You will need excel 5 (Mac) or excel/Office 97/98 (PC) or later and either a program  to plotting the panels or some sort of CAD program when it comes to printing out. This is if you wish to make changes in the plans with your own shapes, airfoil profiles etc.

The idea behind this project, apart from making the mod 3, was to write up the design in such a way as to enable more or less anyone to use the excel program and make their own changes to the shape, size and airfoil profiles. Turns out this was rather ambitious as the work involved in producing complete instructions is considerable but hopefully I've given enough instruction to get you started. You will still need some expertise in excel and maths.
 

The resulting design from these following pages can also be used as is without making any changes. There will be a set of plans at the end.

Recap of the mod 2
Introduction to the ESP mod 3 or "Tug"
Excel file of the Tug                             Note that this file is offline due to space limitations on the server.
Excel file of general formulae and constructing 2D shapes                           Ditto
Notes on the Tug excel file                                                                           Ditto. Mail me.
Results                                                                                                       
 

The plans

Changing the flat shape and vertical shape will require a bit of thinking about and some juggling with the formulae etc, on the other hand changing the airfoil profile and thickness is a fairly simple operation and will only require running the brown and red macros again to produce new data after you've changed the profile. See "Notes on the Tug excel file" below, in particular "Notes on profile and database" and "Notes on the XY projection pages".

Recap of the mod 2.

2001-02-24
So far mod 2 seems pretty good although I wouldn't call it a buggy engine. It seems to have a wider wind window, less pull than the NPW (the profile is only 10% thick) but the max to min ratio is higher. This means that when it's moving fast it pulls more than the NPW but stalled at the top of the wind-window it hardly pulls at all. It's also more tolerant to a drop in the wind as the shape traps the air and it floats longer. It's probably a good one to make in order to perfect sewing and flying techniques. The basic unscaled mod 2 was 1.5 sq m. Mod 3 will be 2 sq m or more. Changes I will be making include increasing the width but leaving the height unchanged, chopping off the wings square, increasing the profile to 15% and changing the vertical shaping to circular. 2 square metres seems a good size and with other changes will bring the aspect ratio up to around 3,5.

2001-03-29
I'm getting more and more pleased with the mod 2 now I've learnt to control it better and as I've made minor adjustments. The handles are still too long but I'm getting used to it. [I've since shortened the handles to around 25 cm, this does make it easier to fly but I had to learn it all over again!]
Removing the reinforcement line behind the nose (where the nose ties attached to the upper skin) made a big improvement in the flight characteristics as it removed the distortion in the profile. It now pulls pretty hard in a good (running speed) wind, so much so that I'm wondering whether a 3 sq m won't be too much (for me at any rate) for all but the lightest winds. I've also noticed that the nose ties seem to be slack at all times when I've been able to observe them but are necessary to prevent the kite from flipping inside out.

The reason for this is that I didn't know which shape the wind would try to curve the kite into and an ellipse obviously isn't right. The wind bends the kite a little more than I'd anticipated although the 3D design counters it to some extent. The result is that the upper skin is taut but the lower skin (nose) is slack as there is a little too much material there. It's only noticeable on close observation and I don't think it can affect the kite much at all.

It could be the wind would actually shape it to a catenary shape (the form of a hanging chain)  if it were a plain sheet but it's too complicated to calculate as there's also the question of the pressure under the kite and how much this will stiffen the structure. This tends to oppose the bending. When the wind increases it doesn't seem to change shape much so the wind window remains much the same. This is contrary to the npw which changes quite a lot with a higher wind and which ends up pulling like crazy straight in front of you. In other words the npw's drag increases a lot in a stronger wind but the lift doesn't. The esp doesn't seem to act the same way, the lift to drag ratio is more constant.

Introduction. Mod 3 or Tug.

I've since examined the ESP mod 2 closely, photographed the leading edge in flight and compared the shape to various curves. It seems that the nose area does follow a catenary curve but the sides deviate from this so they are almost pointing directly in line with the bridle lines. From this I've decided to use a circular shape as being the best compromise. This is very close to the central part of a catenary and fits well at the edges.
David Aberdeen has done some calculations on the shape the wind should induce and it appears that this is elliptical but with the a and b radii transposed from  my design, ie deeper than wide.
The deviation from a circle, however, is small so I shall continue with a circular vertical shape in the mod 3.

In mod 3 I will be trying to remove the bridle lines by removing the wing tips, panel 0. This panel is quite small and doesn't noticeably influence the kite. If I remove it I can remove the tip bridle line of course. Also if I make panel 1 longer, to go further forward, it may be possible to reduce the bridle lines on the leading edge to only one. That would leave one on the leading edge and one on the trailing edge. A total of four. One in each corner. In other words, no bridle at all for four line flying.

