What causes this?
Sometimes it is corrosion, but sometimes when the paint bubble is removed, there is nothing there. The paint seems to have just decided to bubble up.
It could be that a fairly widespread airframe fabrication practice is to prime all the sheet metal components before riveting them, so the rivets are not primed. The paint which is then applied still “sticks” (well, kind of) but isn’t actually adhering to the rivets. I know Socata do this a lot but I have seen it on King Airs also. It probably avoids a 2-stage priming process because you get primer in between two riveted parts with just one primer step.
Solid rivets are typically zinc plated, which gives them their gold color. Depending on the subsequent paint, I’d imagine it may not stick well to the bare zinc. For repaints, the typical process is to strip to bare metal, etch and then conversion coat (Alodine) the surface including rivets before priming. I think that works pretty well in terms of paint adherence.
Excuse me, but I can’t help ranting here.
< rant >
Just like some people rant against traditional engines, as epitomised by Lycoming/Continental, preferring diesels or turbines or two strokes or Rotaxen or whatever, deeming the former hopelessly outdated; just in the same way I consider riveting old technology, and should have been banned from load-bearing applications long ago. Surely bonding is the way to go for joining aluminium pieces under stress?!
< / rant >
Surely bonding is the way to go for joining aluminium pieces under stress?!
If by that you mean applying some kind of “glue” then I would prefer rivets to be honest. If bonding means some kind of welding then I could accept it. But very difficult to repair.
RIveting is labor intensive but structurally efficient and easy to field repair. American Aviation (Jim Bede and company) did bonded aluminum light aircraft structures in the late 60s. They generally worked (‘purple glue’ issue not withstanding) and they continue to work. But for practicality and structural efficiency rivets are fine – especially since the major improvement in manufacturing that came with CNC matched hole construction. That is a huge thing that is invisible to the consumer.
A friend of mine works for a major European insurance company, developing field repair designs in collaboration with German auto manufacturers. I’ve spent time with him there, and it was interesting to see that as car manufacturers have finally started to build more efficient structures, using higher strength non-weldable materials, they’re adding riveted joints. They use ‘productionized’ rivet designs, and then for field repairs they have to spec different rivets. They also use some bonded structures, and there is nothing wrong with it at all except that field repairs are more difficult and that drives insurance rates up.
Reaching beyond bonded aluminum and If composites are not your thing, a spot welded 2219 AL structure (as per rockets) might be cheap to assemble 
Actually I think CNC skins are OK.
Bonding alloy structures is old technology, Redux bonding was used in the late 40’s in the DH Dove and other DH metal aircraft.
I suspect that DH were happy about bonding because of the extensive experience they had with bonding wooden aircraft, unfortunately that metal centric industry has failed to grasp this technology.
Surely bonding is the way to go for joining aluminium pieces under stress?!
And why is that exactly? The best way og joining metals is welding. Unfortunately 2024 alclad is impossible to weld. The next best is bolting/riveting. Ship hulls using 5 or 6000 series of aluminium are welded. But 5 and 6 series are not nearly as strong as 2 or 7000 series used in aircraft. Some small aircraft use 6 series, and could in principle be welded, but welding 0.5 mm sheets sounds very difficult to me.
LeSving, you might be interested in weldable 2219 aluminum. It was developed as a corrosion resistant high strength alloy for aerospace structures – in effect a weldable version of 2024. It has the highest strength of weldable alloys, although not as high as non-weldable aluminums, but 2219-T851 is expensive.
Otherwise, it might be useful to point out that a riveted joint can be made as strong as the patent material (or stable, whatever limits) and rivets don’t weigh much. They also fill the holes completely so the joint’s structural characteristics are predictable, and they don’t affect the patent material. With CNC’d holes you can design a riveted structure so it is self tooling. If you’re going with metal construction I don’t think there’s a lot of scope to find fault with riveted joints except for the labor to install the rivets.
One issue with welding is that you can no longer have overlapping joints in which the overlap is protected (from water ingress and subsequent severe and rapid corrosion) by one of
So if you are welding, you can’t have overlaps unless it is absolutely assured that the overlap gap is totally sealed.
Whatever happened to friction stir welding for aluminium aircraft structures? I seem to remember some VLJ maker a few years ago used to go on about their friction stir welded aircraft all the time in their PR (I don’t remember who it was, but it was one of the start-ups at the time)
