We’ve got far enough down the road to get the aircraft registered with the CAA, and while the piece of paper hasn’t come through yet, they’ve taken our money and Peter Walsh at the relevant CAA department was happy that everything was in order. In fact he got very enthusiastic about it for a civil servant!
It’s a strange fact that aircraft registrations are NEVER re-used, even after the aircraft itself has long since gone to the great scrap-heap in the sky, its registration is only ever used the once. You’d think they would be running out of them by now, but there doesn’t seem to be any danger of that.
You’re allowed to choose your own registration letters if you want to – and provided, of course, that they haven’t been used before. UK registrations are of the form G-XXXX, where the X’s are the letters you can choose. We tried various combinations around the word Scout but weren’t convinced by them. In the end, we chose G-FDHB, which were my grandfather’s initials, andseemed appropriate. In fact, the only place they will appear is on the fireproof plate, but I think it will be nice to remember Granddad’s connection to the aircraft in this way.
Of course, we hope it will be painted in its original colours and markings – or lack of them – and for that we need to ask permission of the RAF, but it seems unlikely they will refuse. After all, it’s unlikely to be confused with a current service aircraft!
While Theo and I were busy with the linen covering, Rick was also hard at work with Ian Harris on the oil tank. The result is, frankly, a work of art, and it’s criminal that it, along with the majority of the internal structure, will be hidden from view on the completed aircraft. So feast your eyes on this wonderful piece of work while it’s still available. It’s been made exactly as the drawing stipulated, and Ian has even managed to acquire from somewhere the original brand of filler cap – a Rotherham’s no. 2.
The front and back are made by forming the flanges over a wooden template. The straight bottom flange is easy, of course, but the curved one has to be gently tapped with a lightweight curved wooden bat, making sure that you never let a crease form. The result is that the brass actually gets thicker towards the edge of the flange.
The ends are slightly dished to stop them from ‘panting’, and Ian’s method of doing this is frankly genius, and I won’t give the game away! It’s finished off in an English wheel.
The sump at the bottom is made using a mandrel – a steel template machined from solid. A disc of brass is clamped to it in the lathe, and the brass is ever so gently forced around the shape of the mandrel using a wooden paddle. By the time you get to the top of the sump, the brass is considerably thicker than than it started out.
There are lots and lots of rivets. Each one has to be cut to exactly the right length, and peened over. Getting exactly the right strength and direction of tap is a real art, and Ian and Rick got down to about 2 minutes per rivet by the end. Even so, its a monster workload.
When that’s done, you apply solder to the joints. For all except the front end, you can apply the solder from both inside and out, but the final joint has to be done ‘blind’, and it’s a considerable worry to know whether it has formed a perfect seal. Happily it did, and the tank has survived its pressure test, which means applying an air pressure of 1.5psi to the inside and wiping a 50:50 mix of water and washing up liquid all over it. Any leaks will show up as bubbles.
The next stage is to mount it onto the aircraft, and here, I admit, we aren’t following the original drawings. The Scout was famous for leaking oil tanks (here is one example, but there are lots more).
The reason is that the tank had originally been located behind the pilot, where it was accessible. It was decided to move it forward because of oil starvation problems, and in the haste to do this, they hid it under the plywood cover forward of the cockpit. The ply cover is almost impossible to remove, making the tank almost impossible to access. The tank had channels rivetted and soldered to the underside of the tank which fitted over the fuselage longerons, and they recognised that vibration of the longerons could cause leaks where the channels were attached to the tank because they fitted bits of rubber between the top and inside of the longerons, and the channels. but there wasn’t room for rubber on the outside of the longerons, so the whole exercise was a bit of a waste of time.
Since the whole thing is out of sight – even if you poke your head in under the instrument panel – we’ve decided to adopt modern technology and bond the channels to the tank via strips of rubber – much like an exhaust mounting bracket on a car. This should reduce or eliminate the risk of these leaks.
One of the compromises we’ve had, reluctantly, to make, to ensure the aircraft is practical to operate.
