Hi Arne I have probably already gone on about it too much, and I am sorry to keep on, but I must comment on your SJR suggestions, because I am not sure if you are correct there, and I think not.

You may be right on the one hand by saying that these are "other factors" which may impede the sails ability to weathercock, but I do not think they were the important factors in the case you are are referring to.

I THINK the boat you were referring to was one of the "coin" series - a "farthing" if I recall, the owner was Bert. The entire conversation can still be found many pages back, near the start of this thread and I have gone back and checked the following: There were two problems (1) lee helm (2) a refusal of the sail to "weathercock". A pretty bad combination. A lot of theories and suggestions were floating around, but Slieve got to the heart of (2) by diagnosing the balance problem as a misunderstanding of his way of calculating balance, and ending up with too much sail area in the jibs. Slieve wrote at the time "Looking at the photo in Bert's photo album it is difficult to see how much camber has been built into the rig, but by taking simple horizontal measurements from the photo it would appear that in the split panels the jibs have 33% of the total area but have been moved forward by the width of the wide slot so that the mast/ pivot appears to be at over 38% of the total jib luff to main leech chord. It is not surprising that this is unstable. It would be interesting to see a copy of the rig and hull drawings to see exactly what is happening. " Bert had thought that his balance was 33% but admitted in a later post that when doing the calculation properly it was 37% so if Bert was correct about that, then Slieve wasn't far wrong. 

In this thread Slieve also made a lengthy post in which he outlined how he had intended "balance" to be calculated in regard to SJR. If you take the JRA definition of balance literally, and calculate a 33% balance purely on actual sail area, you will get more area in the jibs than Slieve intended, and to make matters worse, those large jibs are then placed further forward, by the width of the slot. 

Bert said at the time that his jib sheeting angle was 8 degrees, which is probably inefficiently low, but I doubt if that was any part of the problem, I think Poppy's was about 8% if I recall. (Slieve advised me later to use 12 degrees as they did successfully on Amiina's  second iteration, and it is possible that more might be even better. But 8 degrees should not have been the problem.) Nor IMHO was the problem primarily caused by the mains luffing before the jibs did, as I will explain later. Perhaps these things might exacerbate the problem, in that way you might be right by raising it.

As a follow up, Bert posted later and said that he had made a complete new set of jibs, said that his new jib sheeting angle was 11% (I presume he meant 11 degrees) that he had reduced the area of his jibs by 2 sq m (probably more of a reduction than needed) but anyway he was able to report "Balance is right now, no leehelm anymore, no backwinding of the main. Jib sheeting angle is 11°. " And that is where it ended.

In my SJR (33% lower panel chord balance) even with 12 degrees of sheeting angle and the jibs having more camber than the mains (deliberate), the jibs still seem to back-wind the mains slightly, at times. And close on the wind, the mains always seem to collapse a little at the luffs, just as the tell-tales on the jibs float up and indicate the jibs are in the sweet spot. Obviously something not quite right and maybe backwinding - but anyway it all works well and despite the mains seeming to luff before the jibs do, the sail always weathercocks reliably when I let fly the sheets, which is why I am sure that in Bert's case camber and sheeting angle were not the main problem and that it was mainly a miscalculation of how much area to give the jibs (with probably other little things adding to the mix).

It  must have been a bad experience for Bert and it is greatly to his credit that he was willing to share it openly and let us all learn from it.

Arne, just out of interest, I would be curious to know what you think the maximum tolerable "balance" would be in the case of a contiguous (unsplit) sail, and what would happen if the limit is exceeded. Is it an issue of weather-cocking, or difficulty in getting the sail to set nicely, or simply a matter of mast too near the point of maximum camber and affecting windward performance - or some other problem?

And if I am wrong about the boat, and you were referring to some other SJR problem, I would be very keen to know about it.

The more this subject is discussed and theorized the more confusion seems to be generated.

Oscar is sort of on the right track in his conclusions, and I think what he is trying to say is correct, but Oscar has made a couple of statements which add to the confusion and I think need to be corrected. Oscar wrote: If de-powering the sail is done only ever before entering harbour and starting the engine, then I guess it's not a huge concern to not be able to completely weathercock the sail but if you were to sail engineless you will need to be able to de-power the sails completely in most situations. 

At all times (except, of course, in a flat calm) the sail must weathercock when the sheets are let fly. If it does not, then the rig is hopelessly unmanageable and possibly dangerous.

