Evaluation of closed end rivnuts

I was surprised to see that the fuel tank called for open ended rivnuts. Was one supposed to cover the ends of the rivnuts with sealant? Does one rely on sealing the threads of the M4 bolts that go into these rivnuts with 577? The manual mentions the latter, but not the former. I contacted Sling and they indicated to cover the open ends of the rivnuts with sealant. Presumably with a bolt inserted to stop the sealant filling some of the threads. I do not know how this stuff flows nor how hard it would be to extract a bolt that has been partly glued in with sealant.

A way around this is to use closed-end rivnuts. I found a suitable part from Huck Aerobolts, a Nutsert Xtralock, M4 closed end with grip range 0.5-2.5mm. In the fuel tank these rivnuts clamp a 0.63mm and 1.0mm plate together so this is ideal.

I obtained samples, and conducted a trial, using the correct thickness of 6061-T6 (0.63mm + 1.0mm),check reamed with a 15/64" hole and the rivnut set with moderate force. These were assembled with the gasket, fibre washer, stainless steel washer and M4 x 16mm bolt.

Defining distance D from the head of the bolt to the tail of the rivnut, as illustrated for the D = 23 case

D = 28mm with the bolt thread just beginning to engage with the rivnut thread

D = 23mm with the cork just beginning to be compressed (5mm thread engagement)

D = 22mm with the cork squeezed 1mm (6mm thread engagment)

D = 21mm with the thread bottomed out on the closed end of the rivnut (7mm thread engagment)

This last case represents 2mm (about 50% of a starting thickness of 4.3mm) compression of the cork gasket. I am not sure what the torque is at this level, or whether this is sufficient.

If the cork needs to be squeezed more than (say) 1.5mm I will need to use a shorter bolt, like 14mm. However 14mm is not a standard length, the next shorter length is 12mm. Grinding down 16mm bolts may be necessary.

For if I were to need to squeeze the cork down say 3mm and use a 12mm bolt, this would mean 4mm thread engagement which seems too little.

I did a direct torque test on the rivnut, the grip on the aluminium failed before I reached 17 in-lbs, probably about 14 in-lbs from feel. I will test how this will be improved by sealant, and by different setting forces.

I have 50 rivnuts on order.