Incidentally, on the later Scout D they fixed the problem by making a combined oil and petrol tank which filled more or less the whole space between the cabane struts and didn’t have a ply cover, so if you’re thinking of building a Scout D you won’t have this particular issue!
And now, Ian and Rick have moved on to the petrol tank, which is a good deal more complicated, since it’s a sort of conical shape – wider at the front – and because its made of tinned steel, not brass. It also has a couple of baffle plates fitted internally. On the plus side, the petrol tank isn’t covered up with ply, so the craftsmanship will be permanently on display.
Our two weeks covering the flying surfaces is over. It’s been the most emotionally draining of the entire project, and the most physically tiring.
Unlike the rest of the structure, it’s permanently on show; above all, it’s what it will be judged on, and yet it’s technically the most difficult bit, since (as we’ve seen) it’s regarded as a black art. Despite the fact that it was how the vast majority of aircraft were made from 1911 to the 1940s, there’s no standard procedure and many of the experts give contradictory advice.
We could have chosen to use modern heat-shrink fabrics. These are much easier to apply; we’ve done it many times before and the result would have probably been more acceptable to the untrained eye. But we set out to create an aircraft that was as historically accurate as possible, and that meant using the original covering.
We could have chosen to have the covering done by the experts who have been so generous with their advice, but it would have pushed the cost of the project way outside our comfort zone, and we have tried all along to do as much of this ourselves as possible. It’s cheaper, for sure, but we have learned so very much more along the way.
So it had to be this way.
And although we haven’t finished, we have completed the sewing, fitting and initial coats of dope on all the flying surfaces, and completed the ribstitching and taping on a couple of small surfaces and one wing. It’s a good point to stop; there are a number of wrinkly areas, and while they may be acceptable, we’d like to get expert advice to see if anything can be done about the before carrying on with the final coats of dope. We have been incredibly lucky with the weather; it’s become clear that humidity (or, more precisely the lack of it) is crucial in the whole business and we’ve had virtually no rain the entire fortnight. Now both Theo and I need to have a week of normal life before the next stage of getting ready for the LAA Rally over the August Bank Holiday at Sywell, which will be its first public outing. Most of August will be dedicated to getting ready for that.
This process has been used to cover tens of thousands of aircraft. From 1911 to 1940, it was how the vast majority of aircraft from the smallest sport plane to large airliners, from the slowest glider to the fastest fighter plane were covered. It was applied on an industrial scale by manufacturers large and small worldwide, by workers who were not highly trained experts. It must therefore have been studied and documented by any number of governments, manufacturers and universities and been well understood and repeatable.
So how does it work? How does the doping process shrink the fabric to provide a taut, flat surface? Nobody knows. The experts we’ve spoken to don’t know; I can’t find any explanation on the internet.
How much does doping shrink the fabric?
My test piece, using an unrestrained 500mm square with a single coat of nitrate tautening dope showed no shrinkage at all. After 24 hours it still measured exactly 500mm square. One expert said that it will only tauten the fabric if it’s already in tension, and yet the initial application of the dope will always make it relax, so that wouldn’t seem to be the explanation.
How much tension should be applied to the fabric to start with? Based on the above test, the answer would seem to be that the fabric must be stretched as much as you will need on the finished aircraft, and yet experts say that it should be straight, but not in tension, otherwise the structure can be crushed.
Should you apply water to the fabric before doping? Two experts say you should apply as little as possible. Although it will tighten the fabric beautifully, it will cause it to sag somehow when the dope is applied. The third expert says you should apply a good even spray coat of water to get a nice even tension. Our experience is that no matter what you do, initial application of the dope will cause the fabric to sag alarmingly. As the dope dries it will shrink back again but generally not back to its original state.
All agree that it should be demineralised water applied with a spray. As far as I can tell, this is the only thing on which they agree.
What sort of dope should be applied? The manufacturer firmly states that butyrate dope should be applied. All the experts prefer nitrate, particularly for the initial, tautening coats.
Should it be thinned with solvent before application? Two experts say it should be thinned considerably for the initial coats; one says it should be applied unthinned and warmed.
