Fournier RF4D DIY rebuild: Part Two
PUBLISHED: 13:57 10 May 2019
Part two: learning about woodwork repairs, applying new skills − and making a shocking discovery | Words & photos: Bob Grimstead
After considerable financial and legal difficulties I - no kind of engineer but merely a retired airline pilot - faced the prospect of repairing, reassembling and re-covering my beloved but bare and completely dismantled Fournier RF4D G-AWGN, Wagon.
Having retrieved Wagon from its fourteen-month sequestration, its wing was now safely stored on my newly-acquired '£10,000 trailer' in my Polytunnel hangar, while the fuselage, engine, canopy, tail feathers and a heap of other, smaller components virtually filled my heated double garage.
First I needed to take stock of what I had. Out came the Maintenance Manual - a sparse publication, little more than a booklet, really - plus the more comprehensive bilingual French/German (but unfortunately not English) Parts List.
Who knew that such a little aeroplane would have so many parts? Well, of course, actually I did, but it was sobering to see all those many bits and pieces laid out on a bench top. Regrettably I didn't think to take a photograph.
After much cross-referencing, I was pleased to discover that, although my poor little aeroplane had been stripped down to its last nut and bolt, it seemed that nearly everything was still there.
Unfortunately, all the expensive fabric and paint I had bought had disappeared, and so had several virtually irreplaceable old-style metric bolts. Also nothing was tagged or bagged. I expected that other things might have gone astray too, but I actually felt optimistic that Wagon would fly again.
Learning about woodworking techniques
The next thing was clearly to get proper instruction in aircraft woodworking techniques. Serial homebuilder, former proprietor of Swindon Aircraft Timber Company and all-round great guy, Dudley Pattison runs inexpensive one-day courses under the auspices of Britain's affordable-flying organisation the Light Aircraft Association (LAA), so I enrolled in one of these.
I found it invaluable and, with newly-inspired confidence, I started planning the necessary repairs. Then I called my LAA inspector to come over and have a good look at everything and discuss a way forward. He was very positive and said it all should be within my capability and would be 'just a matter of time'.
Call me conceited−actually, I am conceited−but I have always reckoned that, given enough research and a little practice, I could probably undertake most manual tasks.
So, although aircraft woodworking was alien to me, I was pretty sure that, given my painstaking (some might say 'OCD') nature, I could make a reasonable hash of Wagon's necessary airframe repairs. Turning over the fuselage and lifting it on to trestles enabled me to view the damage and assess the challenge.
One positive thing was that Pat and Bobby, a couple of trustworthy and competent employees, had replaced the cockpit floor. And they had done it properly, using plywood, Aerodux adhesive and spruce I had purchased myself from Dudley and Bygone Aviation's ever-helpful Matt Petit.
They had also started, but unfortunately not completed, replacing the tailwheel block in the rearmost lower fuselage bay. Again, what had been done seemed to have been done well, but it wasn't the correct external shape nor had the four holes been drilled to take the tailwheel pivot bearing holder.
At first I was at a loss as to how to drill those four long holes exactly parallel, so that their bolts would fit the backing plate inside when the exterior of that tailwheel block was now semicircular. After a while I realised the best way to get those bolt holes in the right place was to push the bearing holder into its large circular socket and use it as a template to drill the smaller holes.
But how to ensure these holes were exactly parallel all the way through? It would be easy for a drill bit to wander off line. And where could I find such long three- and four-millimetre bits? The latter question was the easier to solve. As always, Drill Service Horley came up trumps with three long bits of the required diameters, mailed and received the following morning.
I eventually made a three-inch square wooden block, traced on it the tailwheel fitting's back-plate complete with holes, and drilled through this block with my pillar drill. I then precisely positioned this block on the tailwheel fitting and drilled through the whole lot with each of those long bits in succession to make the holes. To my delight, when I offered up the back-plate inside the fuselage it fitted perfectly!
Now I had to shape the outside of this block and the stern plate to conform with the fuselage's convex belly, but minus the two-millimetre thickness of ply skin that I would next be gluing onto it. I started this process by offering up the fin and bolting it into place to ensure that the final surface of the fabric-covered fuselage would align with it.
What glue to use
I know of four types. In chronological order: casein, urea-formaldehydes, phenolic resins, and epoxy resins.
The ancient, milk-based casein glues have long been superseded because of their tendency to be damaged by temperature variations and attacked by the microbes in rainwater which eat away the glue, destroying its strength.
Casein is no longer acceptable in aviation, although early Mosquitoes and most Miles and Percival aircraft used it.
