Thanks for the quick responses.

I will admit that I may have misled you by talking about keels first, and on that subject I reckon most junkies must simply accept the keel on the hull they end up with. My main interest is in the top of the sail, and a set up to have the lowest induced drag/ vortex drag and hence the best L/D ratio.

Winglets seem to be the biggest improver for transport aircraft, but somehow I can’t see that they can be practically adapted for sailing rigs. David wrote, “So my hunch (no more than that) is that we have to decide which way to jump.” I guess that sums up what I am thinking, and in many ways that ties in with the question of sail balance.

Look at the photo below of the elect aeroplane which claims to be the fastest of its type in the world. I’m sure much research will have gone into the wing tip design, and I’ve also seen this shape on many slow speed but high performance model glider designs which also require best L/D ratio.

I’m not surprised that this is considered to be an efficient tip design as back in the 50’s I was building my own design model aeroplanes and found one design with flared wing tips to perform better and faster than expected. Again, about 1970 there was some problem with the wings of the Trident 2 flown by BEA, and some aeroplanes had modified wing tips fitted. The original tips were flared out to the trailing edge, but the modified ones curved in, only a few inches but enough to require an a recalculation of the increased fuel load.

The one thing about these tip shapes is that they are suitable for sailing rigs as they can work on both port and starboard tacks, and do not necessarily increase weight aloft.

Are we to assume that this tip shape pushes the vortex out and back and reduces the drag?

Graeme’s mention of end plates on the yard would increase the weight aloft, and there’s no guarantee that the yard angle might be already concentration the vortex up to the tip. As he says, it would be difficult to appreciate any gain or loss.

I’ve been looking at the videos of the Americas Cup racing and it is noticeable that they have removed the gap below the sails, even putting small skirts to completely seal the gap, and have squared of the mainsail head to horizontal. (This latter is probably some rule about no sail above the black band (Black band with blach sails and black mast!) Anything to raise the L/D ratio.

Just more thoughts!

Cheers, Slieve.

As I understand it, an aircraft's winglet is toed in a few degrees, and cambered on its inner face (a bit like bilgeboards and leeboards), and interacts with the tip vortex in such a way as to add a small lift vector in the forwards direction (at the expense of more friction drag). Well and good, we might be able to put one on the lee side of a yard, the low pressure side equivalent to the upper side of a wing. But hang on a minute, we need our sails to work on both tacks, and having such a winglet on the weather side would cancel out the positive effect on the lee side (a bit like leaving both bilgeboards or leeboards down). So I don't think the type of winglet as generally used on today's aircraft is a workable notion when applied to sails. We might go for the fence type of winglet though, more like a delta shape attached vertically to an aircraft's wing, or horizontally to the bottom of a keel or rudder. We're familiar with those, on shoal draught boats. It might work in modifying the tip vortex in the way we want. Trouble is, the yard would have to be very close to horizontal. So my hunch (no more than that) is that we have to decide which way to jump. Do we go for a near-vertical yard, and lengthen our leading edge and increase our aspect ratio? Or do we go for a near-horizontal yard, and hope to decrease the negative effect of the tip vortex? Or do we go for a yard at 45˚, and fall between two stools aerodynamically but get ourselves a rig which is easiest to manage and sail with? Decisions, decisions...  

I remember David Thomas, the designer of Tystie and no slouch when it came to racing, looking at my cambered panel sail approvingly, and saying "the wind likes to be told where to go". Channeling the wind along the panels with protruding battens might just be a Good Thing. 

It has been good to read some of the recent threads on the website, and they have encouraged me to lift down my copy of Tony Marchaj’s book ‘Sailing Theory and Practice’ from the top shelf, and blow the dust off.

Page 101 has a good drawing of Jester, but little useful discussion, but not finding the diagram I was looking for I latched onto the photo attached to this mail. (Sorry, but I don't know who to embed it with the text.

In the book it is described as, “the appearance of this additional resistance can only be due to ‘end leakage’ , which is the flow of water under the bottom of the fin keel from the region of pressure to that of suction. This leads to the formation of vortices, such as those clearly visible in the photo, which shows the water flow lines around a model of a Dragon’s hull. As with the sails, this kind of resistance is called induced resistance.”

Induced resistance, induced drag, vortex drag are all names for this drag which is formed by anything that produces lift. In the photo the vortex is quite large diameter and is about one third of the height of the rudder. The question is, can this vortex be reduced or pushed out of the way to allow the keel to work more efficiently?

The answer is yes, but it would not be within the class rules of the boat, however we can possibly learn something from this photo.

My thoughts on the subject, and these are wide open for criticism, are -

The high pressure on the far side of the keel is escaping under the keel and diluting (weakening) the low pressure on the near (camera) side and therefore the pressure difference (lift force) across the keel is weakened and more leeway will result. If the foot of the keel slope downward more acutely it would be harder for the water to flow under the keel and in practice there must be an angle when the pressure would not be able to escape and would flow along the keel to the trailing edge. With no leakage there would be a higher total force across the keel, and the inevitable vortex would not be able to form until the trailing edge of the keel, and then it would be of smaller diameter and concentrated at the lower tip of the keel. The net result would be a better lift/ drag ratio, which is the key to better performance.

The actual angle required for this situation unfortunately would depend on a number of variable so would inevitably end up being a compromise. As a keel for the Dragon it would probably be too steep for the boat to dry out on, and the draft would probably be out of all proportion.

We are told that action and reaction are equal and opposite, so we can turn the diagram upside down, change its aspect ratio as air is less dense than water, and think of it as a sailing rig. Imagine a rectangular (junk) rig made of a number of parallel panels with a horizontal top (and no batten rise just for convenience). As in the above photo, the air at the top would curl over from high pressure on the windward side to the low pressure on the leeward side, and form a large diameter vortex. But if the top of the rig, the yard, was angled upwards, peaked up, the vortex could be decreased and even get pushed right to the peak of the yard, reducing the pressure leakage and the diameter of the vortex. To me this suggests that there should be a compromise yard angle that could reduce induced drag and give a good L/D ratio.

