Impala wing repair - By Christo le Roux
This Blog will attempt to chronicle the repair of Malcolm’s big Impala wing after it got badly damaged during a bailout after launch at Chapmans Peak. A Port Jackson tree did a good job of dissecting one of the wing panels.
The basic wing construction is of a low density polystyrene foam core with what looks like 100g/m2 glass fibre cloth at 45° bias with a layer of 0.6mm Obeche wood veneer as a surface skin. There is a Meranti leading edge but no wing spars.
The repair process will be as follows:
Assess the damage
Cut out most of the damaged material
Insert lite ply braces to breach the damaged area
Fill the gaps with new blocks of polystyrene foam
Sand the polystyrene to the wing contour
Insert a Meranti leading edge
Epoxy a 200g/m2 carbon fibre cloth with a 49g/m2 glass fibre veil cloth over the repair area
Sand the repair area smooth
Use filler to fill any low spots
It is quite a nasty break as you can see in the photos.
Basically, this is how it looked when it arrived.
After removing the fibre tape, the wing came apart.
I used a razor saw to make a clean cut towards the flap split line. The bottom looks quite bad, with a large delamination.
A closer look at the problem area. If I cut this whole piece out, the flap will have to come off.
I made another cut with the razor saw, removing most of the “mush”. The bottom shows a missing piece by the flap.
Here is the missing piece. The servo tray is visible. All I can do is to glue this piece back.
The piece is epoxied and taped into position. Sand bags and weights are used to apply even pressure on the glued piece.
After the epoxy cured overnight, the two wing pieces were lined up by clamping an aluminium straight edge to the trailing edge of the wing.
The position for the 3mm lite ply spars were marked with a long ruler. The 3mm slots were cut very carefully and precisely to accommodate three ply spars that will breach the gap between the two wing pieces.
Here you can see how a scrap piece of 3mm ply fits in one of the slots.
Bit hard to see what is going on here, but I basically used a bunch of straight edges to clamp the wing pieces together so that they are perfectly aligned for gluing the spars in.
The spars are dry fitted, ready for gluing tonight.
This is what it looked like after the spars were glued in place. I used Balcotan polyurethane glue. It has a slight foaming action that fills any gaps between the bonding surfaces.
I used a razor saw and a sanding block to shape pieces of low density packaging polystyrene I had lying around into puzzle pieces that fit snugly in-between the spars.
I drilled a hole in one of the blocks to allow the servo cables to pass through.
Here all the pieces are dry fitted.
The foam blocks are glued in place with polyurethane glue.
It took about 1 hour of sanding with a long sanding block to get the foam to the wing contour.
On to the Meranti leading edge. It was cut slightly longer than the break area to help distribute the stress area. A cutout was made to accommodate the piece.
The piece was glued and clamped with packing tape.
Next time the leading edge will be sanded to shape and some of the veneer will need to be stripped back to allow me to sand the foam to the correct level. I suspect that it is going to be tricky to get right, so let’s hope it works out…
I used a small palm plane to shape the leading edge.
I used a marker to indicate where the veneer must be stripped back.
The idea was to strip back the veneer to expose the glass layer underneath around the damaged area so that the carbon/glass skin has a good surface to bond to and to ensure that the final top surface of the carbon/glass is on the right level (not above the veneer level).
I tried to peel back a small piece of the veneer but it is bonded very well to the glass underneath and de laminated the glass layer from the foam. I already anticipated this kind of difficulty, so I switched to plan B. If the veneer is well bonded to the glass to the foam, then it should be no problem to make the carbon/glass patch bond to the veneer. I just need to sand all the paint off the veneer in the area of overlap to ensure a good bond and to make it easier to feather the scarf joint between the carbon/glass layer and the original top surface. I am not too worried that the patch will leave a ridge since the wing surface is not very smooth to start with and there is a large enough area to feather the join area.
After some serious heavy sanding to get rid of most of the paint around the patch area, we are left with a good surface for the skin to bond to. The bottom of the wing is very undulated with lots of high and low spots. The low spots will need to be filled in.
I mixed some slow epoxy with lots of cabosil and glass balloons to make a thick filler that will be easy to sand down. The filler was used to fill all the low spots and to fill the ridges where the spars were slotted in. The spars were deliberately made to sit slightly under the surface to prevent a hard point when sanding the foam blocks to the wing contour. The hard point would have left high spots on the surface that would be very hard to sand out. I also used the epoxy mix to fill in some random damaged areas all over the wing.
Next time I need to fill in the spar ridges in the foam areas with lite fill in the same way as I did the epoxy fill in the veneer areas. After all the filler is set and dry, everything needs a final sanding to prepare it for the carbon/glass skin.
I sanded down the epoxy micro balloon filler that was applied on top of the spars and other gaps. I then marked the overlap area I want to cover with the carbon fiber and cut a paper template that I will use to cut the carbon to size.
I cut 200g/m2 carbon to the size of the template and 49g/m2 glass that is larger so that it overlaps the carbon. The 49g glass is used so that you don’t get millions of pin holes when you sand the carbon surface. The carbon is sprayed with 3M77 before it is applied to the patch area. The 3M77 will ensure that it stays in place while the epoxy is applied.
While the 3M77 is still tacky, the carbon is placed neatly in position and rubbed down.
The 49g glass cloth is cut oversize and also sprayed with 3M77 to apply it on top of the carbon. This is very tricky to do without trapping wrinkles because the glass cloth is very thin and distorts easily when handled.
I used slow set Ampreg21 epoxy and a roller to apply it to the cloth. One can use more epoxy than you think necessary because you try to fill all the weave of the cloth. Most of the 49g cloth will be sanded down together with all the excess epoxy. The glass cloth has no structural function in this application, it is there purely to hold the epoxy and to provide a smooth pin hole freeish surface.
