How to test a newly welded fuel tank for leaks?

You have a lot of questions there. I would pressure test by introducing air at 5 psi. A spray of soapy water will show any leaks. Yes tack welds are a possible site for leaks. It would be best if you grind your tacks feathering them at both ends before welding. Lap joints would be easier to do. I would power wire brush with a stainless wire wheel on a small angle grinder. Do this to all edges before assembly then just prior to welding. Don't get forceful wire brushing. What you are doing is removing surface oxide. Don't weld on top of a leak. Grind it out so that you are welding about an inch or more over the area. Grinding/sanding wheels for aluminum are available. I prefer a skill saw with an aluminum cutting blade to dress and shape aluminum. It cuts clean and leaves no dust to contaminate. This takes a high degree of skill and proper safety equipment. Randy

Questions:

The bottom of the fuel tank is to be made from three pieces. The main piece is 51" long and is bent in the shape of a trough. The front edge is about

Could I run a bead on the inside and outside of each corner?

The top of the tank is filet welded along the length of the 51"x1/2" flange on the front and rear. I have heard/read that tack welds can become a source of leaks. How often should I place a tack along the 51" weld? How can I reduce the amount of distortion that might show through to the top of the tank as it will be exposed?

What is an acceptable method to check for leaks? If a leak is found is there anything special I would need to do to seal the leak? Can I go back over the area with the electrode?

I know some might be thinking I should just hire out the welding and believe me I am not opposed to that but would like to give it a try if it's within my skill set.

Thanks.

Ebby.

-- John "Ebby" Ebensperger Hatz Classic s/n37 Camden, NY

I have welded up a bunch of fuel and water tanks for boats.

BTW you can make weldable pipe fittings by simply buying aluminum pipe couplers and cut them in half. Punch a hole in the aluminum sheet slightly smaller than the fitting and weld the half coupler over the hole from both sides. Do this before assembling the tank.

The fewer seams the better.

Not necessary, As long as you run a good consistent outside fillet weld, you don't need an inside fillet.

Wire brush inside and out, and feather the edges of the tacks with a file or flap wheel.

Start on the ends and the middle. Then start splitting centers until you have a tack every 2 inches or so.

If you don't have a pulser then you should do Back-Step Skip-Stitch.

Picture a seam with marks one inch apart marked 1 through 10. Back-stepping would be welding from 1 to 10, but going backwards between tacks. You weld from 2 to 1, then 3 to 2, then 4 to 3 and so on. Skip-stitching would be welding every other segment across the seam, then coming back and welding the remaining segments.

So 2 to 1 then 4 to 3, then 6 to 5.....

Then 3 to 2, then 5 to 4, then 7 to 6...

This will break up the heat and help reduce distortion.

Go to a hardware store and make up a test fitting. The test fitting is a air hose connector screwed into a small air regulator. The outlet side of the regulator goes to a ball valve then a T fitting. The T fitting is used to mount a air pressure gauge. The other side of the T fitting has a short nipple that screws into a fitting that will fit in one of the openings in the tank. Plug all the outlets with pipe plugs. Use one to connect the test fitting.

Leak check with soapy water until you have eliminated leaks in the pipe plugs. You can pressurize the tank to a pretty accurate pressure and then close the ball valve.

The rule is 3 atmospheres, but 5 PSI is close enough. The tank should be able to hold 5 psi for 24 hours.

Best of luck. Remember to use 5356 filler for corrosion resistance.

Wow I will yield to the pros on this one! The Squarewave I have doesn't have a pulser. This is a critical piece and I've got about $350 in parts and fabrication costs so the prudent thing would be to hire a pro.

Ernie, thanks for the virtual welding lesson. I will practice my aluminum welding on the scraps I have left over. If something less critical comes along I will attempt it. Perhaps a smaller header tank. I am a hobbiest and this may be the only aluminum I need to weld. PS: I will give my 5256 to the pro.

Something else besides your fittings... You didn't mention baffles. They prevent fuel surging back and forth as well as strengthening the tank. They are best if the ends are formed flanges only welded at the edges of the flanges to the side walls of the tank to prevent stress points and allow some flexing. A "pro" will mention the need immediately and he will know where to locate them. Randy

Ernie, thanks for the virtual welding lesson. I will practice my aluminum welding on the scraps I have left over. If something less critical comes along I will attempt it. Perhaps a smaller header tank. I am a hobbiest and this may be the only aluminum I need to weld. PS: I will give my 5256 to the pro.

Right you are Randy. The plans do show two baffles. I will make the rest of the parts for the tank and have the pro do the welding.

It is always safer to hydrostatically test because there is no stored energy.Even 5 psi of air could be very dangerous if the tank is large enough and lets go. If he uses air he definitely should have hearing protection because a rupture could cause permanent hearing loss.

I freaked when I was told what a local fabricator had to do to qualify for the United States Coast Guard. He had to basically blow up a gas tank as a demonstration. The seams had to hold. A light gauge tank often will swell and buckle at pressures around 15. I have seen enthusiastic testers damage tanks during pressure tests. The seams held but the side walls were damaged. 5 psi is more than enough for a soap bubble test. Randy

It is always safer to hydrostatically test because there is no stored energy.Even 5 psi of air could be very dangerous if the tank is large enough and lets go. If he uses air he definitely should have hearing protection because a rupture could cause permanent hearing loss.

That's good advice safety-wise, but if you do it to something the size of a fuel tank you'll inflate the thing into a sphere. It is very easy to get a surprisingly high pressure up with a hydrostatic test, even on a hand pump.

As it's only a containment tank here, not a pressure vessel or boiler, then I'd submerge it under water, blow air (or CO2 from a tank) into it _through_a_regulator_ and only bring it up to a very low pressure, so as to avoid distortion. You're looking for pinhole leaks here, not failures due to over-stress.

If you have to contain explosions in fuel tanks, then that's a whole different design problem. Nor is it particularly relevant - fuel tanks don't contain flammable mixtures (they're too rich) and just don't explode in practice. It's a leak into the bilges that kills people, not a fuel tank that finds some improbable new way to light off.

Scare yourself. Look up the "Elizabeth Brim" forging technique 8-)

The term hydrostatic does not mean nor does it infer the use of high pressure The media, be it a gas or liquid will create the same amount of stress.

It is very easy

That depends on the effective area and the mechanical advantage of the pump. It's also what pressure gages are for.

Try submerging a truck gas tank. Snoop is a lot easier.

Only when they are full. If there is air, the potential for explosion always exists.

It's a leak into the bilges that kills people, not

No need. While working at Ashcroft Gauge I saw enough images from on going litigation to last a lifetime. We would routinely perform a 6000 psi oxygen test on tank gages by prepping them with a special dirty oil formulation concocted in the chem lab. After you see the shrapnel from the bourdon tube imbeded into dial face you will never look at another tank gauge when opening a tank valve again.

The way I test microwave wave guide lines for leaks is to use a freon leak detector. If you have access to a electronic detector, they make spotting leaks easy. The way I do it is purge the wave guide by blowing some r22 through the guide. Then cap each end. I have a fitting in the cap that allows me to put about 1 pound of pressure in. The leak detector will find an extremely small leak of any freon.

John