This is achieved by changing the shape from the mod 2
 
 
 
 
 
 
 
 
 
 
 

by stretching it and lopping off the wings thus. (It would be very interesting to try this kite with it turned upside down so the leading edge is the flatter curve. If you want to try this then "all" you need to do is to turn it upside down in the excel file before adding the airfoil profile.)
 
 
 
 
 
 
 

This is going to turn out looking a lot like Peter Lynn's Arc but it will be a single skin and cannot possibly be deemed a copy (it certainly isn't anyway). The one striking resemblance apart from the shape is the use of the wind with the 3D design to maintain the kite's shape with a minimum of bridle lines.

There are some design notes to aid you in using the excel file but these are not complete and I guess never will be. You can ask if you're not sure of something and I'll reply to questions but I will also use the questions to add a FAQ later as questions come in. Check for a FAQ (frequently asked questions) below before asking me.

The mod 3 has turned out to have an area of 2.3 sq .m. and aspect ratio of 3.4 (uncorrected).

Tug (Mod 3) Excel file
The excel file is now one large file with many pages covering:-

Outline (shape of the material lying flat)
Vertical shape (pull the middle of the material up off the floor in an arc)
Profile (same as the mod 2 but increased to 14% thickness)

The three pages above are combined in the pages below into a 3d database of points which describes the kite in masses of small rectangles. There are x,y,z coordinates for the corners of each rectangle. The rectangles are the width of the panel (around 15.5cm) but fractions of a cm deep. David Aberdeen has also pointed out that the only shape which will be flat is a triangle. This is quite correct of course as any rectangle can be bent over the diagonals. The projections and explanations I found before attempting this project have all used rectangles so this is the way I started too without contemplating the why's or wherefore's. The rectangles I use are really pretty small so any discrepancies shouldn't amount to too much. The shape of the mod 2 turned out pretty good and it flies well so I've continued using rectangles rather than start over.

Chords (varying in length and thickness over the kite in x,y,z coordinates)
ChordsAngled with profiles superimposed (these are angled to be perpendicular to the skin)
XYProj (Projection of the 3d coordinates back to 2d. This delivers the dimensions of each rectangle and angles between the sides).
XYProAligned (The trailing edge of the above rectangles are then aligned to the angles of the leading edges of the preceeding rectangle to produce one continuous panel made up of x,y coordinates.
 

That then is it, apart from plotting the coordinates to a 2d drawing or plan of each panel.
 

As to a name, I've decided on "Tug" and if it doesn't fly it'll be "Flop".
 
 

Excel useful formulae and constructing 2D Outlines Note that this file is offline due to space limitations on the server.

Tug.xls

Notes on Outline    2001-07-07 Note that this file is offline due to space limitations on the server.
NotesOnVrtShape.Html  2001-10-13 Same here.
Notes on profile and database2002-01-24 Same here.
Notes on the XY projection pages2002-02-05 Same here. Mail me.
 

In order to plot the excel data I've exported the sorted data to a text file, imported it into a Claris Works (Mac user) spreadsheet. Made a chart of it in xy line format, blown up the diagram real big for one panel at a time then copied it to a drawing sheet also in Claris Works (later versions are called Apple Works). From there it's exported as a pict image. If the diagram is big enough you can get a resolution of around 288 dpi for an image which is 30 cm x 120 cm. The pict image file I've reopened with Photoshop and adjusted the scales before exporting as a gif file. PC and Linux users will no doubt be able to find shareware or freeware programs which will do the same job.

Plans. Note all plans are in 288 dpi resolution and include a 5x5 cm grid to facilitate size checking. Note that the panel pics are quite larger although they are in gif format but beware, if you open them in your browser they become enormous and may eat up all the available memory and crash the program or show a broken link. It's better to download them right off and open with a photo program afterwards. The image resolution may then need to be set to 288 dpi to obtain the correct scaling or scale down for 72 or 96 dpi by dividing by four or three respectively. The leading edge of each panel is at the top and the trailing edge at the bottom. Apart from the 5x5 cm grid you will see thicker lines that run from side to side of each panel. These are the rectangles from which the projection is made and are left in as an aid to alginment when sewing or sticking together. The grid may not be in the same place on adjoining panels but the thicker lines are, so the left side of one panel will match the right side of the one next to it. The nose ties prevent the nose section from flipping inside out. The length along the skin is measured from the end of the underside of the nose along the seam or panel outline and the tie is attached from the end of the underside of the nose to a little reinforcement placed on the seam at the appropriate point for each tie on the inside of the kite.