All the experts and the manufacturer agree that there should be sufficient time between coats, but no-one agrees how much time that should be.
How many coats should be applied? Some experts say no more than three; others say up to eight, provided they are thinned right down.
If there are wrinkles after application, how can they be removed? The manufacturer clearly says you should apply more coats of shrinking dope. Some experts advise the use of boiling water, others recommend re-activating the dope with thinners, and others recommend the use of many coats of well-thinned dope.
Everybody warns that dope will continue to shrink throughout its life to the extent that it can crush the structure underneath. Our own experience with wrinkly fabric after being left a week is that it’s no different.
So far, we have covered eight small pieces of the aircraft. Of these the first, the rudder, has come up successfully. It was the first to be sewn and doped, and was therefore the least expertly done. All the others, no matter what initial tension has been applied, or how much water sprayed on, or what sort of dope was used, or how much the dope was thinned, are wrinkly to a greater or lesser extent.
We have had limited success in removing wrinkles, but it can take an hour or more to have any effect on an area of a square foot, and with a total fabric area of 400sq ft, we could be there for some considerable time, and in any case, subsequent coats of dope will cause the fabric to relax, undoing all the good work you did…
So today, we applied water to the last aileron and, as always, it came up perfectly flat and taut. We let it dry – it’s been a roasting hot day – ideal conditions and we were keen to make the most of it.
Then we applied a coat of dope 50:50 with thinners. When it was dry, miracle of miracles – no wrinkles! On that basis we decided to do the same to the first wing, and lo and behold, it too dried to the same perfectly flat surface as when it had been water shrunk.
Theo had been applying dope and thinners alternately to the other wrinkly surfaces and putting them outside in the sun, and the improvements in the centre section while not dramatic, were sufficient to suggest it was probably okay, so I got on and ribstitched it and applied the rib and edge tapes.
The result is certainly satisfactory. Meanwhile Theo was busy doing the last of the sewing on the last wing.
Then in the late afternoon we applied a second coat of dope, thinned 70:30, to the aileron, and all the wrinkles appeared. By now we were committed to doping the first wing, so we decided to apply a second coat, but still thinned 50:50.
And the result? Wrinkles. We will have to wait now to see whether they sort themselves out, but by last thing at night we felt that they were looking pretty much okay, and felt a sense of provisional relief.
It’s been a pretty depressing day today. In fact I’m not sure when we’ve come across such an intractable problem.
We’ve spent ten days sewing up the linen covering for all the flying surfaces, and got eight of them – all except the wings – covered. Seven of them have been doped, but only the first – the rudder – has worked successfully. The others are still too wrinkly to be acceptable.
The sewing shop is running very smoothly under the supervision of Theo. In fact he’s now developed a very successful way of sewing round corners, which is neater, more accurate and of course much, much faster than doing them by hand. We are most of the way through making the bags for the wings, and all the other pieces are complete.
It’s the doping shop (my responsibility) that’s causing the problem, and we can’t work out why. If you read the manufacturers website, they firmly recommend using butyrate dopes for natural fabrics and that’s what we’ve done most of so far. As I said in the previous post, we’d heard from several experts that nitrate dope was preferred for at least the initial, tautening, coats, and so we’d decided to switch over. We also gathered that we should only apply the absolute minimum quantity of water – apparently, although it shrinks the linen, it also makes it softer and saggier.
So this morning when the new dope arrived, we took one more aileron, made sure the covering was as flat as possible, applied a very few drops of water in a couple of areas that weren’t quite flat and applied the first coat of nitrate dope.
The result? No difference.
We only have one more aileron to experiment with and we certainly don’t feel confident about carrying on to dope the wings at this stage. We need first-hand assistance and advice before carrying on, which is very disappointing, since we had planned to have all the covering done by the end of this week. There are other things we can do in the meantime, and shall be working on getting some first-hand advice ASAP, but tonight there’s a cloud hanging over the workshop!