Urea-formaldehydes like Aerolite 306/Prefere 5325 are inherently waterproof and mould-resistant. Aerolite is a two-part adhesive in which a powder is mixed with warm water to form a paste which is painted on to one component and then cured with a liquid acid hardener painted on to the other.
Aerodux 500 resorcinol-formaldehyde adhesive was invented by Duxford, Cambridge-based CIBA-Geigy in 1943 and is now produced by Dynea under the name Prefere 4050/5750.
Also waterproof, mould and UV-resistant, this fifty-fifty two-part adhesive is strong, durable, and waterproof and cures at room temperature. All Fourniers were constructed using resorcinol glues, so that's what I used for my repairs.
Epoxies like West System and the Araldites are also two-part synthetic resins. They have gap-filling ability and are more tolerant of poor joints or clamping pressure, but some have been found to be intolerant of high temperatures or excessive moisture, so only certain brands are acceptable in aviation use.
Then, using a three-foot length of rigid Dexion racking as a straight-edge with glasspaper attached to it, I sanded back and forth over the block, central spruce fillet and stern plate until they nearly conformed to the required shape. Finally, unable to find my engineering blue, I smothered one side with felt-tipped pen ink to reveal the high spots, gently sanding them down until I had the shape absolutely perfect.
I had to skin this reshaped area with suitably-curved aircraft-quality Finnish birch plywood. I would have used the remains of that ply sheet I had bought from Dudley, but was told it had been burned, like all the other wood I'd originally supplied.
But then I had a better idea. Why use new plywood of unknown moisture content, when in my garage was some good Finnish ply of the same age as Wagon, and thus similarly well seasoned? Some years previously I had been given the sawn-up remains of F-BMKC, the French-built, third pre-production RF4.
Although the actual tail skin section that 'GN needed had already been cut out of this, the wider piece immediately ahead of that was still there. Made from the right thickness of plywood, with its grain correctly aligned, it was even almost sufficiently curved.
So I cut it out and carefully chiselled away most of the fir internal structure. Next, I used sanding flap-wheels and a Dremel tool to grind away the remaining fragments even more cautiously. Obviously this liberated a lot of dust, including a good proportion of resorcinol glue powder, so face-mask and gloves were essential.
Having cleaned out the interior of this old tail skin, I needed to strip off the fifty-year-old exterior cotton fabric and its embedded nitrate or butyrate dope. I didn't like the idea of using paint-stripper because I couldn't be certain of washing off all the residue, and the consensus of opinion on the very informative American Fournier forum was to use a heat gun.
Unfortunately I didn't have a heat gun so−shhh, don't tell my wife−I used her hair drier. It doesn't get as hot as a heat gun, so it took longer than I hoped. Also, after an hour it stopped working (shhh again) so I substituted the spare one from our guest room.
Now I had a clean, curved piece of plywood, the right size but only half as tightly bent as I needed.
Paul Hendry-Smith of The Light Aeroplane Company told me they made their Sherwood Rangers' D-section leading edges by soaking ply sheets in hot water overnight, and then leaving them in an appropriately curved jig for a day to dry out to the correct shape.
So I ran half a bath of hot water, put a length of three-inch aluminium tubing inside the skin, wrapped it around with half-a-dozen strong elastic bands, and left this assemblage in the bath overnight.
In the morning it was more curved, but not yet ideal, so I tightened the elastic bands and changed the water, adding extra boiling water from a kettle and several saucepans. Four hours later it was virtually perfect. I put it in place on the upturned fuselage, held in position with bungee straps and one G cramp to make the sides parallel, leaving it to adopt its final, correct curvature while it dried out.
A couple of days later I was able to remove this now perfectly-fitting curved ply sheet to make the required scarfs (tapered gluing surfaces) around three of its edges, plus the edges of the mating plywood skin on the fuselage−not forgetting to allow for the thickness of the compressed bead of adhesive and allowing an overlap at the very tail.
The ply's scarfed edges were thinner than cigarette paper and not much stronger, so to support the sheet while sanding away at the scarfs I put it into a length of six-inch drain pipe. Once those scarfs were satisfactory I varnished the rest of its interior.
Finally I mixed up some Aerodux, painted it on to the top of the block, the spine and the scarfs, and around the adjoining areas of my new skin, put it in place, and stapled the whole thing in position. I covered it with a sheet of cling film to prevent any extruded adhesive from contaminating anything and finished off by re-fitting all the bungee straps to hold it all down while the glue set overnight.
I subsequently removed the clamps and staples, trimmed off the back end and drilled drain holes to let out any moisture from this most vulnerable part of the airframe. Job done!
Now for the wing...
A wet November morphed into a cold December and I got fed up paddling about in an inch of water with intermittent gales whistling through my wind-tunnel hangar, clothes and body.
