJR is indeed not so tolerant of sloppy helmsmanship – mine isn't anyway - and I should know as I am not much of a helmsman.

The testing a few years ago did seem to suggest that the best performer with the best alpha tolerance might be the soft wingsail, and balance-wise the Weaverbird planform would undoubtedly be the closest match to the hoped-for Ljungstrom rig, possibly the only one which could really fit that mast position – and quite possibly the best performing junk sail.

The only reason the soft wing sail was left out of the discussion was because Karl was talking at the time about wanting to get set up as quickly as possible, with the simplest solution, should the Ljungstrom sail not meet expectations.

However, for someone with the technical skills and ingenuity that Karl has shown, a Weaverbird sail could well be the shot – what about it Karl?  It would be the first such sail “down under” and we need to do something a bit more impressive on behalf of the junkies, than recent performances at the “Tall Ships”! (The four of us took the last four places - strictly speaking, all DNF).

I still come back to Karl's experiment and the subject of this thread. It is all very well dismissing the "two leading edges" proposition - and it is doubtful - but where is the plausible explanation for the experimental result? The answer may be in the paper David refers to. That's a bit above my ability to really comprehend so I don't know. It did seem though, from the little bit of the experiment actually shown in the video, that those tell tales were flying quite nicely and maybe the jibs really were helping there. I have never really thought the so-called "slot effect" had much to do with the SJR, but maybe it does a bit.

Karl's experiment does not demonstrate what I always thought might be an advantage of SJR and that is the entry shape of the foil - never crushed and always much less interfered with by the mast, compared with the contiguous sail. The key to the SJR jib is that you can make a really good entry shape for the foil. That's where the soft wing stands out also, maybe even better.

Balance wise, the weaverbird sail would probably be the best fit to that mast position and underwater profile, and the SJR probably the worst. (A little mizzen sail would fix that problem though).

Cheers Graeme and before I move on to my answer, I'd like to acknowledge all the emails you have sent me with advise and answers to my questions over the last week or two and while I'm thanking people, Arne has also been of great help to me and tried to steer me towards the light but I can be a bit stubborn and for that matter, Paul Thompson and I go back a long way as friends and has been a great help and may even build my next sail. I helmed Paul's La Chica to a win over Zebedee a few years ago and after that day, seeing the highs and lows people experienced, you can't tell me Junk Riggers don't like going fast, as long as it's fun of course!

 Yes Graeme, I actually had way too much footage of the tests and was a bit overwhelmed by it all along with trying to build a yacht so I decided to pretty much voice over some of the content and yes, I would have made the video differently for the JRA audience  but it was mainly aimed at my yacht build community on YouTube who I think have experienced a bit of a roller coaster ride (tell me about it) as I go back and forth between the Ljungstrom and a Junk Rig. What did surprise me was how many are supportable of me going to a Junk Rig and I get the feeling there are a growing number of people interested in what a Junk Rig can offer.

"I do find this improvised, low-tech type of testing very interesting, and believe much can be learned from it. Well done."

I do to and it probably wouldn't surprise you to know I've done a lot of comparative tests during the build! I even designed and built my own hygrometer (Paulownia timber moves a huge amount with small humidity changes) and before I built it, I didn't even know the word. It was useful and worked great!

Well this is a surprise and thanks for posting the video Pete, I was thinking about it and have never posted on the JRA website before.

I've just had another hard day on the Ljunstrom and it's made it another day closer to a Junk Rig! I do think they have their merits but do think I've pushed the envelope too hard, the best part so far are the bearings I cast, machined and made which can be seen in earlier videos. The sails are truely impressive when they open but it's all still on the trailer.... My last words on the Ljungstrom are, Arne indulged me some years ago, with some back and forward emails on my Ljungstrom intentions and I really wish I'd made the sail to the profile he suggested, one very similar to the one he made for the large dinghy and which seemed to work so well. When I look up at my sail, putting stiffer and stiffer battens in, it just gives the appearance of wanting to hang like an Arne sail!