On Sunday we more or less had a day off to recharge the batteries, watch the Silverstone Grand Prix and the Wimbledon Men’s final, and eat a delicious Sunday lunch provided by Theo’s lady Fran.
It’s been a tough week, and here’s a list of what we’ve learned during it.
Before You Start
Check to make sure the structure is absolutely complete, that the paint, varnish, etch primer and any other coatings are complete.
Check that all fasteners are permanently secured; locknuts tightened down, screws fully home, split pins doubled back.
We applied Wax-oyl to the ends of the internal bracing wires to try and stop them going rusty.
Wrap the metal parts that the fabric comes into contact with. In our case this is the wing and tailplane leading and trailing edges, and the entire structure of the elevators and rudder. You should use tape wound round the metal in a spiral, with a small overlap. We found 1” wide was the best, and made our own from offcuts, splitting it using a craft knife pinched in a Workmate and drawing the fabric across it, using a pencil mark as a guide. Fix the first end with dope, then wind all the way down to the other end. Where it has to be fitted round a fitting, such as a rib end, try and make sure that any creases are on the inside, out of contact with the fabric. When you get to the end of the tape, fix it with another dab of dope, and then dope all along the wrap to secure it. I only bothered with the back of the wrap, since the front side would get plenty of dope from the fabric.
We also applied anti-chafing tape, 1/2in wide, wherever the fabric would be in contact with wood, or might come in contact. For clear doped fabric, this also helps to prevent any colour from the wood bleeding through. The fabric doesn’t follow the line of the ribs – in between the ribs it will tend to pull into a straight line. This means that in addition to the rib caps, you will need to consider many of the spars and stringers inside the wing. I started using Superseam cement, but found that dope was easier and quite good enough.
This process also makes you check for any lumps and bumps that might be in contact with the fabric. This might be screw heads, lumps of glue, etc.
Now check the internal structure all over again. It’s the last time you will be able to see all that beautiful work you’ve done, and it’s amazing how difficult it is to spot every last error of omission. It’s a really good idea to get someone else to look at it. Even if they know nothing about building an aeroplane, it’s likely their eye will light on something completely obvious you’ve overlooked.
Fixing the fabric
Traditionally the fabric was sewn into a ‘bag’, which was sewn by machine as far as possible, then closed with hand stitch, and some places where it’s glued in place with dope.
Machine stitches are much faster, but can only be used on straight seams. Hand stitching is slow but can manage curves.
Gluing with dope can only be used where there’s sufficient area of wood for the glue joint; you couldn’t use it on the leading or trailing edges, for example.
These are done with a balloon stitch, which requires the two edges to be folded inside each other, and lines of stitching run just inside each fold. You can get a twin needle machine that will do it all automatically, but doing it with a normal machine takes a bit more effort. It’s made a lot easier if you do a first tacking stitch to hold the two edges together, then iron the folds in place and pin them, before doing the two main stitches.
It’s generally a two-man job, particularly if you are doing a long seam; an extra pair of hands is necessary for handling the bulk of fabric and holding the folds in place while they are ironed.
For something like a wing, sewing the first seam (at the leading edge, for example) is relatively straightforward, but doing a second (such as the trailing edge) is a lot more complex, since it has to be in exactly on the trailing edge, and the fabric has to be exactly the right tension. After a very great deal of experimenting and head scratching we found the best solution is to make pencil marks (DON’T use biro – it bleeds through onto the finished surface!) on both bits of fabric where you want the middle of the seam to be. Then mark a line exactly 15mm from the pencil marks and cut the fabric. Now tack the edges together with one edge 10mm from the other. It’s extremely important which one stick out more, as it determines whether the tacking stitch is on the same side of the tube as the first seam!
Folding, ironing and pinning are done as before, but the seams may have to be sewn from each end, as the whole tube may not fit through the machine.
The American bible, AC43-13, says to use a baseball stitch, but we gathered that the standard British way was different, more complex and time-consuming. Since the whole thing is hidden under edge tape and the scout was made before any such manuals were written (and maybe the seamstresses in the factory in 1915 didn’t follow the instructions!) we decided to go with the baseball stitch. The edges are cut about 1/2in long and tucked inside, and you sew through the main fabric and the tuck under. The tricky bit is getting the fold in just the right place so that when the stitches are pulled together you have just the right tension in the fabric.
