Crack in the weld transverse to the direction of weld run 3

[rtiwari123]

Hi Al,

Can anyone take a look at the two pictures below and advise the cause for the fracture shown here.

It is an oil separator tank made from SA 455 material. To my knowledge there are no temperature extremes and vibrations in service. Material is 0.25" thick; it is code vessel. Thanks in advance.

 

[DVWE]

More than likely hydrogen cracking, but there are several other possible causes listed in this article. Transverse cracks are toward the end of the article. Make sure the weld filler metal was correct if at all possible. Another factor for induced stresses is the rolling of the plate, but unlikely at that thickness.

http://www.lincolnelectric.com/en-us/support/process-and-theory/Documents/weldcracking.pdf

 

[rtiwari123]

Thanks DVWE. Hydrogen cracking was my first thought too.. thanks for the Lincoln Electric article.

 

[EdStainless]

It would be interesting to look at the ID of the weld there. That is a high stress location and less than full penetration would make the stress much higher, though H is a likely cause weld quality could have aggravated it.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, Plymouth Tube

 

[r6155]

1)Excessive fillet weld
2)Check calculation of reinforcement required

Regards
r6155

 

[rtiwari123]

Thanks EdStainless and r6155.

 

[gtaw]

I would not attribute the crack to hydrogen unless the crack occurred when the vessel was relative new. The most critical time period would be with in the first 48 hours or so. As time passes, the hydrogen effuses into the surrounding atmosphere, thus the probability of cracking is reduced over time. Granted, a small hydrogen crack may have initiated and grew over time, but I would expect it to manifest itself relatively quickly.

I suspect there is a discontinuity in one of the initial weld beads. Perhaps there is incomplete fusion at the start of the weld bead where two beads overlap, possibly incomplete joint penetration in the root of the root bead, who knows unless the weld is sectioned an examined.

Transverse cracks can also occur if the weld filler metal overmatches the base metal. This can happen if a highly deoxidized filler metal is used on multiple pass welds on "clean" steel. The deoxidizers become alloying constituents if the deoxidizers are not utilized. As alloying constituents, the silicon or manganese increase the hardness and strength of the weld, i.e. the weld overmatches the strength of the base metal. The resulting residual stress in the longitudinal direction can exceed the tensile strength of the base metal and cause a transverse crack.

Having thrown that possibility out there, my money is a defect in the root bead.

Best regards - Al

 

[gtaw]

By the way, nice photograph of a transverse crack. I would love to have a copy of that for my presentation on welded defects!

Best regards - Al

 

[rtiwari123]

Thanks gtaw.

I have been told that most oil separation tanks (identical to the one shown in the picture and built by same fabricator) show the same defect. Don't know if the defects manifested soon after being put in the service.

The picture is not mine; however I will find out if I can get permission for use in presentation.

Regards, RKT

 

[LittleInch]

Perhaps someone just got the design wrong. If its a common fault that would tend to rule out manufacturing error as it would be difficult to get the same spot to fail.

This looks like the highest stress point. Pressure in service and number and size of pressure cycles would be good to know. when you have a hole that big that close to a dished end, the stress calcs can get a bit indistinct.

The issue might well be the weld if its much stronger than the base metal, but there look like other things at play here and maybe the weld is only one part of the story. - More data please!

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[DekDee]

"I have been told that most oil separation tanks (identical to the one shown in the picture and built by same fabricator) show the same defect "

And people are still buying them ?

 

[metengr]

Hydrogen cracking based on the orientation and appearance.

 

[rtiwari123]

Thanks LittleInch. But that is all the data I have.

DekDee - perhaps these separation tanks were all delivered at the same time for a project. Seems very unlikely that people will keep buying them knowing fully well of this problem.

Thanks metengr.

 

[chicopee]

Is this oil water separator a pressure vessel and if it is what is the name of the manufacturer, and did you get a copy of the manufacturer's data sheet from the National Board for information on the tank and its tower material, allowable pressure and temperature.

 

[weldstan]

Transverse cracking in steel weldments are most caused by hydrogen (vast majority of cases) or other local contaminant. It would be interesting to know the filler metal chemistry because one would not expect hydrogen cracking in .25-inch wall thickness welds even without preheat; however, they may have used a high strength filler like ER80S-D2.

The location of the crack in the weld and apparent extent of the crack into the base metal may also indicate fit-up problems during welding causing a localized high stress in the location. I would also check the root for a major discontinuity per EdStainless.

 

[rtiwari123]

chicopee - yes, this is a pressure vessel. But I don't know the name of the manufacturer, nor do I have the data sheet.

Thanks, weldstan.

 

[Compositepro]

That crack has opened-up pretty wide and extends into the heat affected zone so there must be a lot of residual stress. I believe that preheat of the vessel before welding or post weld stress relief may have been lacking.

 

[rtiwari123]

Thanks Compositepro.

 

[r6155]

No information, no documents: Discard this pressure vessel

Regards
r6155

 

[Compositepro]

The vessel does have a name plate on it in the picture. It has been painted-over.