 Panel 1           Be sure to read the note above.
 Panel 2                                     "
 Panel 3                                     "
 Panel 4                                     "
 Panel 5                                     "
 Panel 6                                     "
 Panel 7                                     "
 Panel 8                                     "
 Panel 9                                     "
 

Results:                                                                                                                                        2003-08-14 The Mod3 has finally been made and tested. It had only five problems . 1) A sewing error caused alignment errors (up to 1 cm)on the trailing edge). 2) It needed nose ties or else the leading edge folded back on itself. The nose tie table is below. 3) It wouldn't fly. 4) The wings flapped on turning (once I'd got past numbers one to three). 5) It's so powerful that it yanks the handles into another angle and spills air when it pulls hard and so collapses. If you have tried the mod 2 then this will be reminiscent, it's the whoomp, flop, whoomp syndrome again. Dismay not, it's all getting put right. The discussion below I figured didn't warrant a yellow box.

Results - Discussion
To start with I'd changed sewing machine and didn't get the foot pressure right so the two layers got out of sync and I ended with errors at the trailing edge of up to 1cm as in the gif below

Next time I'll use double sided tape or stick them together some other way first.
Anyway, I hot cut at the outer ends of the chords on the trailing edge, folded over and made a reinforcement channel for the 50kg dacron line (anchored at the outer 3 panels with stitching) and proceeded to test. (Dacron seems to have about the same degree of stretch as chikara.)

Add nose ties.

It just sat on the trailing edge and didn't want to go anywhere.  Also the wing tips looked aesthetically rather wide. Next step, hot cut away the reinforcement channel at the offending trailing edge (it's too long and gives too much braking action) by 1cm in the middle, 1.5 cm at the panel junction 8/7, 2cm at junction 7/6 and gradually increasing the cut off up to the edge, (this reduced the width of the wing-tips) then made a new channel for the reinforcement. Note that the sewing error has now well and truly been cut off and I'm just eating into the flap section at the centre and the actual profile at the wing-tips.

Test.
Guess what, it sits on the trailing edge and doesn't want to go up. The wings look better proportioned though .
Remove another cm equally over the whole trailing edge, make a new channel and test again.

[When all the trailing edge chopping was complete later, I marked the plans to show how much was cut away.]

No wind.
Wait.
No wind.
Run backwards until I fall over, but it looks like maybe it wants to go up now so I leave it as it is and head off on holiday - to the coast - must get wind there.

Day two of holiday, yep there's wind alright. Lots of it. Never good at judging wind speeds, this was like a good sprinting speed. To be on the safe side I started on the leaward side of the ridge at Chesil Beach, Weymouth. It shot off like a rocket and then hit the full force of the wind coming over the ridge, yanked me forward half a metre changing the handle angles too, stalled until I'd returned the handles to the correct angle so it shot off again pulled like mad and so on. Each time the kite aoa changed the wings flapped in the stall. Take a break, trying to grip the handles like a vice and try again, I just couldn't hold the handles steady. Wait for another day.

Less wind, much less, walking speed, but I tried anyway. Difficult to keep it in the air in lulls but nevertheless, it flew. The wing-tips do seem to stall and flap on turning. No sign of nose collapse at all.
Wait for more wind.
Got it.
Confirmed that the wings did indeed flap on turning, the nose is fine and it stays inflated well. Pulls hard when it gets up speed too. There seemed to be a distinct half oval shape at the wing-tips which stood out as the flapping bit. I guess there's a lack of wind pressure there. Can't have that, I'll chop it out when I get home. By the way, when cutting always fold the kite in half so you get an exactly symmetrical cut on both sides.
 

This is the result of the chopping and cutting out Note that this change is not in the plans as it's easier to make the kite first and then cut this out.

Seam the cut out section and test again.
No wind.
Off to the local slag heap. Wind up there all right. Yes, it has stopped the wing flapping even on turns.
It still shows no sign of nose collapse at any speed. The shape is good. The nose ties are neither under tension nor are they slack.

The handles I use (one for left and one for right) are around 25 cm long and originally the lines were attached at each end. I've now sawn up through the handles to get the lines only 12 cm apart (large enough for me to hold in between the lines.) There's a pic of the handles in my npw article. I use the handles to wind up on too so didn't want to shorten them.
The result of this is that the pull is now better balanced between leading edge and trailing edge and it isn't a problem to hold the kite at a steady angle. No more whoomp, flop syndrome. Sure there would be other ways of tackling this, some sort of control bar instead perhaps, but this was the quickest way. I've tested it again today in a very strong wind (small trees moving and I can't run that fast), couldn't hold the kite, it just dragged me as soon as I let it go up. On the other hand I could hold the handle angle even in that wind.