As I said in the previous post, it’s unbelievably time-consuming; about 18in (500mm) per hour, and you can’t watch television while you’re doing it or the stitches go wrong.
Originally they would have used dope, with loads of little tacks to hold it in place while it dried. The tacks tend to go rusty and don’t do the wood any good, so we used modern glue – called Superseam cement – in order to comply with the airworthiness requirements.
It’s horrible stuff to use; glutinous and stringy, and the tin has a tiny lid which makes it very difficult to get at, and becomes completely glued up. You need at least 30mm of wood to stick to, and we found the best way was to make sure the wood surface was horizontal, then pour a small amount onto it – enough for around 4-6in of glue joint. Spread it with an old kitchen knife, then lay the fabric into it, adjusting the position to get the tension correct. Now use the knife as a squeegee to force the glue through the fabric. If necessary, dip the knife into the tin to get more glue on top. It starts to dry very, very quickly, which is why you only do a little bit at a time. We used glue joints on the sides of the centre section, and the curvature of the wing section means that the fabric won’t lie perfectly flat on the wood. With a bit of luck you can persuade the wrinkles to lie flat if the curvature isn’t too great. If you can’t do that, you’ll need to make slits in it, but it’s better not to leave this until you’re half way through gluing!
Modern polyester covering materials are shrunk using heat – often using an iron. It will shrink a huge amount, and the tension can be increased by using a hotter iron. This means that it’s not particularly important to have the fabric taut when it’s applied.
Linen is different, however. The amount of shrinkage is much, much smaller, so you will need to get it sitting almost perfectly before starting the shrinking and doping process.
Machine seams are the least easy to adjust for tension. Hand sewn seams are better, but glued seams are the best. On the wing, we used machine seams on the leading and trailing edges and on the straight part of the tip. The curve at the tip was hand sewn, and the inboard ends and the cutout for the aileron were glued.
The next step is to apply water to the fabric. You must use demineralised water and apply it with a spray. It’s a simple and very satisfactory process, and within a few minutes the creases have all disappeared, the fabric has stretched itself gently taut, and you’re likely to feel confident that once it’s fully dried, it’s going to be perfect. But you’d be wrong.
The next stage is to apply dope. There are two types. The original substance, developed in 1911, was cellulose nitrate and was the first to provide a permanent, waterproof finish that didn’t stretch over time. The downside is that it’s very flammable, and a chance spark from an engine backfire or any other source can ignite the whole airframe. In the 1950s, an alternative, butyrate dope, was developed. Both systems have the disadvantage that they will continue to shrink the fabric over its lifetime, and we were warned that it was very important not to get the fabric too tight to start with, or it would distort the frame underneath in only a few years.
Both systems can be bought as tautening and non-tautening (which contains a plasticiser), and it’s standard practice to apply a couple of coats of tautening dope to start with, followed by subsequent coats of non-tautening in order to limit the continuing fabric shrinkage.
Applying the dope, like spraying with water, is a real pleasure. You thin the first coats down about 70:30 with acetone thinner, then slap it onto the fabric. There’s no need to be careful with the brush strokes – it will soak into the fabric and be quite even. If you apply vast quantities it can form drips on the inside surface of the fabric, but it’s not very easy to do this.
What then happens then is very disappointing. Instead of tightening, the fabric actually relaxes, going really quite slack. It’s touch dry in about an hour, and we found that although it tightened up a bit, it was rarely perfectly tight, and there were wrinkles in some places. Applying a second coat of tautening dope sometimes helped, but not always enough.
We had used the modern butyrate system because of its fire resistance, but after considerable experimentation with smaller pieces (ailerons, rudder, elevator) and advice from other experts, we’ve decided for the wings to use nitrate tautening dope (which tautens the fabric more) followed by finishing coats of butyrate. This is a very commonly used system on aircraft, and should meet our needs.
