Gas fuel tank leak pressure test check before paint

I ruined a new paintjob back in 1980 because I did not check if the new tank was leak-proof. Never crossed my mind to check the tank, turned out it had a very small hole in a weld. A beautiful paint it was too, made by a good friend. I just fixed it temporarily so it was ok over the summer, then it was fixed and repainted the next winter.
 
I ruined a new paintjob back in 1980 because I did not check if the new tank was leak-proof. Never crossed my mind to check the tank, turned out it had a very small hole in a weld. A beautiful paint it was too, made by a good friend. I just fixed it temporarily so it was ok over the summer, then it was fixed and repainted the next winter.
Test, don't guess" 😎

I've weld repaired quite a few (for friends & family) while minimizing paint damage. A little touch up and so it went.
 
I didn´t guess, it just never crossed my mind to check the brand new tank.
Understood.

And, when my truck was overheating, 4th of July weekend, I troubleshot the new GM branded thermostat that was only partially opening.
Worked perfect, from when I installed it, in March.
UNTIL, the engine was WORKING, WORKING HARD.... pulling a 32' camper on a 95F day.
Turns out it opened at 195F, but only .060" (insufficient flow under high load)
Now, I test EVERY ONE, brand new & be damned, before install.
Live & learn
 
Last edited:
Please don't remind me, there is a thread somewhere where I asked about the stud spacing at the front of the tank.

That thread confirmed why when I got the bike one hole in the frame attachment wings was widened to accommodate the studs.
It must have been pressure checked and over pressured even though not obvious to the eye spreading the tank and then widening the stud spacing.

I had been tempted to fit some form of turnbuckle to the studs and give it a tweak (once a week perhaps) to try and slowly get them back to something stock but was a chicken.
They got machined offsets (insets ?) instead.

View attachment 111500

There was a member in the past that I recall (built his 850 from a frame and then bought parts as he progressed) got a tank tested at a radiator shop and they near turned it into a puffer fish.... Whoops.
The anti-puffer fish method, to possibly help it from happening to others. 💡
Gas fuel tank leak pressure test check  before paint
 
When I pressure test I only use a car foot pump or a bicycle pump.

Typically I use a cork / rubber bung and drill a hole in the centre.
Inserted into the hole I use a car wheel valve ( Schraeder sp ? )

Attach the pump to the valve and typically I find two or three pump motions is plenty to find a leak and I try to avoid more than 2 psi, never failed yet.

I use this method for tanks, radiators, car heaters and even whole engines.

Bung size depends on application.
 
I temporarily fitted the gas cap, compressing a piece of thin spongey rubber sheet, had one gas tap fitted and turned off, then connected the Mityvac pressure tester to the other one. Easy control and went to 2-3psi with no leaky.
 
Nowadays I use my lung power
Just put a sandwich bag over the fuel cap fitting on the tank
Fit the filler cap
Then blow into the tank through the petrol pipe and turn the valve off to hold the air
Then go round the tank with soapy water
If no leaks the tank will hold air until you open the tap
I have done this many many times
 
I worked many years in the chemical industry. Sometimes we were faced with pressure testing with air instead of water (long and irrelevant reasons); point being the safety implications are major. If a tank full of water ruptures, the pressure drops immediately without the massive release of potential energy that comes with a failure involving pressurized air, even at fairly low pressures. I like Concours' and others' techniques posted here. The decoy failure could have been a lot worse.
Having shared this, I will confess to taking a Honda gas tank up to about 15 psi on air to 'pop out' a dent. It actually worked and I got away with it.
 
I worked many years in the chemical industry. Sometimes we were faced with pressure testing with air instead of water (long and irrelevant reasons); point being the safety implications are major. If a tank full of water ruptures, the pressure drops immediately without the massive release of potential energy that comes with a failure involving pressurized air, even at fairly low pressures. I like Concours' and others' techniques posted here. The decoy failure could have been a lot worse.
Having shared this, I will confess to taking a Honda gas tank up to about 15 psi on air to 'pop out' a dent. It actually worked and I got away with it.
Cold hydro on BCP & RCP loops.

Many people who don't work in industry are blissfully unaware of the various kinds of stored energy that is around us all. Properly managed, it's safe & reliable.
That's the reason I shared, to try and guide the "youtube"mechanics a bit.
 
 
I welded a tank recently and did my usual trick of using just my lung pressure and it was fine with no leeks
But because the tank wasn't for me and was going to have an expensive paint job
I rigged it up and gave it 4 psi
I found 3 tiny leaks
Only just visible with soapy water
 
So, I know my Atlas tank has a leak but it's small enough that over 25 years it's only discolored the paint and lifted some but never a drip. I know approximately where it is -- on the seam at the left front corner. Probably from vibrating against the frame because of flattened mounts.

The question is how to fix it. I'm scared to use the tank sealer I see advertised 'cause I've heard so many horror stories about it. A couple of inches of the paint will have to be redone but I'd like to save the rest, it's a stunningly good candy apple red paint job that was initialed by the painter.
 
Back
Top