Graeme K Wote:

I question whether it is valid to regard a sjr as two separate sails, and I question whether the two models are truly equivalent in camber. This leads to a question about the experimental apparatus itself.

I may be quite wrong, but it seems to me that a Split Junk Rig is effectively acting as two sails, mimicking to some extent the interaction between a Bermudan mainsail and a non-overlapping headsail, as described in Arvel's paper. Here's another paper that makes a similar point—I'll quote a relevant paragraph and include the link.

Regarding your question about cambers, my experiment suggests that I wouldn't be starting from a bad place by building the bottom panels of a Split Junk Rig with 8% camber on both sails and an 8deg sheeting angle. Interestingly, many Contiguous Junk Rigs are also thought to be around 8% camber, so that's encouraging.

I'm still very open to suggestions and alternative ideas, but the experiment produced an indicative result that I didn't expect. At the moment, I also have a working theory that explains what I observed, although I'm happy to revise it if better evidence comes along. My initial attempts to explain the results were off the mark, but that's all part of the learning curve, even if it has been a little time-consuming!

"The interaction phenomenon causes two sails (the jib and the main sail) working together in an upwind condition to produce a larger lift force than the sum of lifts generated on sails operating alone. In spite of the fact that this phenomenon is widely known to both sailors and sail designers, its cause is still arguable."

Research Gate

Karl, I would like to de-romantisize those "miracles" you mentioned. They are well understood!

Maybe in trying to understand the mysteries of why a Junk sail works, there's a lesson or two in nature. A friend who has a very high end trailer race yacht, now sails with 'bumps' on the leading edge of his rudder, bumps like you see on a humpback whales pectoral fins.

These bumps on the humpback whales fins help to prevent stalling at high angle of attack. That's very useful if you need high lift at high AoA's and can live with high drag in return - which is the case for whale fins in sharp manouevres or rudders at high angles.

--> Different situation than in a sail!

Then there is the gold ball which works better because of the dimples!

Also, this example is being teached in the first semesters of aerodynamics. The dimples transfer energy into the boundary layer, shifting it earlier from laminar to turbulent flow. A turbulent boundary layer is less prone to separating (see David's paper), thus reducing the air resistance of the ball. No rocket science at all...

And we can't forget the Bumble Bee with it's flattish wings that technically can't fly at all!

That's an urban legend! It results of a missunderstanding, that somewhen in the sixties someone compared the wing area of a bumble bee with its weight and draw the conclusion that they are too heavy. That calculation was made with fixed wings - just like airplanes. However, that's not how a bumble bee flies. It flies by constantly moving its wings in a way that creates lift vortexes with every stroke, which creates way more lift than any fixed wing could ever hope to. It is a very different, powerful way of generating lift - but not at all useful for gliding (which is what we do with our sails while sailing).

Should we make Junk sails with even more imperfect surfaces to keep the air flow attached?

Sounds interesting, but: what you describe is basically the golf ball principle. It is a possible way to reduce drag/resistance by energizing the boundary layer. Thus, higher AoA's could be realized before the sail stalls. That would need some proper design and layout. Shooting from my hips, I'm pretty sure that the boundary layer of a sail in any decent seastate is fully turbulent...

Cheers

Paul

It turns out (→ page 4 of the article) that a tool has long existed that would allow us to dispel many of our doubts. All we would have to do is shell out a few bucks, buy some conductive paper in Amazon , and take it to the nearest technology museum that has such a tool. It would certainly be cheaper than renting a supercomputer or a wind tunnel in Southampton.

Karl,

that paper of David is quite old. The mentioned potential flow simulations nowadays run on a standard laptop in a few seconds ;-) I'm still looking for capacities to continue my CFD simulations of the junk rig... Maybe this winter.

Best

Paul

Anonymous wrote:

Arvel Gentry’s articles are very interesting. However, I am not so sure that his findings about the slot effect apply to the SJR. On all the diagrams Mr. Gentry has drawn, the jib is as big (wide) as the mainsail, and the overlap is about 50%. The SJR is more like a mainsail with a smallish working jib, with no overlap at all. There still must be some interaction between the jiblets and mainlets, but probably not as much as in Mr Gentry’s papers.

As for the contiguous, cambered JR, like Ingeborg’s sail, Graeme recently described it as (from memory) “the sail that should not work, but which still does work”. I could not agree more. What happens behind my sails is a mystery. The leading edge gives anything but a smooth entry, and the rest of the sail is full of bumps from the mast, mast lift and lazyjacks. Somehow, the airstream still leaves the sail in a tidy order, with the telltales at the leech flying (as long as I don’t over-sheet the sail).

• ·         Area for area, my junkrig may deliver a little less drive, with the moderate camber of 8%. I compensate by adding more sail area (same as mainsail and Genoa 1).
• ·         The lift/drag must be quite good, as we point and sail within a degree or two (either way) to the pointy rigs.
• ·         The CP, centre of pressure, sits roughly in the same place as in the original Bermuda rigs. If I install the rig with the CE at the same place, or a little aft of that of the original rig, helm balance will be good.

Conclusion:
I am still able to sleep well without understanding everything...

Arne

..The sail that should not work...( full size photo in Photo Section 8-20)

I like that title " The Sail that should not work" and agree with your take on the limitations of the Arvel's observations. It's also pretty 'awesome' you get the comparative performance you do and surely the small % you don't get upwind is more than made up for on the downwind and you'll be carrying less crew weight as well, although you need at least one crew to make you a cup of tea because there's not much else to do but drink tea while racing  with a Junk Rig!
Maybe in trying to understand the mysteries of why a Junk sail works, there's a lesson or two in nature. A friend who has a very high end trailer race yacht, now sails with 'bumps' on the leading edge of his rudder, bumps like you see on a humpback whales pectoral fins. Then there is the gold ball which works better because of the dimples! And we can't forget the Bumble Bee with it's flattish wings that technically can't fly at all!Should we make Junk sails with even more imperfect surfaces to keep the air flow attached?
I like your point and observation about what you see happening at the Leech, a clean air exit tells a good story about the rest the sail ahead of it.
All this reminds me of what I hope I recall accurately, that Annie H said to me once years ago when I was discussing a JR idea for performance and that was, she wasn't bothered with the 95% effort to get the last 5% of performance gain and that made sense to me. If one was bothered and well resourced, I do think you could close that gap even further Arne, a mast with less drag, lines inside the mast etc etc but I like the 'culture' that accepts quite acceptable results for relatively minimal input and the fun that goes with it.
Good to hear these matters don't disturb your sleep!

Arne,

I also find it quite unintuitive, that the cambered JR draws well (or even better) on the "wrong" side, though I can confirm the same on Ilvy. It would be interesting, if the effect varies with lower/higher mast balance.

My educated guess would be, that the airflow separates in front of the mast and reattaches aft of the mast, creating a "virtual" camber and even reducing the drag of the mast. But it is only speculation up to now...

I like that it works how it works! I would also very much like to know why it works how it works.

Cheers

Paul