As we saw in the previous post, one surprise bonus is that the butyrate non-tautening contains a small amount of tan dye in it to make it easier to see where it’s been applied which reasonably accurately recreates the look of the final varnish coat which would have been applied to clear doped aircraft of this period.
With the first two coats of dope applied, you have to ‘sew’ the fabric to the ribs, or it can pull away in flight, modifying the aerofoil section with potentially disastrous consequences.
Waxed linen cord is used for this with very long needle to post it from side to side. First of all, however, you have to dope a piece of 1/2in wide cotton tape along the line of the rib. Then mark out the positions of the holes – on the Scout they should be every 3in and carefully make the holes with the needle. The first loop is secured with a reef know on each side, and then the cord is run alongside the reinforcing tape to the next set of holes, where it’s looped round the rib and tied in a modified seine knot and so on. For the undercambered wing, of course, the fabric will need to be pulled in quite a bit until it contacts the rib.
Now dope tapes over all of the rib stitching. The Scout has tape 1.25in wide, and we had to trim these down from 3in tapes. It may be a good idea to add tapes wherever there’s a risk of the fabric contacting the internal structure or other parts of the airframe.
Finally, tape must be doped around the edges of the piece. In the case of the rudder, this involves four quite tight curves, and as we saw in the previous entry, what’s required is brute force. The tapes have sat down very satisfactorily round the edges, with no wrinkles. After that it’s a couple of coats of non-tautening until you get a satisfactory finish. For more modern aircraft, this means a deep, lustrous, mirror-like finish which takes any amount of care and attention, and would be way beyond our level of skill. Thankfully, the Scout won’t need anything like that, and more or less anything that’s waterproof will be good enough!
We applied dope to the elevators this morning, and rang Skysport Engineering to ask about the colour in the dope, and Steve, who has been hugely helpful throughout the process, said that the aircraft would originally have been painted in a top coat of varnish, so the slight discolouration is entirely realistic – and once it’s flown a few times it will all get covered in castor oil anyway!
So, feeling much more optimistic, we applied more tautening dope to the elevators and managed another leading edge wing seam – something that takes two of us to manage the large quantity of fabric.
A machine seam requires the following steps;
Mark off the fabric at the point where you want the centre of the seam.
Cut it exactly 15mm from the marks.
Pin the edges together facing the same way, 10mm apart. Sew a tacking seam 5mm from the inner edge.
Fold the lower protruding edge over the upper by exactly 10mm and iron in place.
Fold the top fabric over the seam, creasing, ironing and pinning it exactly 10mm from the other fold.
Machine a seam 1.5mm from the visible folded edge.
Turn it over and do the same for the other folded edge.
This is relatively straightforward (if time-consuming) for the first seam of a wing. If you plan to machine sew both leading and trailing edges, there are a couple of other complications; first, the positioning of the seam is critical to get the tensioning correct; second, you’ll need to be very careful that your tacked seam ends up on the inside, and third, access will be made difficult as you’re making the fabric into a tube which can’t all fit over the sewing machine, so you may have to sew from each end of the tube.
We decided not to do any more tubes until after the wings had been signed off by our inspector on Saturday morning, so I spent time going through the wings making absolutely sure everything was completely finished inside, since we’d not be able to get access again without disturbing the covering. This included a final check on bolts, screws, wire locking, leather wear pads and cable splice whipping, and then applying some Waxoyl to the ends of the piano wire to try and eliminate corrosion.
Theo started sewing the trailing edge of the tailplane. If you thought machine sewing was time-consuming, just try hand sewing! I suppose the 2.1m of seam to be sewn here must have taken about 4 or 5 man hours in total.
And I applied the edge tape to the rudder. It was something else I’d asked Steve at Skysport about, because Iwasn’t sure how the tape would fit round the corners, but Steve said to use plenty of muscle to stretch it round, and he was absolutely right! The end result is that the rudder is now ready to go flying, apart from the tail stripes to identify it as an RNAS machine, and we’re pretty proud of it!