I would like to respond to some of the comments but start by saying as I said in the video, I just did the scaled experiment to help me decide on a Junk Rig for my yacht I'm nearly finished building and also wanted to understand what appeared to be the differences I observed and I also said what I found may not scale and I also said, don't take my results seriously, until someone else repeats/peer reviews the experiment.

The 8% camber on all panels has been raised. I did this partly because I think I'm correct in understanding Slieve saw 8% as possibly being better for his Jib. For the SJR Main I made it 8% because it's a rigid sail model when in reality, for my boat, I'd probably make the main 6%, thinking in reality, 6% would probably become 8%, as they do.Then it became clear, to make it a fair comparison, I should make the Contiguous camber 8%, all with 37% maximum draft, once again, trying to keep things equal.

The split width has been mentioned and I didn't apply much science to that, it 'looked'  about what Slieve recommends so I may not be doing his sail set up, justice.

The 8% Sheeting angle is an interesting one and I think confirmed by what I've briefly studied (because of a comment on YoutTube) in the last 48hrs in the way of multi element wings. One thing I noticed with many of them is you can take a fair line from the exit of the wing 'ahead' and it will run into the camber of the wing element behind it, even if there's more than two elements. So I believe the sheeting angle for the Jib is OK and I was noticing, as shown in the video, a good flow through the split, onto the Main with no separation observed. This is speculative and I briefly noted it and looked no further at it because I didn't test it properly, but the telltale seemed to be hugging the SJR Main a little closer than the on the Contiguous Sail, make what you will of that and what it 'might' mean to the lift drag ratio and the results I may have seen.

If someone repeats this experiment and I hope someone does, note that I made all the camber profiles the same, same entry angles and same 40% straight exit. I made the camber close to what Slieve saw/see's as the optimum (for a practical sail build in flexible materials) and I simply 2D scaled the camber for all sail elements. I'm open to suggests as to how I could have done this better! I'm happy to pass on the CAD files for the sail drafts cambers and sheeting angles if anyone want's them to make a model and compare to mine.

The Aspect Ratio aspect has been commented on. In trying to explain the results I 'think' I found and I started to think about it like this. I don't think you can literally compare aspect ratios literally between a SJR and a Contiguous sail plan. Why? Because a SJR literally and relatively has to high aspect sails. What is similar to that? A Bermudan Rig and they go well to windward. Connection there?
As well as the length of leading edge difference I mentioned earlier, there are many examples that two or more high aspect foil elements working together, have a greater lift effect than a single element in 'some', maybe not all, circumstances. Plane wings at low speed use this effect to be seen when you look out the window on take off and landing. Formula 1 down-force foils have multiple elements because they work better than one. Maybe they have more drag but this may be offset by greater lift? Not everything is transferable of course! I'm just saying there might be something in this, given the results, I think I observed, and that don' t mean much until someone else repeats them.

To finish, I really appreciate the feedback and discussion my video has created and to be honest, it's given me an appetite for further model testing! Maybe that will lead to something like some testing on two lasers, my brother won't notice if his goes missing for awhile and surely my friend Paul Thompson could make me two sails of equal area and CE position, without biasing the construction in favour of the Contiguous Rig!? It's a serious question, how would you make the SJR sail and the Contiguous one, so it's a fair comparison?

To finish, I totally get it that there's much more to choosing a Junk Rig than the issues I've commented on. Things like your mast position, who makes the sail, what is easier and more reliable to sail etc etc.

I thought the small scale model test which Karl has done is very interesting.

I have to respond to what Arne has written, however.      I have the highest respect for Arne, but I think here he has missed the point a little bit.

First of all, I think Karl was being very tentative in any "conclusions" -  rather more that he was seeking an explanation for these surprising results. The explanation he put forward as a possibility (twice the amount of leading edge on the split sail, compared with the same size contiguous sail) was a suggestion – an interesting suggestion, which may prove to be false -  but worth consideration. It is generally regarded as true that most of the useful lift in a sail which is close hauled, comes from the first part of the foil, and that the entry is important. It is quite possible that the split sail provides a cleaner entry and more leading edge – two of them in fact. Whether this leads to better performance is still open to question.

