Building an aluminum gas tank

[Bart Carter]

I need to build a custom fuel tank for a kit car. It will be made of aluminum, have baffles, a sump with one-way flaps and a return column to swirl the bypassed gas from the fuel injectors. The tank will be .090 or .125 and will not have any bends.

A few questions for those of you that build gas tanks.

1. 5052 or 6061?

2. What rod would be better?

3. I would like to do the complete bottom first and then weld the top on. Any problem with welding the outside seams first and then weld them from the inside? I am looking for strength and to not have the possiblity of any leaks. I don't want to have to weld up any pin holes after the tank has been in use!

[precisionworks]

Bart,

The last diesel tank I made was from .100" aluminum
(6061) diamondplate. I had the plate cut & bent so that one piece made the bottom with both long sides attached, there was one flat piece for the top, and two small end caps.

The baffles are the first things to install, they provide some support for the structure. And they're lots easier to weld in before the top goes on.

The top will need a tiny tack about every 4". Tack one long side, then the other. Go back and run a continuous seam down each side. No need to weld the seams from the inside as long as you get full penetration from the outside. Try to get both edges to meet so that you have a 90* open corner to fill in, much easier than a lap joint.

4043 filler should work well, it did for me.

Holesaw the openings for the fuel filler cap before the ends are welded on, otherwise it's nearly impossible to remove all the metal shavings.

After both ends are tackes on and seam welded, use a bicycle pump to pressure the tank to 3 or 4 PSI. Mop a soapy water solution on all seams. Any tiny bubbles indicate a pinhole that needs to be remelted and filled.

My customer chose the diamondplate, which is a real PITA. Imagine driving your car down a smooth road, hit a bump, smooth again, another bump, etc. That's what it's like to weld this stuff. But the end result is killer.

If in doubt about anything, use more tacks

[dyn88]

ive never had success or heard of someon bending 6061 without cracking. 3000 series is what i would use.

[Bart Carter]

I thought about bending, but it is a pretty complex shape for the area it will occupy. That's why I want to cut each piece and weld it together. I know aircraft use 5052 for it's ability to bend and more strength than 3003.

But with no bends, would 6061 be better? Or does 5052 have more fatigue and crack resistance.

What would be best for day to day auto driving over the usual roads with some off-road thrown in.

[Richs]

Been there...done that! 3003 H14 (half hard)...

Is the only way to go. Triangle Engineering, Jaz, etc all use 3003. Easy to work and weld and has more strength than you'll ever need. Try to make as many bends as possible and as few welded seams. Make patterns outta poster board, get your bungs and fittings from Chassis Shop or others. Use 4043 rod. Aluminum work hardens so if you fudge a bend you'll have to anneal it to straighten it out or it will be crack city. Anneal it by sooting up the mistake with an acetylene rich flame, then add oxygen and heat up the soot until it burns away. Then pound out the mistake. If you are content with not making bends then put some beading on large flat areas. 3003 is also quite corrosion resistant.

[arcalaska]

fuel tank

5052 and 5086 are the only grades I weld up here, the 6061 that I weld is usually extrusions, so I've never had to worry about cracking when you break "bend" it. I've never had 5000 series aluminum crack due to breaking "bending".

[Bob La Londe]

I know this is an old thread, but .125 5052 will definitely crack when bending to 90 in a brake. No issues bending up to 45. Maybe even 60. If practical I'll run two 45s about 3/4 inch apart if I need a 90. If I have to have a sharper 90 I'll anneal after bending to 45. I don't recall the temp, but 600F sticks in my mind. I am pretty sure I have a 600 F templestick in my tool box so I can use a torch and templestick along the seam. Others I have heard of is set your torch to a sooty flame, soot up the area, and then set a rosebud to a clean flame and warm it up until all the soot burns away. Never tried that.

I only felt it necessary to revived this zombie thread because I was doing some searches on various alloys for making gas tanks. I figured if I am still finding a 12 year old thread somebody else might too.

On the flip side I've made 90s with my brake in .0625 5052 and never notice even a stress line much less an actual crack.

Also a trick I have used to get "rounder bends" in .125 5052 is to move the fingers back about 3/8 inch and stack about 1/4 of solid cardboard sheets (like the back of a notebook) on top of the aluminum. This give a radius bend and I've done it without cracking. Regardless you need to go to near net angle in one operation. It is still work hardened. Not all brakes have that much adjustment. Another trick to do the same thing is to buy or make some radius edge fingers.

[Willvis]

Thats not true at all. Maybe on a finger brake (glad I didnt buy one) but it all depends on the radius you are bending. Even 6061 can be bent without anealing if you use a large radius.

Here are a bunch of "boxes" I bent a few days ago out of 1/8 5052 h32. I worked at a place that bent this stuff constantly on a brake.

[Bob La Londe]

So you don't bend at all or you have a big press with press dies of your own in your shop? (Not your employer's shop.) AND I noted how you can bend it without annealing by bending with a radius or progressive bends IN A FINGER BRAKE. What exactly are you disagreeing with? Nothing I said is untrue. Do you just dislike finger brakes? Even they can be setup to bend a radius if you know how and I took the time to tell people how.