This ability to “weathercock” is affected by (among other things) the amount of “balance”.

The JRA definition of “balance” is out of date because when applied to the split rig it is ambiguous.

To clarify: The axis of rotation of the sail is assumed to be the centre line of the mast (and as it happens, in a “conventional” SJR or "split junk rig", the mast is assumed to be vertical and the luff of the “main” panels is assumed to be on that line.)

The area of all that part of the sail in front of that axis of rotation (including the area of the slot) is expressed as a percentage of the total sail area (including the area of the slot.)

In other words, the calculation is done in the normal way except that when we talk about “sail area” we must be sure to understand that the slot is included as part of the area of the sail.

As long as this convention is followed, then comparisons are valid and we all mean the same thing when we use that pesky term “balance.”

…one could argue that you could make it work to have a 50% balance split sail, as long as you only have 5% etc etc… Just forget about that,  Oscar, maybe you are right, but it is confusing. Lets just keep it simple.

Most people can calculate the area of the parallelogram panels without too much difficulty but the fanned upper panel (or panels) becomes a complicating factor and therefore most people use a short cut when doing this calculation, and simply base the calculation on the lower parallelogram panels. If the shape of the sail is fairly simple, then all that is necessary is to find the ratio of that part of the chord which is in front of the axis, to the total chord. This might not be 100% accurate, but the error is conservative – the balance calculated simply on the basis of chord length will be a slightly on the high side because of the lower-balanced upper panels being ignored. In other words, the true figure will be slightly less than calculated by the chord method, so any error is conservative, or “safe.”

Split junk rigs with a balance of 33% - 35% calculated simply by considering the chord of the regular-shaped lower panels have been found to weathercock satisfactorily when sheets are let fly and at this point in the development of the rig, this might perhaps be regarded as about the limit of what is practical. We can take 33% as a safe “rule of thumb” for a conventionally shaped and normally cut split junk sail. Slieve is comfortable with up to 35%. Others taker a more conservative approach and there is nothing wrong with being on the safe side.  I found 33% to be fine - Dave Z seems to have pushed it closer to 35% and found it OK, but as he explains, the chord calculation he used means his actual balance may be a little less than calculated. Personally, I would take 33 - 35% based on chord as a limit, not a target.

(Of course, anyone experimenting with extreme amounts of camber, or unusual sheeting angles, or sail shapes which are far away from what is conventional, or non-vertical mast-rake etc will be aware that this 33% rule of thumb for split junks may need to be re-assessed.)

Other factors which may affect the sail’s ability to weathercock (though none have been reported as far as I know) could include excess friction in the sheeting system, and binding of the sail due to insufficient halyard span or insufficient span for running parrel/downhauls. An aft-raking mast wouldn't help things either, though I am sure no junkie will be doing that.

The above remarks apply to the split junk rig. I am not sure why the balance of non-split or “contiguous” junk sails seems to be limited to 20% or less – whether the issue is weathercocking, or simply the windward performance of the sail, or difficulties in getting it to set properly,  I do not know. I am sure someone else can clarify that.

Problem is then how you would properly establish the CE/CP in order to get the correct lead (another fuzzy concept) for the rig. There is nothing difficult or “fuzzy” about this, other than the need to realise that there is no such thing as an actual centre of effort or centre of pressure. The so-called CoE is a purely geometric measure which is static and not real, but is a useful substitute and is conventionally used to help decide where the sail should be placed. It is calculated in the normal geometric way, the slot being considered part of the sail. With a conventional, simply-shaped SJR sail, it is near enough to simply use the mid-point of the chord. There is no point in being paranoid about accuracy here – the actual centre of effort is a moving target – the geometric centre is merely a useful convention. Slieve has suggested that with a SJR conversion it is sufficient to align the geometric centre of the sail (approximately the chord mid-point) with the so-called centre of effort of the original bermudan rig. That makes sense, and it worked for me. Again, this assumes conventional shapes and cut. A severely hooking leech, for example, will move the actual centre aft of where it should be. This placement of the sail may not apply to a contiguous (un-split) sail, I do not know about that, and would prefer to follow Arne's recommendations in that case.

I'm considering converting my current boat to split junk due to the split (or aero) allowing the most balance and thus the best mast position without having to make considerable modifications to the interior. From some quick sketches I would need something around 33-35% (including the slot).

