Your point is well taken Arne, but it is worth considering how the camber is spread across each panel as well, a single measurement of maximum camber in the middle (vertically = 1/2P) doesn't tell the whole story.  Panel height, max camber and in the case of a shelf-foot sail, the foot angle all change how the built-in camber is spread across the sail.

Only Sebastian could answer why he changed to shelf foot sails, but there are two reasons that come readily to mind.  First is that paying customers may not want wrinkly sails, if you look at the sails Sebastian makes they not only set well, but also look great (scroll to bottom of page here).  I completely agree that this does not matter to everyone and that effort building the sail and difficulty are also considerations.  With a rounding-only, you have to decide if you want to stretch the sail along the battens to get rid of the wrinkles, or let it a bit slack to get the designed camber and have some wrinkles.  This is totally fine, it is just the way barrel-cut sails work and they clearly work very well.

The other reason that makes shelf-foot sails more interesting is that a sailmaker has more control over the shape of the sail, you can vary not just camber, but also the angle of the shelf foot.  This is never going to be case of WYSIWYG, but having more control over shaping the sail opens new opportunities not present with barrel cut. I don't think anyone has said barrel-cut isn't good, but surely there is room for other methods of making sails?

I haven't played around with each of the methods enough to convince myself I fully understand how things are work, but here are my thoughts:

I suspect there are sweet spots for each kind of sail making.  8 -10% camber with barrel cut sails seems to work very well with the sailplans you use.  If you look at a shelf foot sail you can see that a steep (45^o) shelf foot on a narrow panel sail yields an odd shape (very narrow-waisted hourglass panel).  So, it seems to me that the more camber you want, the shallower the shelf angle should be although the panel height matters.  At the extreme, a flat-shelf foot and high camber will lead to a sail that is baggy until there is sufficient wind to fill it.  Pango's sail while not extreme, is a nice example of this where Paul used a flat shelf foot, but lighter sailcloth to help the sail fill easily.  Similarly, if you want less camber (6%), then you have less "excess cloth" in the sail and a lower shelf angle lets you put the effective camber across a greater portion of the panel height.  At low (4%, 2%) camber barrel cut sails make a lot of sense because there isn't much rounding to try and force onto a straight batten.  I drew the top panels of Leeway's sail with just rounding, I think this makes sense for the low-camber triangular upper panels.

Whether any of this turns out better than barrel cut, who knows.  I'm trying to make a shelf-foot sail that I hope won't be much more difficult to build than a barrel-cut sail.  Maybe, I won't succeed, we'll see.  The point isn't to make hyper-accurate sails, it is to explore which techniques make the most sense for a given mission.  I'll post the actual camber when the sails set on the boat (which will be a joyous day).

Darren wrote:

Either reference can work if it is made explicit.  I'm not the first to make a 30^o shelf foot sail, although the search function has frustrated me in showing another example.  Given that flat and 45 don't change as a result of the reference point, I'd suggest using a horizontal reference as creating less confusion with the already existing posts in the JRA.

I agree. To be pedantic, orthogonal to the panel, not horizontal.

A shelf that is at 30˚ to the panel isn't worth taking the trouble to make, as it is not very different from a simple barrel cut. 

Darren wrote: If we are talking about 30^o and 45^o shelf feet, I think the reference plane has to be horizontal (orthogonal to the sail).  If you change your plane of reference to vertical (parallel to the sail), then a 30^o shelf foot no longer makes sense.  Given that the only references on the JRA are to Flat, 30 or 45 degree shelf feet.  I think we have already settled on horizontal as the plane of reference? 

I don’t think the reference has to be horizontal. If you wanted to design your sail using the vertical plane as a reference, you would simply choose 60 degrees as the parameter instead of 30, and get exactly the same result.

I suspect that Slieve had the vertical plane in mind when he wrote his notes some years ago on shelf foot sails (but it was never entirely clear, as his shelf angle of choice was 45 degrees, so the question did not arise).

You are breaking new ground, Darren. (Unless there is a body of literature somewhere I have missed). The term "shelf angle" has now become ambiguous.

I am faintly in favour of referring to the vertical plane when discussing shelf angle – rather than the horizontal plane which you have preferred, but it’s an arbitrary choice. I don’t think it has been settled on at all . That is the issue, and why I was asking if the established junk sail designers who have used the shelf foot method (in particular Paul and Slieve – and I think David) – could perhaps agree on a standard.