Thanks to the generosity of Airfield Manager David McAlister I was able to shift my wing on its trailer into the big ex-Hawker Hangar 57 on Dunsfold's north side. It was dry, there were transparent roof panels and high-power lights, electricity, toilets and, best of all, a café within walking distance.
I spent many days in that echoing hangar with my good friend John Watkins, cleaning the wing of its thick layer of old Super Seam Cement with nitrile gloves, gallons of MEK solvent, a heap of rags and tons of elbow-grease. Once finished, I felt I was making progress and was very glad to get that eye-watering MEK smell out of my nostrils.
It was interesting to find so much Super Seam Cement plastered all over this wing. I guess it would normally come off with the fabric, but I was told "The fabric just fell off when cut along the trailing edge". I can only presume that the cement wasn't properly sucked up through the fabric by MEK in the first place.
Professional fabric workers might throw up their hands in horror, and that certainly isn't the way it should have been done, but it lasted 31 years like that and in my ownership I had flown more than 300 hours of vigorous aerobatics without any sign of the fabric detaching from the wing−which I guess goes to show how very cautious and conservative we all are in these matters.
I made a couple of small repairs where there had been storage damage, pushed out four areas of convex oil-canned ply, and leached out some oil staining on one leading-edge with naphtha and diatomaceous earth. I also made a rudimentary stand from four-by-two timber on which to turn over the wing to clean off its underside.
Once I was certain all chemical residues had been cleaned off and had thoroughly inspected its structure inch-by-inch to ensure it exhibited neither long-standing nor recent damage, I prepared to fix the ready-made plywood patches to cover iPhone sized holes cut into the underside of each wing's structural D-box for the new outrigger leg sockets.
It was then I came across something shocking.
I was about to varnish the insides of these plates prior to fitting them myself, when their significantly lighter colour attracted my attention. Then, as I studied them more closely, a couple more issues became apparent.
The plywood skin here is 2mm thick, with the outer grains running spanwise. The patches were also of 2mm ply with spanwise grains, but there the resemblance ceased. The wing's ply was clearly made up of five fine laminates glued together with reddish-brown resorcinol, whereas these patches had only three plies and were not resorcinol-glued.
Model aeroplanes were being built in the workshop at the time of this work, so I strongly suspected this material was not proper aircraft quality plywood at all, but model aeroplane ply.
I was warned by an experienced aero model builder that they looked like obechi ply, which has very little strength, so I took them to the local model shop for confirmation. "Oh no, they're not as good as that," said the guy. "They're what we call 'light ply'. See that central layer? That's balsa wood. It's about as strong as wet cardboard." I'm not often rendered speechless, but that day I was!
A wooden wing's leading-edge D-box is of course vital torsion-resisting structure, and these holes had been cut immediately outboard of rib No 9−coinciding with the outer end of the spoilers' torque tube, the aileron cut-out's inboard end and the outrigger leg attachments.
This was where the prototype wing broke during structural testing (albeit under a massive 13.8g) so it was important these plates and their adhesion were at least as strong as the original structure. This is not the thin aluminium leading-edge skin of a Cub or Champ's strutted wing, which is merely a fairing, but the seriously important torsion box of a cantilever wing.
I had no idea how much weaker those patches would have made my wings, but was really glad that they had not already been glued into place. Otherwise I would have had no clue that such a sub-standard repair had occurred until I found myself tumbling earthwards.
Furthermore, the aircraft maintenance 'bible' AC 43.13-1B stipulates scarfs of at least 1:12 for structural components, so with 2mm thick ply these scarfs should be 24mm or more (one inch) wide.
Measurement with a vernier calliper soon revealed that they were much narrower than this, their gradients ranging between 1:6.5 and 1:3.5, meaning that the finest scarf was only half the taper it should have been, and the worst was only a quarter. This meant there would be nothing like enough gluing area for safe adhesion.
So I spent the evening converting the edges of a couple of larger pieces of proper aircraft-quality, five-laminate plywood from F-BMKC into fine dust with a block of wood and some 240 grade glasspaper. It was time-consuming but strangely satisfying, and very reassuring to know things had now been done properly.
I couldn't just glue those patches onto the wing's skin because the clamping weights might distort the surrounding plywood, reducing the gluing area, so I next had to make two overlapping reinforcing squares of ply to be glued under the edges of those holes.
Then I could finally glue the patches in place, being sure not to use either too little or too much glue so that the patches would set level with the existing skin.
Holding them down with a stack of old gel-cell batteries, I left it all for 48 hours and was delighted to see that the end result was absolutely flush with the wing's original skin.
So now the airframe was ready for its fabric covering....