Fuel tank welds cracking

[72bwhite]

I made a fuel tank out of 5052 .125 aluminum used 4043 filler
and it has a leak after just a few months
I don’t have a brake that could bend it so bottom and sides are one piece
tank is 28 in long 16 in high 16 in wide.
I used T type joints the bottom sticking out a little past the sides.
tank is for a Catalina 36 sail boat so will probably start over and make new tank
and not try to fix the leak.

original tank had one end that sat in the bilge took it 30 some odd years to start leaking
I modified the new tank so it stays dry like they did on later boats.
what did I do wrong?
Options try again with 5052, use 6061, or 304 stainless

 

[Bondo]

A description of where the leak is, where the leak is in comparison to the boat (sitting in water or crud or whatever), and a general understanding of prep work into it is really needed to assist you.

Stainless will always be the best bet when it comes to marine stuff especially when touching steel.

 

[72bwhite]

Can’t see exactly where crack is but appears to be on the bottom joint.
was new clean saw cut sheet cleaned with stainless brush.
it only touches fiberglass and rubber and is dry
salt water so 304 tends to rust

 

[seiner]

I've built quite a few tanks out of 3/16 5052. I use 5356 3/64" wire for mig and 3556 .125 tig.
I always try to design the tanks so i can weld inside and out on as many seams as possible, or at least the bottom. I TIG the external corners. I think .125 is too thin for a boat tank. going to 3/16 is cheap and makes everything easier.

I also do a pressure test with water at 10psi or 24' head pressure. do not use air. if it survives the head pressure test it is good to go. you probably just mucked up a weld and now it leaks.

clean the tank and do the pressure test, mark all your problem spots with a center punch. unless it was used for gasoline go ahead and clean the interior, car wash works well to steam it out and then fix your leaks. for gasoline find advise from someone else.

or build a new one using 3/16 and don't forget to pressure test it. Fixing a leak before you install it is way easier than after.

 

[seiner]

A description of where the leak is, where the leak is in comparison to the boat (sitting in water or crud or whatever), and a general understanding of prep work into it is really needed to assist you.

Stainless will always be the best bet when it comes to marine stuff especially when touching steel.

Stainless has it's issues and there are so many types that I would not make that blanket statement . I've seen 5" diameter shafts corrode in half from from low oxygen conditions and not from galvanic corrosion. 3/16" 5052 is the correct material for the OP's tanks.

 

[Bondo]

You first said you built a tank from .125, from 5052 using 4043 wire. Then you say you have built tanks from .190 of 5052 using 5356. What changed your mind on the filler for this one?

 

[sfriedberg]

You first said you built a tank from .125, from 5052 using 4043 wire. Then you say you have built tanks from .190 of 5052 using 5356. What changed your mind on the filler for this one?

Pay attention to the names, Bondo. 72bwhite is the OP. seiner is the 5356 guy.

72bwhite, 5356 might be slightly better filler than 4043, but 4043 is supposed to be OK for 5052. It's not OK for other 50xx alloys, so maybe what you got [for sheet/plate] was not what was on the label. The other thing is that you may not have used enough filler! If you were being real minimal about filler, you may have run into the problem that 5052 (like 6061) cannot be welded autogenously (with its own material). So that gives you two things to try: change of filler rod, and use more filler rod to dilute the alloy (I believe the problem element is magnesium) in the puddle.

 

[drcoelho]

It would also be interesting to identify where the cracks are showing up. In the weld itself, along the edges of the weld, etc....might even suggest you do a clean cut and analyze / confirm you are getting a quality weld.

 

[Fish On]

You first said you built a tank from .125, from 5052 using 4043 wire. Then you say you have built tanks from .190 of 5052 using 5356. What changed your mind on the filler for this one?

I think it you're confusing someone else's post with the OP on the second one...

But, to the OP, 5356 is your filler of choice on 5000 series alloys in non high temp environments. 5052 has a lower mg level than 5083/5086 (where 4043 isn't an option) so 4043 is marginally acceptable in some cases, but 5356 is definitely the preferred option, especially in corrosive environments. I don't ever use 4043 on 5052 alloy.

This is about as far as anyone is going to be able to help without seeing pics to get a better idea if how/where the crack is forming.

I don't have issues with 1/8' material for fuel tanks - I've built plenty out of such, but 30ish gallons is a little on the large side for that, especially if the tank was fully open inside. Did you have a baffle inside? Did the original?

 

[BT Fabrication]

Can’t see exactly where crack is but appears to be on the bottom joint.
was new clean saw cut sheet cleaned with stainless brush.
it only touches fiberglass and rubber and is dry
salt water so 304 tends to rust

first off, lets see these welds and how you welded them, it will tell us much more in a picture then just words.

 

[72bwhite]

Well appears I got a minimum of 2 things wrong used wrong filler
and probably not thick enough, and more filler.
still have to drain and remove tank

 

[Bondo]

Oops, that didn't go well lol

 

[mymachinery]

Could be stressed joint, oil canning causing cracking at the root. My customer had a Cat hyd tank (steel)from an excavator leaking at a cracked seam, while they were waiting for me to weld it they bought a new tank that started leaking a month later. I cleaned out the welds and rigged it with larger weld, so in that case it seems like the weld size was insufficient to resist the flexing.

 

[Strostkovy]

I tend to see cracks from poor mounting more than anything.