When applying the epoxy I started sweating bullets because the carbon started delaminating and forming bubbles al aver the place! No matter how much I tried to work it down it kept making bubbles. To make things worse, I did the top and bottom at the same time, so just as you think the one surface is sorted, the other side was full of bubbles… I didn’t know if I had to pull everything off or try to tape everything down with packing tape. Packing tape over wet epoxy, now that sounds like fun! In the end I decided to let it rest for an hour to give the epoxy time to gel a bit. It made all the difference, once the epoxy started gelling, it conformed very nicely to the surface. I have used this technique for laminating glass epoxy before and I thought the 3M77 would keep everything in place, but this is the largest area I ever tried to cover in one go, so hence the difficulty. I got up at 3 in the morning to check that everything was still OK, which I am happy to report it was.
Yesterday I left the wing standing in the sun all day, so the epoxy cured nicely. The wing is now very stiff and probably 10 x stronger than before.
The next steps will be to sand down the excess glass/carbon flashing on the leading edge. Rub Combifill all over the patch and especially on the overlap areas. Then use 220 grit wet sand paper with a long sanding block to flatten everything out.
I used the long sanding block to sand off the excess glass/carbon on the leading edge.
I then went bedonnered with the Combifil. I used an old credit card to squidgy the stuff all over the patch area. It is quite difficult to do a neat job of it because the filler starts drying within seconds, so by the time the squidgy comes past for the second time it starts crumbling. The key is to work fast and to put it on thick and not worry about making it look neat.
The next step will be to use 220 grit wet sandpaper with a long sanding block to sand down 99.9% of the Combifill. Because we used a 49g glass cloth over the carbon, there should be very little pin holes to fill. The Combifill will remain in the “step” where the carbon cloth ends and also at the edge of the glass cloth. I expect that I will sand away some of the carbon cloth where it makes this step, so I will have to go back with some Combifill a second or third time to fill in the pin holes it will leave.
I used the 220 grit wet paper to sand down the Combifill. As you can see, there is still some filler left in the transition areas where the carbon and glass cloth makes a small “step” at the edges. You can also see that the patch area where the foam blocks were inserted is slightly lower than the immediate surrounding areas where there is veneer + carbon. Even though it looks like there is a ton of filler left behind, in reality it is only a few microns thick.
After all this messy wet sanding I went back and spot filled some low spots and new pin holes as expected. Unfortunately the photo for that doesn’t show the detail.
I took the big imp wing inside my work room in the house to do final preparations for painting.
So this is the patch area all filled and sanded, ready for some paint.
I used some benzene to clean the entire wing from all manner of dirt, grease, packing tape goo and general grime from fynbos juice.
I then used masking tape and sheets of paper to mask off all the areas that needs masking from paint.
I must have been bored or something, because I started looking at a delamination and hole around the one rear mounting. I mixed some epoxy and micro balloons to re laminate the glass and fill the hole (forgot to take a picture of the result).
That’s it, basically.
Now I just need to lay up some paint and reconnect the servos.
I used the white Duco primer to cover both top and bottom surfaces. Some of the underlying colors are still showing through, but I wanted to complete the next step before applying the final layer of primer.
As always, some pinholes showed up that needed filling with Combifill. A hard squidgy is used to fill the holes only and leave very little to sand down. The pinholes were all understandably located on the periphery of the patch area because these are the areas that were standing proud of the patch area and hence the 49g glass cloth was sanded through in these areas, leaving pin holes in the carbon cloth.
The Combifill dries quickly when it is applied so thinly. I used 360 grit sandpaper to rub down the excess filler. I used the sandpaper without water to prevent me from having to wait a long time for everything to dry before proceeding with the spray-painting. The sandpaper clogs up quickly this way, so one has to change it very often.
I used the white gloss Duco on the bottom of the wing as well as on the white panel area on the top surface of the wing. I use a professional low pressure gravity feed Devilbiss spray-gun. I pour the paint through a filter to prevent dirt from clogging the gun.
After the white paint cured, I carefully masked the bottom of the wing as well as the white panel area on the top of the wing.
Orange Duco was next. The funny thing standing up in the center of the wing is the bundle of servo cables wrapped with plastic and propped up with a piece of wire to keep it out of the way when painting. I left the wing standing outside in the sun for a couple of hours.
After removing the masking it started looking like an Impala wing again…
On with the job of threading the servo cables… I foresaw some troubles with this, since the aileron servo wire channel run underneath the flap servo. This is the exact area that had a triangular chunk of foam broken out that I had to glue back early on in the repair process. The urethane glue that I use will expand into cavities which is normally a good thing, but in this case it meant that the servo wire channel got clogged up. I ran a very strong stainless steel cable in the channel to allow me to pull very hard on it. The problem is, how do you tie something onto the steel cable without making a big bulge in the cable? I made little slits in the end of the servo wire and weaved the steel cable through it, then soaked it in CA glue and covered it in heat-shrink… The cable pulled fine, until it got stuck under the flap servo tray. I pulled and yanked and pulled but no luck. I then proceeded to cut a slot in the flap servo tray bottom to access the cable. This worked, but again the cable got stuck, this time in the middle of the patch area. Then to crown my frustration the steel cable came loose from the servo wire and I pulled the puller cable clean out of the wing without the servo cable! Two cups of coffee and a walk to the beach later, I managed to push the puller cable back into the wing and I could re-attach the servo lead and this time managed to pull the servo cable through! This was single handedly the most challenging part of this whole repair project!
I then soldered all the servos and leads back together. I left the labels on as per original so that there are no mix-ups or confusion, but most of them are in a bad state and should be redone before they fall off or become completely illegible. I used a servo tester to check that all the connections worked.
So there you have it, from this… , to this… in a blink of an eye.