Not logged inAmerican Welding Society Forum
Forum AWS Website Help Search Login
Up Topic Welding Industry / Technical Discussions / Remelting HAZ during rework of inconel 600 tubing
- - By Kix (****) Date 01-09-2017 18:15
A run down of the process is Orbital autogenous GTAW welding of half inch .049" wall  Inconel 600 tubing. There were some rejections for porosity in some of the welds. The existing weld was cut in half and refaced back to the HAZ removing all of the weld. The seam was then put back together and re-welded totally incorporation the old existing HAZ into the new weld. The concern is, how much different will the microstructure of the new weld be compared to the old weld and will the mechanicals be that dramatically effected. No pre or post weld heat treat is undergone in this process. I know you can't possibly know for sure without testing, but what are your thoughts behind the theory of what might be happening?

Kix,

Ray C.
Parent - - By 803056 (*****) Date 01-10-2017 04:30
Do a full section tensile test. The focus should be whether the joint is as strong as the adjoining base metal and whether it meets the required mechanical properties of the base metal.

The next question is why the nickel alloy is being used. More than likely it is because there is a corrosion issue. Chlorides would preclude the use of austenitic stainless steel for long term use, so seawater perhaps? That being the case, subject a couple of samples to accelerated saltwater corrosion testing.

I suspect there will be no issues, but as you said, without test you cannot be sure.

Best regards - Al
Parent - - By Kix (****) Date 01-10-2017 12:33 Edited 01-10-2017 13:42
Thanks for that Al. The fluid coming into contact with the material is a glycol mixture at about 10psi. I understand that the best warm and fuzzy will come from some full section compare and contrast tensile pulls. However, in theory, if I face the weld back say past the fusion zone into the fine grain zone of the HAZ and re-weld, wouldn't all of my alloying elements still be present and the re-liquidfication of all the remaining HAZ through the same original welding process produce a grain structure in the weld similar if not the same as the original weld with all alloying elements present? The HAZ of this orbital weld is vary narrow. I'm almost positive that we are facing back past the HAZ, but overall dimension is a huge concern so they can only be reworked once. I am just being inquisitive about the what if's ya know. :wink:

Here's a macro of the weld and a what if scenario. The dark arches in the center of the weld are the tail out pulses from the orbital welding process.


Kix
Parent - By 803056 (*****) Date 01-10-2017 14:07
It is an autogenous weld shielded by argon, I wouldn't expect any change in chemistry.

I would not expect any problem with a reweld since there would be little difference from a multiple pass weld. If a multiple pass weld would be acceptable, I would expect the repair weld would be acceptable. The difference here is that the HAZ from the previous weld will be completely remelted and consumed by the weld. In essence, the HAZ would be "virgin" material and it should have the same grain structure as the current HAZ.

If this was a ferritic stainless, where knifeline attack (corrosion) could be an issue, I wouldn't be so sure of this repair.

Right at the moment I have to leave, but I'll check a couple of my references when I get back this afternoon.

