troubleshooting v groove weld with backing ring 1

[dmh2819]

we are having an issue with a V groove weld and I'm looking for some help is solving the problem. We are welding .250 thick SA 53 E/B pipe shell to .188 thick SA 414 G elip head with an SA 414 A backing ring that is being left in place. Welding process is FCAW using .045 wire. 22.5 volts, 218 wire speed, using miller matic 252 with 75/25 shielding gas.

The problem we are seeing is if we grind away the backing ring, we see a line on both edges (shell and head). I'm not sure is this is lack of fusion or penetration. My welder says he can see the base metal being melted away. We have tried to increase the land from feathered edge up to over 1/8" with no change. We have turned down the voltage from 22.5 to 19, with no change. In the future, we may need to xray this type of joint on another vessel design and these lines will show up on the xray.

Any suggestions on the cause and possible corrections would be greatly appreciated.

 

[metengr]

Sounds like lack of penetration between the edge of the land and backing ring along the root. How deep are these lines? Have you tried blend grinding to determine their depth?

It could be that the welder is not getting enough access into the root location to fuse the land with the backing ring.

 

[stanweld]

When welding into steel backing strips/rings, they must be intimately fit uniformly about the ID circumference. The gap between the head and shell should be 1/4". The root face should be near 0.

 

[dmh2819]

Stanweld,
We started with no land (feathered edge) on both pieces. The backing rings come with a 3/16" spacer on them which has been the spacing we have been using. There is a tighter fit on the pipe (shell) side than on the heads since the heads are slightly thinner, but the issue shows up evenly on both edges.

I had the welder try it without doing any beveling of either piece and the 3/16" spacing to see if it was the size of the land not being enough (based on his input it looked like we were burning away the base metal), but the results were the same. Then we tried the lower heat setting, slight improvment but still there.

Metengr,
they are not very deep. I haven't measured them, but when we are grinding down through the backing ring to them, we need to be careful, or they disappear. They do show up on an Xray though faintly.

 

[stanweld]

I'd recommend increasing voltage - not decreasing.

 

[dmh2819]

OK Stanweld,
We'll try that.

 

[Duwe6]

What you are seeing is typical to backing bars and rings. If the RT shows that the root face is no longer straight, and is now slightly 'wavey', the root face has been consumed and no Lack-of-Penn or -Fusion LOP / LF exists.

The line you can see is from [I think] the weld shrinkage making the backing bar/ring 'squirm' and loose its intimate contact with the back side of the weld. I do know that those lines show up on UT and RT, and do not indicate LOP / LF.

One other thing: you loose the advantage of a backing ring if you decrease voltage, or spend the time grinding a prep to a feather-edge. You *increase* amps, and especially volts because you have the backing to retain the large 'wet' weld puddle. This is why rings are usually called 'chill rings' by the welders.