this is something i haven’t seen before and is blowing my mind. i’m coating an aircraft control panel made of aluminum. there are rivets holding mounting brackets to the back of the panel. my customer wanted the rivets, which were already very flat to the panel, filled/smoothed. i said no problem, they’re shallow enough that a primer coat should take care of them. so i hogged on some KL primer so that no rivet was visible. when i baked, the powder seemed to actually flow away from the recesses instead of into them. so i wet sanded them out using 100/220/340 grits wrapped around a flat wood block, down to metal in spots. this time i used powder-to-liquid to apply dabs of primer on the rivets only. when baked, this left raised bumps of primer (so i *knew* i got those suckers this time). i flatted them down then applied the grey powder. upon baking, i could still see the rivets – clearly – and something odd had happened… anywhere there was bare aluminum, the finish was dull (this won’t matter bc its final coating will be a liquid urethane). but i thought it odd. this time i used a DA at low speed with 120 to again flat the rivets. i finished off the whole panel with 220 wet on a block. i *did not* go down to bare metal this time. cleaned it up and applied grey again. baked again and son of a bitch, it looked almost exactly the same with the rivets visible and the dull spots.

1) what would have been the correct way to deal with these rivets from the beginning?
2) what can i do about them now?
3) why the heck is the primed part glossy and the bare metal part dull (i could understand the first time maybe, but after i laid a second coat on a first coat that was sanded uniformly?!)?

you can see how drastic the difference in this pic.