Carbon Steel Airtank ripped open at the Weld

Date 02-19-2007 12:50 Edited 02-19-2007 16:13

Hiya guys!
We use air tanks that are MIG welded with CO2 gas (Not manufactured by us) for pneumatic application in our automobiles. Max pressure these tanks reach is no more than 170 PSI. Recently one of these air tanks just ripped open exactly at the weld while in service. I have to make a failure analysis. Im not a welder so, pls help. I've posted 2 pictures of the ripped open airtank here:[img=http://aycu37.webshots.com/image/11916/2003238230274757809_rs.jpg]

AND here: [img=http://aycu06.webshots.com/image/11365/2001545452249225150_rs.jpg]

I want to know:
1. What could have triggered this problem? Was the weld or base steel used of bad quality since its ripped open exactly at the weld?
2. What is the standard testing procedure for quality checking new tanks?
3. To what factor of safety should it be built?

Thanks in advance guys.

By pipewelder_1999 (****)

Date 02-19-2007 15:36

Is there a weld at the longitudinal seam ? I can't tell by the pictures. The circumferential seam would be unlikely to be area where the failure started but it could be possible.

Gerald

By ownmind (*)

Date 02-19-2007 16:05 Edited 02-19-2007 16:15

Yes! there is a longitudinal weld as well. I just changed the picture. Pls do check the link again. Thanks

By MBSims

Date 02-19-2007 17:55

Not a good picture there of the long seam fracture surface. Possible cause could be incomplete penetration or throat thickness in the long seam weld. Suggest using ASME VIII Div. 1 calculation for minimum thickness of the tank and apply a joint efficiency factor to the weld depending on what NDE was performed.

By jon20013 (*****)

Date 02-19-2007 18:18

Marty, I agree. In trying to evaluate the condition from a less than ideal photo, it appears to me that there was no penetration on the longitudinal weld (if in fact it was supposed to be a full penetration?) and on the circumferential, there also appears to be lack of fusion, this based on the *clean* break line of the weld image. It's hard to properly judge but thats my read on what I see.

By hogan

Date 02-19-2007 19:34

did the shell overlap the heads at the circ welds?

By jon20013 (*****)

Date 02-19-2007 20:02

It sure looks that way, doesn't it? Wonder where the hell that thing was manufactured????

By CWI555

Date 02-20-2007 04:35 Edited 02-20-2007 04:50

My two cents worth.

bear in mind that the views are limited, and therefore the accuracy of my statements is limited.

That tank looks like it was made from a piece of pipe, with two circ welds to tie the heads in.
The Long seam appears to be LR-ERW welded or EFW welded.
It appears to have given up the ghost on the long seam and ripped open. I suspect selective seam corrosion.
If this is the case, the amount of pressure that it could take would be reduced, in which case you may in
fact not have overpressurized but still produced enough pressure for this brand of failure. The left corner
of the 7809 full size picture shows signs of this brand of corrosion.
As for the circ welds themselve the picture looks like the circ welds did not penetrate into the heads themselves.
(you can see the affected clean edge of the heads in both pictures.

In short, I think it gave up the ghost on the long seam, and the circ welds were in turn torn open by the explosive
release of pressure. Without having the piece in a lab and more direct pictures, that's my read.

An interesting article on ERW http://www.kiefner.com/ERW.PDF

Regards,
Gerald

By Bill M

Date 02-20-2007 16:13

I had the same initial observation as Hogan...is that a single fillet weld lap joint at the head to shell?

Whats the w.t. of the head and shell? Whats the inside dia.?

Does this vessel have any ASME stamping or nameplates?

By aevald

Date 02-21-2007 07:05

Hello ownmind, I would have a few additional questions. Does this tank have a pressure relief valve installed on it? and has it been determined that this valve was working properly or set to the proper relief pressure? The tank definitely looks rather light duty for the 170 psi. that you say that it could see in service, however, since it is relatively small in size and diameter, one that is properly constructed and welded should stand up to this service. Do you make any allowances for clearing moisture from these, as they will end up having condensation in them as they are used? and not clearing this moisture on a regular basis can cause corrosion issues that could cause them to burst. Also, do these tanks bear a pressure vessel stamp? Being used on a motor vehicle they may very well require this sort of documentation. If not you may want to go a different route. Just my $.02. Regards, aevald