Now, that makes sense and I think you are on the right track. Assuming you are converting from bermudan rig to any of the well-proven junk variants, I am sure you will be delighted with the result, best wishes.

PS  Without wanting to add to the confusion, but for the sake of completeness, people with conventional contiguous junk sails have reported that a forward-raking mast is desirable, among other reasons because when sailing off the wind in light airs the sail has a tendency to "sleep" in the sheeted out position - a kind of built-in "weathercocking" tendency. A small disadvantage of the SJR rig is that, so far at least, they have all been designed around an assumed vertical mast, for obvious reasons of simplicity - so with the split rig we don't get that extra, little benefit.

Hi All,

I'm a johhny-come-lately, having only recently become aware that there is some concern regarding overbalance in SJR sails with larger balances, especially approaching the top end of what Slieve McGalliard considers possible (35% of sail chord). Indeed, I had no idea that any balance 'reasonably' well under 50% might be a problem.

At 34.4% our SJR balance is very near that top top end. We sailed extensively in inland waters (no engine) this last year (write up here).

We can report absolutely no hesitation in weathercocking when sheets were let run beyond normal friction from the blocks. We noticed no difference in sheet responsiveness from our previous (flat cut) main, which had considerably less balance (20%), but was otherwise rigged identically.

However, there are some relevant differences between typical SJR and our prototype, include the following:

1. The higher peaked yard moves the CE of the uppermost panel aft. This would enhance weathercocking.
   
   
2. Our sheets are six part vs. the more usual three,  which induces more friction. This would impede weathercocking.
   
   
3. Opened by Thai style lashings, our after panels had gaps running along each batten. As their maximum point of camber (and therefore of gap) was 40% of panel chord, I suspect the after sail’s CE was moved aft to some small extent. This would enhance weathercocking.

We are planning to build final sails as prototyped this winter.

So... umm... data points for discussion.

Dave Z

Just a point I'd like to make about when talking about balance.  

When talking about balance in front of the mast,  I think that it is important that the posters clarify whether they are talking about "sail area" in front of the mast, or "chord length".  

When I made the Split Junk Rig for the Wayfarer, I both misread Slieve's files and in reading discussions on the forum, got it into my head that there was 33% sail area in front of the mast plus the split and built the jiblets accordingly. This would have resulted in my sail then having the luff of the jiblets approx 40% of the chord length in front of the mast.  

Luckily before I set sail, some postings earlier in this thread , warned that this was wrong and more likely correct that the leading edge of the jiblets should be no more 33% ahead of the mast.  So, not wanting to face building a complete new set of jiblets I cut a bit off the trailing edge of the jiblets, ending up with 29% sail area in front of the mast, but still 35% or so of chord length because of the split. 

The sail does feather with this amount of sail area and chord length, but it is slower to feather than I like.  So I intend build another set of jiblets to reduce the amount of sail area in front of the mast to 25% plus the split.

If the Weaverbird rig has 25% sail area in front of the mast and the Split Junk or Aerojunk has 25% sail area in front of the mast, I think the actual balancing effect would be different due to the split and thus increased chord length.     

So, in the interests of clarity and help in ending confusion among us newbies, I think it is important to clarify whether you are talking about actual "sail area" in front of the mast, or "chord length and sail area" in front of the mast.  

regards, Dave D

..drifting limits...

I guess my attitude to what the maximum balance in a one-ply JR should be, has drifted a bit over the years. In the early days of my cambered panel sails, I aimed for only 10% balance to minimise distortion of the camber on one tack. Fine in theory. Then Paul Thomson rigged his La Chica with 18 and 22% (?) balance, and got away with it. I still don’t dare to recommend a cambered panel sail with 25, 30 or 35% balance to others, and that is for one reason: I haven’t sailed with this much balance myself. I may well try such a hi-balance cambered sail, if I get access to a dinghy. If it turns out good, I will recommend it, but only then. I guess I am conservative...

However, by adding a split in the sail ( see Slieve’s SJR), the equation will change, to allow much more balance (ask Slieve how much).

Anyway, I wonder if we are discussing a not-problem here: As far as I remember, Christopher had no need for more than 20% balance.

Arne

Ps: A while ago I made the diagram below to help me find a what yard angle will match different amount of balance. As a result, if I were to set the sail with 20% balance, I would design it with a 65° yard.

(PPS: Only now did I spot Asmat’s warning against 28% balance...)