Weld Cracking In A Stainless Water Tank

[superdupertrooper]

I have a problem with a weld cracking on a 303 stainless tank. At the base of the tank is 42 1/2" by 13 1/8". Along the longer edges of the tank there are 3/4" flanges that protrude from the base of the tank to the inside wall. Along the shorther edges, there are no flanges and it is welded from the outside and ground down.

We have had many reports of these tanks cracking in the field and leaking water, but as of now have only recieved one tank, and there is a clear crack, centered in the middle, along one of the 13 1/8" edges. The crack runs right along the weld material.

Statically the tank holds about 600 pounds of water, and the material thickness (14 Gauge) is sufficient to hold this weight without any sign of failure.

So the current conclusions are:

1. That it is a design issue. Being that the crack occured on the edge without a flange, there is poor weld pennetration in this thin material. (but we only have data frome one tank...)

2. It is a welding issue. Maybe there are some things that we could change in the process that would eliminate this problem (like welding the seam from inside the tank or not grinding the weld).

So what is everyone's take on this issue? Your input is greatly appricated.

 

[btrueblood]

Well, for starters, 303 is not considered a weldable alloy.

 

[superdupertrooper]

I'm sorry it's 304 for the welded sections. The tank cover is specified to be made out of 303. It's dumb that our drawings specify two different materials for this tank.

 

[Goahead]

304 is not a good selection for welded sections. The filler metal used should also be considered. Usually 304L or 321 are preferred as base metals and 308L for filler. But it should be first determined that corrosion is the cause of the cracks.
If metallographic examination could be performed one would know.

http://www.welding-advisers.com/

 

[superdupertrooper]

The tank that we recieved was used for like 2 months when the failure occured. The time for failure for other tanks has varied.

I agree that 304L and 321 are better choices for material in this application because of how they respond to the heat of welding.

 

[EdStainless]

Check the carbon, it might actually be 'L'. There is very little straight grade amnufactured.
Filler or autogenous? Is it the correct filler?
Check some micros, is there sensitization? If so you could get CSCC just sitting there.
Is the crack actuall in the weld? or off to the edge? If it is smack in the middle, and there is good penetration, then I would suspect centerline shrinkage or porosity.

= = = = = = = = = = = = = = = = = = = =
Corrosion, every where, all the time.
Manage it or it will manage you.

http://www.trent-tube.com/contact/Tech_Assist.cfm

 

[superdupertrooper]

The crack itself is right in the center of the weld. Nothing is evident off to the side in the heat affected zone.

I don't know precisely what our fabricator is using for a filler material, no material is specified in our drawings.

As for weld quality, there is no evidence of porosity within the weld that cracked. It looks like weld pennetration is real good as evidence of the 1/4" wide blue marks along each side of every weld seam.

 

[unclesyd]

Centerline cracks in a weld bead are indicative of a welding problem, either the filler metal itself or not enough filler metal. Centerline cracking is very prevalent in autogenous welds as mentioned above.
I would try to ascertain the exact weld procedure.
A lot of times if lack of weld metal is the problem and you can see evidence of the problem in uncracked areas of the weld. Also look at the cracked area to see if it tends to be sunk in.
The blue line really doesn’t indicate very much other than it's a tell on the welder's uniformity or precision in welding. It tells about his speed and technique.
There should be no problem welding 304 SS with 308, in fact I've never seen a problem with the alloy combination only the welder.

There is another possibility:

Where are the tanks used?
Are they subject to vibration?
Is it possible for the side plate to act like a drum head?

 

[rolledalloy]

Is the water at room temperature or is this a water heater tank? I have seen water heater booster tanks made of 304L sheet stress corrosion crack (SCC) due to chlorides on the inside. They also have cracked from the outside due to foam insulation applied to the outside containing chlorides.

Easy check for chloride SCC is through metallography. Stress corrosion cracks will run transgranularly and are branching in nature.

One centerline crack in a weld though does not sound like SCC, but you never know.

 

[SMF1964]

Rolledalloy: SCC will run transgranularly except when it runs intergranularly, and branches except when it doesn't. ;-)

Superduper-: I agree with Unclesyd that centerline cracking in the weld points focus to a welding problem. Metallography to look for oxides in the crack (lack of fusion for an autogenous weld), hot tears, etc.

 

[superdupertrooper]

The water tank is at room temperature.