Conversely, if the head of the rig sloped down, as the leech of a Bermudian mainsail does, then the vortex will spread down the sail and be of quite large diameter. Some of the diagrams on Tony Marchaj’s book do indicate rotating airflow over the upper area of a Bermudian mainsail, and certainly not flying of the top.

Does anyone have any thought on this idea?

Modern airliner wing tips are fitted with winglets to achieve this tight vortex right at the tip to achieve much higher L/D and better fuel efficiency. They are not practical for junk rigs, but I believe there are ways it can be done.

What say you?

Cheers, Slieve.

Paul.

I don’t know how relevant this is to the tip of a junk sail.

Tip vortices are hard for me to imagine, but I saw a video clip some months ago which set me wondering if a square topped sail might benefit from an “end plate”.

I saved the link to this rather beautiful little video clip. It is here.

It shows a piper cub turning in close to the ground, for a landing, through a layer of fog just above the runway. As the aircraft cuts through the fog, the wingtip vortices show up very clearly. I guess you have seen plenty of this sort of thing.

In case the link doesn’t work I captured a few freeze frames.

It seemed to me that it would not be too difficult to add a thin plywood “end plate” to the yard – at least to that part of the yard which is aft of the mast. It would be easy enough to do. I asked Slieve and he did not see much value in the idea and, too many projects anyway, so I left it at that.

One of the major unsolved questions that remain with the cambered junk rig is, why the rig draws so well even if on the “bad tack” with the mast distorting the camber profile totally.

(At least I read this between the lines while reading through JRA forum threads. Please be so kind and direct me to the answer to that question, if it was already found and I am writing here about something already solved. Much appreciated!)

I’ll try to keep the theoretical introduction to this issue short: When you look at a sail, you look at an airfoil. If you slice our beautiful junk rig in the direction of airflow, so more or less horizontally, you get a 2D profile (please, only slice virtually in your head…). There you go, with different profiles for different junk concepts:

Now with these 2D-profiles, you can draw lift and drag vectors, discuss upwash, alpha-tolerance, laminar/turbulent flow, boundary layer characteristics, attached and detached flow and all the like. Though these are interesting issues, one is lacking one important factor: induced drag.

Induced drag is not explicable on a 2D profile, as it is a 3D effect: The pressure on the windward side of the sail and the suction on the leeward side of the sail create a huge vortex at the highest tip of the sail. This vortex creates a lot of issues with airfoils, for example the induced drag. I do not want to go into detail here, but with airplanes this induced drag can be contributing up to about 40% (or higher, depending on the kind and speed of the airplane). I do not know the exact numbers for sails on a boat, but the induced drag is potentially the second highest contributor to overall sail drag – after friction. A short (really short) google research offered me the following web page from NorthSails:

Quite an interesting read, concerning sailing aerodynamics! Though it deals with Bermudan rigs, the physics are the same for the junk rig. The article also shows some very understandable views taken from CFD simulations, which are able to explain that tip vortex at one sight.

Again, the tip vortex induces a significant amount of drag onto the sail: the induced drag. Induced drag has a very, very high influence on overall sail drag, close after friction drag.

Now, windward performance is strongly dependant on drag. Slieve perfectly explained why, in his “some thoughts” document. So, as the induced drag from the tip vortex has a significant effect on the overall drag, this induced drag also has a significant effect on windward performance. Right?

Now, in aviation the induced drag can be significantly reduced by wingtips. Everyone who ever sat in a airplane (fast passenger jet, not slow propeller machine) will remember the little vertical wings at the end of the main wing. They are winglets (or wingtip fences, or endplates, …), and they are there to limit the unwanted airflow from the pressure side (lower side) to the suction side (upper side). It works so good, that it is commonly seen on most airplanes (sorry to any aviator for such a generalisation!).

We don’t have that on sails of a sailboat, do we? I’ve seen endplates on ruder tips, I think on fotos from Arne. But never on sails… or do we?

It was when we were tacking up the Kalmarsund against F5-6 yesterday, when it struck me. We were sailing heavily reefed, upwind, making good progress in one of the most ugliest, choppy seastates the already very choppy Baltic sea has to offer. By the way, we reefed and unreefed 16 times that day, with me singlehandedly sailing Ilvy while Toni was under deck due to her hurt ankle. It was the most easiest exercise, and really showed off the junk rig capabilities. A lot of huge, white, shiny sailing yachts with carbon/kevlar sails motored past us while we steadily kept sailing, and after we tacked even into the harbour, we were the talk of the evening in the Kalmar harbour. I think it we produced some good junk rig advertisement that day.

The battens might act as winglets, only not located at the wingtip but distributed vertically along the sail. Thus limiting vertical airflow, thus decreasing the tip vortex, thus lowering the induced drag, thus lowering overall drag significantly, thus increasing windward performance significantly. And as those camber pockets form on port as well as on starboard tack, this tip vortex lowering effect makes for nice upwind performance on both tacks.

The theory I describe here explains the good upwind performance of sewn-in cambered sails. It is a theory, yet, not verified by experiments/simulations. I do not know the actual numbers/percentages involved, and I do not intend to risk an educated guess. It might be that the tip vortex lowering effect of the cambered junk rig batten pockets is negligible, it might be that it is significant if not dominant. I try to stay on the non-speculative side. My explanation here is rather a hypothesis, as long as not backed by experiment or simulation. It would, however, be an explanation for the unexpectedly good performance of the “bad tack”.

What do you think? Please do not hesitate with critics!