As you state, 45^o shelf feet are simpler to understand in that they don't change depending on the plane of reference,

That is not quite what I said. I don’t think there is anything simpler about 45 degrees, the trigonometry is exactly the same, its just a matter of choice of sail characteristics you want.

With 45 degrees, the scale factor between camber-of-cloth-shelf  and  designed-camber-of-sail is 1.4142 instead of the 1.1547 which you used. That's the only difference. Your sail will be a little baggier (or “fuller” if you like) than if you had used 45 degrees (for the same amount of camber).

The simple case is horizontal shelf, in which case the scale factor is 1.0000  (in other words, no scaling required). eg Pango’s sail. For a given amount of designed camber, this is the greatest volume you can get, when fully inflated. Beautifully made by Paul, this sail looks magnificent when filled with a puff of wind. I happen to think the less “bulgy” and perhaps slightly easier to fully inflate 45 degree shelf is a good choice too, and I am happy with it, intend using it again.

In shelf angle, your sail will be somewhere in between mine (45) and Pango’s (horizontal) – perhaps an even better compromise – I look forward to seeing the result of all your careful thought and attention to quality.

(I am not sure if the differences between 45 degrees, or your proposed shelf angle, or the horizontal, are all that earth-shatteringly great, actually. We are only talking about “tin plate” models anyway. The reality is, they will all be baggy, some a little baggier/saggier than others, some a little easier to inflate in light airs). 

I would have thought that for off-shore work the least baggy would be the better choice, but that's outside my experience, and anyway, as well as shelf angle, the amount of camber is also part of the mix.

As I think more about it, intuitively, the more baggy should have the higher number, so I will now advocate for the vertical plane as being the reference plane when discussing shelf angle. In that case, 30 degrees would be a fairly shallow shelf angle (the least baggy), 45 would be half way, 60 would be a steeper angle (like your proposed sail) and the maximum volume  (horizontal shelf) would be 90. That would be my choice of nomenclature.

But the other way round works too – as long as we don’t get mixed up which is which.

The most important thing is, we should all be speaking the same language.

So, Paul, Slieve and David: which way do you want it, the horizontal plane or the vertical plane as the reference for “shelf angle"? Arne, you can have the “casting vote” if needed!

[PS Thanks Darren for pointing out that some sail-makers have full-size plotting tables - I never thought of that - I'll look into it.]

Below (Pango) shows the massive volume you get in a horizontal shelf, fairly heavily cambered sail, when fully inflated. I still prefer 45 degrees for my next sail - but... perhaps I might follow Darren and go for a little more volume...

(that's Marcus up there...)

I would like to thank Darren for creating this thread,  and giving us a PDF of his step-by-step instructions in the use of QCAD to loft the panels of a shelf foot sail.

It inspired me to make the effort to get started and learn how to use QCAD myself, something which has been on my bucket list for some time.

I was able to follow the QCAD manual and to understand how QCAD is used to make 2D and isometric drawings of objects. Though I do not yet understand how Arne and David have been able to design hull shapes using QCAD (perhaps someone can provide some pointers here), I was able to understand Darren's instructions which show how QCAD can give us the patterns required to make a shelf foot cambered sail.

If I could find a print shop with large enough equipment to print full size patterns, I think this would make an easier and more interesting way of lofting my next sail. (So far I haven’t been able to find one in Auckland, but hopefully there is one somewhere?)

Thank you Darren.

I did have one serious problem in my understanding of this article, and to save the same mis-understanding occurring to other readers in the future, I would like to clarify something here.

In the “Instructions” PDF Darren wrote: “…since we’re using a 30o shelf-foot the airfoil needs to be taller, as running at 30o it has to cover a longer distance than a straight line…”

Well, er… I understand what is meant anyway.

The cloth aerofoil which makes the shelf, will need to be taller than the desired aerofoil shape in the sail, as it is on the hypotenuse of the above triangle.

You can work through the trigonometry if you like (given you know your camber height you can calculate the 30o shelf (hypotenuse) with b/cos a), but I was really pleased to find out that it all reduces to just scaling the airfoil vertically by 1.1547 for a 30o shelf foot…”

Here is where I wasted a few hours in desperation, struggling to know where I was mistaken in this very basic piece of trigonometry. I was certain, for all money, that the scale factor should be b/sina giving a scale factor of 2.000 – not 1.1547   - a discrepancy of potentially disastrous proportions.

Eventually it dawned on me: Following an earlier exposition of shelf foot development by Slieve McGalliard, I had a mind set and was intuitively following Slieve's diagram and his nomenclature.