As for an explanation of the experimental results – if the only difference between the two model sails is that one is split and the other is not, then we can be excused for surmising that the split might be contributing in some way.

That is not a conclusion, but it is a hypothesis worth considering and perhaps disproving if that is possible. (A good scientist will always try to disprove his/her own hypothesis, and encourage others to replicate the test and try also to disprove. I think Karl is open to that).

The point I am trying to make is that Arne has not disproved anything, he has merely ignored the result of the experiment.

I will give an example – Arne says that size and aspect ratio matter, and that planform has a big influence. He is probably right. So what? Karl has compared two different sails which have the same size and the same aspect ratio, and the same planform - and measured a difference in behaviour. The comparison between these two otherwise identical models remains to be explained.

Where I agree with Arne are in these three points: (1) More testing needs to be done before we could draw firm conclusions from these interesting scale model tests, (2) a low balance/high AR Johanna sail may be a very good solution for this boat in regard to the mast position, though I am not sure if it is still too far forward of that centreboard. (A high-balance SJR would be the least suitable, I think, unless the mast position be shifted aft, or a mizzen added, neither of which Karl wants to do. If it were me I might consider the extreme low balance planform of the fanned sail which Paul Thompson put on Freebie - and even then, a small mizzen might be needed. That centreboard is placed very far aft. But all that stuff is a dark art to me, I think Arne would have much better judgement than me)  and (3) having put so much effort into his Ljungstrom sail and rotating mast, it would be good if Karl would persevere with this and get it working properly, because, at the very least, he can get some good data from his hull, which would be useful if a junk rig is contemplated later.  (I am aware too that Arne has experience and practical knowledge of the Ljungstrom rig, so his advice on that is worth considering).

However, the rest of Arne’s post, true and wise as it may well be, is not really relevant, as far as I can see, because it throws no light on why these two matched models behaved differently under the test conditions.

That (I would have thought) is what needs to be addressed.

(By the way, a small point of detail - I think the split in the model sail is not quite right. I think (not sure but I think) the jibs on a SJR should have a little more camber than the mains, or the jib sheeting angle needs to be increased slightly on that model, in which case I think it would perform even better.)

SJR versus contiguous? I think Karl "nailed" that question with a pertinent counter-question: - "What type of car do you buy?"

I have watched most of Karl’s video.

I don’t think that his findings from his little wind-tunnel test of small scale models is enough to draw any conclusion:

• ·         When sailing a real boat fully close-hauled, say at 30-35° from the apparent wind, the main contributor to heeling the boat is actually the sail’s lift. The drag of the sail mainly slows the boat down. When heading closer and closer to the wind, the driving component of the lift rapidly shrinks until the boat stops.
• ·         Karl mentions that the slotted sail will have twice as long leading edge. Sorry, that doesn’t help. What matters is how much the total sail deflects the incoming wind. Big deflection angle gives much lift at the cost of relatively more drag. Little deflection (less camber or angle of attack) gives less lift and even less drag. The result is better lift/drag, which is useful for upwind work.
• ·         Size and aspect ratio matter. The total planform of any sail (and wing) has a big influence of the L/D ratio. Taller sail, higher L/D.

A few days ago, Karl sent me the shown JR sailplan and wanted me to have a look. It appears to be based on a Johanna 70 master sail with AR=2.10. This sail shows how much area and luff length this sailplan can produce for a given mast length. I am biased, of course, but I think that looks good.
However, Karl has already fitted a Ljungström rig, which is almost ready for test-sailing. I suggest he does some test-sailing with this rig first, and uses its performance as a reference if he later tries a JR. My hunch is that the Lj-rig is second to none for upwind work.

Arne

(Full size diagram at Arne's sketches, section 9-17...)