Forest for the trees
Dave,
I doubt proseal per se is the problem. No doubt your highly experienced painter is on to something. There are many variables in play not the least of which are often wide differences in construction quality between one builder and the next. Consider a basic #40 hole. The very nature of a dimpled hole...already slightly over sized from the dimpling process makes it highly desirable in fuel tank construction that the shop head of the rivet be squeezed/crushed completely around the perimeter of its dimple. Also, if the hole is even slightly wallowed out you've got other dynamics working against a perfectly set rivet. In addition, unlike regular holes you simply cannot expect a rivet to completely swell within a dimpled hole so fully and so tight that even tiny traces of fuel vapor cannot eventually escape. You can get away with improperly squeezed rivets in other dimpled areas, but not when you expect a series of holes to be fuel and vapor proof. To insure the shop head of the rivet has a good chance of completely sealing off its dimple, I personally think it is a good idea to use a slightly longer rivet than as called out in the plans. And even then, if the rivet is even slightly off center...no way is the rivet going to completely seal off 100% of that dimple's diameter and render that ? length of the hole absolutely fuel proof. Granted, it is not strictly necessary to use a longer rivet than called out because of the forgiving nature of proseal. I just like to allow myself every advantage when I can. In the end, the only way to make that hole completely fuel proof is PROPER application of proseal and that includes shop head encapsulation which is very important. Apparently, your quickbuild tanks lacked that shop head encapsulation. This occurred, despite Van's own written instructions directing the standard kit builder to "apply a dab of sealer to the rivets."
If what has once been said here is true (I forget who said it) and Van's does not require its vendor to (in effect) add additional time and labor cost to the construction of its pre-made fuel tanks by requiring shop head encapsulation....the tanks merely have to pass a pressure test, I have to openly wonder why a standard kit builder is expected to do things one way and its vendor in another. Some months ago, I reworked a Piper Cherokee fuel tank. Upon opening it up, as expected the shops heads of the fuel tank rivets were properly covered in sealer. For whatever reasons, your quickbuild fuel tank rivets clearly are not. It is plausible...even convenient to opine that some unknown contaminant from the painting process may be causing the paint blisters but I just don't buy into that theory.
Still, I must add that my pet theory about shop head encapsulation is in and of itself no silver bullet. Proseal cannot do its job perfectly if not applied correctly and that includes making certain that mating surfaces having been properly prepared, cleaned and fay sealed. Omit this essential step and you may well create another problem. For instance, fuel will take advantage of any migration path between a rib and skin if any voids exist in the fay sealed boundary between the parts and much like a leaky roof on one side of a house can cause water to drip from the ceiling in another part of the house, you might end up discovering a leak and then drive yourself nuts trying to find its source. If a void in the proseal between the parts does exist, then the last remaining defense is the barrier of sealer that is suppose to be underneath the manufactured head of the rivet which is located in the second ? length portion of the hole.
I have, in effect discussed links in a chain.....hole quality, length of rivets, and properly set rivets neither undershot nor overshot. Proseal should be applied under the rivet's manufactured head, inside the hole, between the mating parts, and finally dabbed on the shop head. If all these things if were done correctly all the time by everybody everywhere, I believe the paint blistering problem would magically disappear.
re-installed tanks. any ideas what to do about this problem? robert