The scale factor is b/sin45 here, and of course b/sin45 has the same value as b/cos45 anyway, so there is no possibility of confusion leading to a mistake.

But for a shelf foot angle of anything other than 45 degrees it is vital to be clear how shelf angle is defined.

Intuitively, to me, shelf angle is the angle shown in Slieve’s diagram, the angle in relation to the plane of the vertical sail. In Darren’s exposition it is implicit (though there is no diagram which makes this clear) that Darren defines shelf angle in relation to the "horizontal" plane, the plane which is orthogonal to the sail.

Going back to the first diagram above, Darren defines shelf angle as angle “a”, whereas I had intuitively followed what I thought was Slieve’s definition and was using angle “c”.

I actually prefer Slieve’s implicit definition (though I don’t know why, it probably doesn’t matter - though it is consistent with a reference I have seen to "horizontal shelf foot" as being "90 degree shelf foot") but the really important need is this: Would the experts please confer, decide and define unambiguously, what is meant by “shelf angle”. Is it in relation to the vertical plane of the sail, or to the horizontal or orthogonal plane of the sail?

The definition should then be added to the JRA "Vocabulary" of technical terms, and thereby prevent possible disastrous mis-understandings on the part of future DIY designers of shelf foot sails. (And, if I may suggest, perhaps Darren ought to include a clarifying diagram with his instructions).

As for me, I'm, sticking with 45 degrees!

........................................................................

[PS I would be extremely interested if a QCAD aficionado could give some instructions on the use of this software for designing or drawing boat hulls. Arne seems to have designed his dinghies using QCAD, including lofting the cambered plywood panels. The QCAD manual does not give instructions on how to do this, apart from the use of splines. A how-to PDF similar to Darren’s would be of great interest.

(I understand that nesting the panels on a plywood sheet can not be done with QCAD, but drawing the plywood hull shape in 2D and developing the conic sections? Striking waterlines and calculating displacement? Centres of area?)]

.......................................................................

A second matter, which has been pretty well canvassed now so should be understood by all, but again for future readers, perhaps should be laid down clearly in the JRA "Vocabulary": the term "camber" is ambiguous. A shelf foot sail with 8% of camber should NOT be be compared with a 8% camber sail of some other cut (such as Arne's barrel cut). They are not the same, and since the first question often asked by DIY sail makers is for advice on "how much camber" the following diagrams should be understood. The designed or theoretical camber which is built into a shelf foot sail results in an actual camber which is more - as James G., David T. and others have pointed out. (The same ambiguity possibly occurs when barrel cut sails are designed to have a certain camber, but the actual camber might not be the same?) Anyway, one thing is for sure, you can only very loosely compare the "tin plate" or designed camber of a shelf foot sail sail with the actual camber of a barrel-cut sail.

This diagram is attempting to show a comparison between the theoretical (tin plate) camber of a shelf foot sail, with the actual camber of a shelf foot sail (which bulges in the way tin plate will not.)

A good project would be to try to measure and find the relationship between "tin plate" camber and actual camber. It would then be possible to generalise with more validity as to what is the best camber to give to a shelf foot sail.

It seems to me that at the present time, the only valid way to make comparisons in the camber of various sails is to measure the REAL camber, in the way Arne is doing here.

(By the way, I built 10% tin plate camber into the jibs and 8% tin plate camber into the mains of my SJR shelf-cut sail. I don't know what the real cambers are, but the sail works just fine. I would be tempted even to push the numbers up a little, for a coastal (sheltered waters) cruiser. James's sail looks good to me. I can't help wondering if River Rat is simply slightly over-canvassed, or perhaps burdened by a slightly heavy mast, rather than considering the camber as being too great).

David wrote:

Oh, and another point: a 60˚/30˚ triangle has sides in the ratio 1:2:✓3 so that a shelf width of 2 units, sewn together with a panel concavity of 1 unit, gives a theoretical panel camber of 1.732 units. As the cloth isn't stiff like tinplate, but will be blown into a curve by the wind, the actual camber is something in the same neck of the woods as 2 units - this is sailmaking we're talking about, not precision engineering ;-)

So cut a shelf width equal to the camber you're aiming for, and cut the panel concave by half of that amount, and you won't be far wrong.

I found that out the hard way!  I calculated the panels for River Rat's sails using a complex spreadsheet, but didn't allow for the floppiness of cloth.  The result was that the panels had quite a lot more camber than I was aiming for.  It meant that going to windward in very light airs, she was fantastic... but easily overpowered in stronger winds.