Here is another way I have done:
I've taken a section out of a piece of steel pipe slid it over the fingers on my brake and used it the same way as if I had radius tipped fingers. There is nothing inferior about that except that most brake just can't bend material that thick full width without distorting.

My Tennsmith finger brake can do the job (just barely), and it cost an order of magnitude less than a big press.

The fact is if you want a fairly sharp 90 in .125 5052 you will have to anneal it about half way through the bend or it will crack. What I said was "If I have to have a sharper 90 I'll anneal after bending to 45."

[Willvis]

I know this is an old thread, but .125 5052 will definitely crack when bending to 90 in a brake.

This is what I was disagreeing with

[Franz©]

Just wondering out loud here; Is a sharp 90 really a good idea on an aluminum tank, particularly a fuel containing tank?
My limited observation of fuel tanks both steel and aluminum shows most employ radiused bends rather than sharp bends.
I'll also throw in I have seen 2 instances of corner crack leaks in oil tanks on used oil burners that resulted from shipping.

[FusionKing]

I know this is an old thread, but .125 5052 will definitely crack when bending to 90 in a brake. No issues bending up to 45. Maybe even 60. If practical I'll run two 45s about 3/4 inch apart if I need a 90. If I have to have a sharper 90 I'll anneal after bending to 45. I don't recall the temp, but 600F sticks in my mind. I am pretty sure I have a 600 F templestick in my tool box so I can use a torch and templestick along the seam. Others I have heard of is set your torch to a sooty flame, soot up the area, and then set a rosebud to a clean flame and warm it up until all the soot burns away. Never tried that.

I only felt it necessary to revived this zombie thread because I was doing some searches on various alloys for making gas tanks. I figured if I am still finding a 12 year old thread somebody else might too.

On the flip side I've made 90s with my brake in .0625 5052 and never notice even a stress line much less an actual crack.

Also a trick I have used to get "rounder bends" in .125 5052 is to move the fingers back about 3/8 inch and stack about 1/4 of solid cardboard sheets (like the back of a notebook) on top of the aluminum. This give a radius bend and I've done it without cracking. Regardless you need to go to near net angle in one operation. It is still work hardened. Not all brakes have that much adjustment. Another trick to do the same thing is to buy or make some radius edge fingers.

I find this very interesting.
I have thousands of dollars worth of work bent each month from 5052. We mainly only use .100 or less. Normally 10 ft or sometimes 12ft.
I generally specify an inside radius on 3/16ths and 1/4''. It will tend to get surface cracks that have to be addressed or remade if bent too tight.
Lately I have had a pretty fair amount of 1/8th bent this last couple of months and haven't had any issue. In fact I don't recall ever having an issue with 1/8th.
All press brake with 5052 pvc coated.
If we had to anneal or go with 2 45's it would be cost prohibitive. I'm gonna go out on a limb here and say that I believe that if you worked on setting up different somehow you will find that you can bend 1/8th 5052 quite sharp in your shop also without heat and all that. You probably need a radius on your fingers. It may be that simply clamping one side is the problem. A press brake allows both sides of the bend to slide.
5052 isn't even heat treatable, so not sure you are really accomplishing much there anyhow.
I anneal 6061 effectively to soften but 5000 series is a whole different story cause it comes back hard. I work it hot when I use heat on it and let 6061 cool then work it.
Just to clarify, we are basically a pontoon log body shop. We straighten and repair extensively daily use whatever methods needed whether it be by heating and straightening or simply patching or by hammer and dolly.....you name it. They are made from 5052 exclusively.

[MMW]

One other thing to think about is bending with or across the grain. Bending across the grain you have less chance of cracking.

I also agree that you can bend 5052 without cracking by bending a proper radius. When someone says "sharper 90" that doesn't specify how sharp. In a press brake I would use 7 to 8 times material thickness for a bottom die opening, top die with a radius equal to material thickness and bend against the grain. Sometimes however you just get cracks.

[lonewelder]

http://www.cumberlandmetals.com/alum...um-bend-radii/

note above, minimum bend radius

https://www.clintonaluminum.com/wp-c...num-alloys.png


i have bent a lot of material. in industrial job shops, precision and hvac. at the precision end and to some degree the industrial we generally had simple equations we worked with for determining setbacks. variables were the material thickness (t) and the radius of the top die. factored in was the type of material, which in this conversation is aluminum. see the above link.

i have annealed t6 on the bend line but only out of necessity and that is the only al alloy that i ever did that too.

regarding with or against grain that was often determined by size of part or size of sheet in stock or both.

i have bent on a finger break and would bend a 2 x 2 x 1/8th pc of material and then mic the bend from both directions to see what amount of gain there was then adjust accordingly. a finger brake could not be counted on for consistency over different lengths though so you just gave it your best shot.

setback is the amount of material deducted for each bend. this number is not that
critical on light gauge sheet, hvac for ex. but factors heavily on your thicker material.

[griff01]

Ummm. "T6" is not am alloy designation.

Griff