Over the past few weekends I worked on closing out the right fuel tank. I attached the last accessory bits (fuel pickup, vent) and then proceeded with applying the sealant for the back channel/cover. Having the wing turned 90° definitely made it easier to set the fuel tank in place for the 3-day (minimum) cure time before I check for leaks. Nothing big to report compared to the left tank; the process went well, and as before, clean-up was annoying, especially cleaning the sealant off the clecos. One thing I did a little different, I applied sealant to the AN fittings for both the vent and the pickup tubes; it seemed unlikely I will ever need/want to remove these parts for maintenance, and I’d rather be assured the nuts won’t back off. I also noticed from factory build tank photos that they also apply sealant around the whole AN fitting.
I didn’t have time to check for leaks during the week after closing out, so the sealant had a full week to cure. I found out the hard way that setting the fuel tank like the above photo had an unintended consequence: just enough sealant had oozed out from the back plate onto the main wing spar cap, effectively gluing the tank onto the wing. I had to slide a scraper blade between the fuel tank and wing to free it again! As a result I have a few scrapes on the inside of the fuel tank skin that I need to smooth out.
With the tank finally separated, I moved onto the leak test. I decided to take the approach favored by fellow builder Pascal, which is to pressurize the tank and first check for gross leaks with soapy water, then monitor for leaks over 24 hours using a manometer (he has a good write up here: https://www.latten.net/sling2/fuel-tank-leak-test/).
Some differences in how I performed the check: I did not have to seal off the fuel tank cover with tape (no leaks detected at all), and I used the fuel supply and return fittings for my connections. The fuel vent was sealed off with a cap, and I and relied on the sump valve to seal properly – which it really should be doing anyway, being the lowest point in the tank. I also used English units (inches WC) instead of this funny metric stuff :). There are a few photos below of the test setup. I pressured my tank up to 14” WC indicated, which is actually 28”, a little over 1 PSI. You can definitely see the tank balloon up a little when it’s pressurized!
To get air into the tank, I made up an adapter for a floor pump by cutting off a Schrader valve from an old bike tire, and sealing it to the end of 3/8” clear vinyl tubing. I wrapped some black electrical tape around the base of the valve to ensure a good interference fit inside the vinyl tube, then put a zip tie around the outside. At first I tried just pushing the vinyl tube onto the AN6 threads of the fuel tank in/out fittings, but the threads don’t provide enough of a seal. Luckily, I had some nice AN6-to-barbed fitting from the firewall-forward kit, so I just borrowed a couple of those; they fit much tighter onto the 3/8” tubing.
Some good news and bad news for my first leak check – no leaks around any of the fittings nor fuel sender cover, which were my biggest concern areas. However, I did lose about 14” WC (about 0.5PSI) over the course of 24 hours, indicating that something was leaking slowly. I tracked down the leak to one of the sealed rivets along the stringer on the top side of the tank; interestingly, air was leaking out through the mandrel hole, which tells me the bottom of the rivet (inside the tank) is somehow damaged. I am going to drill this rivet out and replace it, since I should be able to reach inside through the fuel sender cover to retrieve the drilled-out rivet tail. After that, it will be another 3-day wait before I can re-check for leaks.