What type of cracks are these? 2

[brandonb73]

The piece of plate in the photo came from a Sec. I steam boiler. This is a section of the longitudinal seam. My questions are: what kind of cracks are these and what caused them?

 

[kelowna]

Difficult to say, are they cracks at all? I do not think you can get an informed opinion from a photo. It looks like a weld deffect, lack of fussion? some porosity?

Do you have access to a welding inspector/engineer?

 

[brandonb73]

They are definetly cracks, we had RT and UT performed on them. I find it strange that the cracks jump from one side of the weld to the other. The total length of cracks on this seam was about 51" which is more than half the seam. Would these be considered toe cracks. I thought about cutting a cross section and doing some PT.

 

[metengr]

In steam drums, the girth and long seam welds that contain undercut are sites for local corrosion attack in service, below the water line. So, I would suspect that weld toe undercut was the cause resulting in local crevices that promoted corrosion attack in service. If these crevices are not removed this will or has eventually resulted in corrosion fatigue crack propagation.

 

[metengr]

brandonb73;
I would locally excavate the cracks to removal and perform weld repairs using Welding Method 1 of the NBIC. This is a straight forward repair in these kinds of steam drums. Make sure that when weld repairs are performed you blend grind the weld toes of the repair weld flush with the drum ID surface to avoid undercut and crevices.

 

[MikeMet]

To get to your original question, they are cracks in the toe or fusion line of the weld. What caused them can only be determined by getting a sample with the crack analyed at a lab and evaluating the operating history of the boiler. From only a picture they could be fatigue, corrosion fatigue, original fabrication defect such as LOF, delayed hydrogen cracking etc. As for crossing the weld, cracks in many cases follow lines of stress and can cross the weld bead if the stress directs it to do so. I would not repair without at least doing some field metallography to try and characterize the cracking and determine some sort of cause. If you repair as metengr described without knowning the cause, the cracks can easily return and you are back where you started.

 

[metengr]

I would concur with MikeMet's[b/] recommendation by removing a small boat sample for analysis, if you are unsure of the cause. Regardless of the lab analysis results, weld repair must be performed in accordance with the NBIC.

 

[brandonb73]

One thing I did not mention, which could play a part in this. When we first visually examined the seam, we could tell the rolling was poorly done. There were flat spots of approximatly 3-4" on each side of the seam and the seam actually met at a point (see photo) about 1/4" higher than if it had been rolled correctly. Could this have created a high stress concentration as the boiler expanded and contracted which was further accelerated by undercut on the weld?

 

[metengr]

Yes.