Doesn't seem to have a large wind window. I'd estimate vertically 60 degrees and less to the sides
 

To sum up, It is a single skin foil-kite with no bridles. This makes it an efficient kite with no drag from dozens (or even a half dozen) bridle lines. Compared to the ESP mod 2 it's less efficient and this is probably attributable to it's different shape. The wing to main body ratio is higher in the "Tug". This I felt was necessary to handle flying directly on flying lines with no bridles. It does mean that the Tug needs more wind to fly well but on the other hand it should be more controllable.  I say "should be" as I still haven't been able to check it out thoroughly in varying winds. Further testing shows that it isn't easier to fly or more stable if you push the TE cutting to the limit which is at around wind-window 75 degrees. At this point the wings start to flap and the nose starts to show signs of collapse occasionally. I've now backed off to the 60 degree vertical window. The wind-window remains constant over the manageable wind range unlike the NPW which has a lower w-w in higher winds. This is due to the fact that the shape doesn't change with wind force. It pulls hard ("Tug" is quite apt as it turns out so "Tug" it is. Let me know if you should find a kite released before mine with this name). The wing-tips don't stall or flap about except when the wind dies or you push it to the limit with cutting TE.. (Tried 3mm GF wing-tip spars at first but lost one on the first try-out and didn't notice until the kite was down again, they aren't needed anyway with the wing-tips cut out.) It has a fairly high max-min ratio, so is quite fierce. The max-min ratio, however, seems to decrease on removing some of the trailing edge. If you increase the lift to drag ratio by cutting away too much trailing edge then it will become uncontrollable. I've now cut away the amount shown on the plan drawings and the wind-window I'd guestimate is now around 70-75 degrees in a gusty wind. Finally it's not particularly easy to fly, it requires constant small adjustments of the brake inputs at least in the gusty winds (with which I'm plagued) but this does become second nature after while. Turning is most controllable by rotating both handles at the same time as one would a steering wheel but can also be done with the brake lines. I usually use a combination of both. Flying backwards is harder but that's the best way to land it as it leaves the kite ready to go up again.   If you want to make one, then you can use the above plans. As author I own the copyright so you are not allowed to copy and distribute any part of the plans or the internet published documents (download you can of course do for your own use) without my permission but you are free to make a "Tug" for your own use. Note that the plans haven't been corrected in any way from the original and it's still an experimental kite. This means that I'm not sure how much you'll need to chop the trailing edge as I've explained above under "results". A sewing error on my part made this necessary but some chopping will be needed anyway. I have marked on the plans the part which I have cut away. From the outset the design obviously was overgenerous with the trailing edge flap so some of this has to be taken away. I haven't changed the plans though as different materials, sewing techniques (no sewing error) etc may make this trimming variable. It's also a matter of personal preference as to how high a lift to drag ratio you want. Removing trailing edge reduces the drag component. I suggest that you make the panels including some of this area which I cut away. About two or three cms should be enough. After assembly then test on short lines (a few metres) and hold the right and left lines wide apart (the kite is designed allowing for the line angle with 20 metre flying lines). Try to get the kite up high enough to be out of the turbulence that you create standing in the path of the wind and see if it seems to want to go up. If it won't go up then remove about half a cm at a time from the trailing edge until it will. Once it will go up you can increase the wind-window and speed with further cutting. At this stage though you should go forward carefully, make sure you can always back up one stage by folding the trailing edge in and sewing it temporarily so you can remove it if you go too far. Test at this stage on long lines. It can make quite a difference. The wind-window (lift to drag ratio) increases as you remove trailing edge and the kite speed increases (so the pull increases proportional to the square of the speed). The pressure inside the kite holding it inflated will also decrease so there will be a point where there's not enough pressure left to hold the shape then the nose will cave in and the wings will flap. So it's a compromise. The wing-tip cut out is easiest to hot cut away after assembly and doesn't waste an awful lot of material.

Note that I give the tie lengths (as all other measurements) to two decimal places but it's not that critical. Get within a mm if you can, that's fine.

Future work will be aiming at a design somewhere in between the Mod 2 and "Tug" and investigate if I can improve on the profile further with a view to reducing the max-min ratio, make it easier to control and more stable. Also to improve the wind-window. Part of the problem with experimenting with this type of 3D kite is that changes can't easily be made without destroying the profile or shape. I may have found a reasonable way round this by getting professional print-outs on plastic. These could be hot cut joined which would do as a temporary test object. It would save a lot of time and probably be cheaper than using chikara.
 

The "Tug" flying in a low wind,
	Comments, mistakes, typos ? Mail me. Until the next model... With many thanks to all the foil and kite designers who have published on the net. They have provided the inspiration without which I would never have started. None mentioned, none forgotten. Thanks also to all family, friends and colleagues who've put up with much ear-bending about Cm, AoA, skin projecting etc ... without understanding a word of it. Tel                     33rd of August 2003.