Best regards -Al
Parent - - By js55 (*****) Date 01-10-2017 17:46
if you have completely re-melted your HAZ then you have a pristine weld just as you did originally. Once a metal goes liquid EVERYTHING that existed before is gone.
Parent - - By Kix (****) Date 01-10-2017 18:29
That's theory I am supporting, but I wanted to see what others thought. I had a metallurgist telling me otherwise, but I'm thinking there might have been some confusion going on with what exactly I was asking.
Parent - - By js55 (*****) Date 01-10-2017 19:55
Our assumption is the correct one. He must have misunderstood.
Parent - - By 46.00 (****) Date 01-10-2017 22:40
Are you sure?
Parent - - By js55 (*****) Date 01-11-2017 14:01
As long as the HAZ has been completely remelted. Carbides are broken down, intermetallics, if existent, are dissolved, and elements are re-homogenized. Nothing of a pre-existent structure will last the re-melt. There is no different chemistry between the original material and the HAZ. Once remelt has happened it will be the same. The remaining HAZ will be as the first HAZ.
The problem with multiple passes or welding regimes is compounding HAZ's or increasing the time parameter for heat regimes of existing ones.
Parent - - By js55 (*****) Date 01-11-2017 14:04
Once you cut out the existing HAZ you have pristine material again. Conjecture cutting the weld out 3" back from the joint on both sides. Would we be having this discussion?
Parent - - By 46.00 (****) Date 01-12-2017 00:39
I am sorry, maybe I am missing something in the dialog here. You talk about re-melting the HAZ (which by definition would create another HAZ?) then about cutting out the HAZ? Which is it to be? What about 'boil off' of alloyed materials? What about specific heat treatment/solution treatment of original material? I really think we are talking about different scenarios here!
Parent - By Kix (****) Date 01-12-2017 16:30
The existing HAZ from the 1st weld is not getting cut all the way out. See picture above and where the new weld seam face line is in yellow. When you put the yellow line on the left and the yellow line on the right back together to make another weld seam after the 1st weld has been cut out. There should be no boil off of an alloying elements in the HAZ because nothing ever came to a boil. The solution/heat treatment is lost in the weld and in the HAZ of the 1st weld. It will also be lost in the weld and the HAZ of the re-weld. You would be moving the HAZ and it's structure over to the left and right into what was virgin unaffected material during the 1st weld. Hopefully that's kind of explains it.
Parent - By js55 (*****) Date 01-12-2017 20:05
Boil off what? You won't boil off any metallics.
Parent - - By js55 (*****) Date 01-12-2017 20:08
But the confusion is duly noted. Yes, when re-melting the existing HAZ structure you will, by welding, create a new HAZ. So what? Its in essence a new HAZ just like the first one. If you have concerns about the new HAZ then you damn well better have concerns about original HAZ's because they are the same.
The main operative here is if the original HAZ is completely re-melted.
Parent - - By 46.00 (****) Date 01-12-2017 22:35
Thanks, my next question would be the use of the word 'autogenous' which to me means dilution, two or more 'autogenous' welds on the same HAZ means more dilution with no addition of stabilisers/de-oxidizers or alloys etc. Is this right?
Parent - By js55 (*****) Date 01-13-2017 13:07
There is no dilution in an autogenous weld. There is only melting. Dilution is a term used when weld metal of a different chemistry mixes with the base metal.
Parent - - By js55 (*****) Date 01-13-2017 13:17
I would also say that if you wish to take this discussion deeper then a call to specialty alloys or rolled alloys would be of benefit. They are far far more knowledgeable in all of this than I. Those guys are top notch and they know these alloys inside and out.
Parent - - By 46.00 (****) Date 01-14-2017 00:14
Thanks once again! You have explained alot and made the subject much clearer. I have found this paper which seems to concern itself with inco alloys and dilution. A very interesting subject.

http://www.lehigh.edu/matsci/faculty/dupont/docs/Dilution_in_Fusion_Welding.pdf
Parent - - By 803056 (*****) Date 01-14-2017 03:30
interesting, but it deals with the joining of dissimilar base metals, that isn't the case here.

Al
Parent - - By 46.00 (****) Date 01-14-2017 23:26
Al, if you read the text and the equations, I think you will find them relevant regardless of the base metals.
Parent - - By js55 (*****) Date 01-16-2017 14:18
There is no dilution in similar metal autogenous welding. There is nothing to dilute. If one is going to conjecture that there is then one is using the term 'dilution' in manner inconsistent with its actual definition.
I did not read the whole article but it makes it quite clear that the application it is considering is with dissimilar metals wherein dilution can be an issue of course.
Parent - - By 46.00 (****) Date 01-16-2017 22:22
Of course there is a certain amount of dilution even in autogenous welds of the same material. A 'Fusion weld' by definition is the joining of two or more materials, unless that weld is by use of a process using pressure to force the parent material together resulting in a reduction of overall length such as a flash-butt weld, then there will always be a gap to 'fill'. If you have a gap with limited amount of parent material, then you have 'x' amount of parent, trying to fill 'y' amount of gap. This is why autogenous welds do not have significant reinforcement height on the welds. Once, and you may achieve a flat profile but repeat and you simply do not have the parent material to achieve thru thickness, so dilution is present to some extent? Am I missing something?
Parent - - By js55 (*****) Date 01-17-2017 14:01
Yes.
In an autogenous weld it is ALL parent material. If you wish to quibble over minor variations in chemistry so be it. But then it must be asked just exactly what does that mean? If you have two materials that have been manufactured within a specification range your final weld will be in that specification range, assuming of course an inert atmosphere. If you are going to use a more stringent definition of dilution then what logically stops you from taking it even farther. Say, variations in chemistry in differing regions of the SAME component (which could actually be farther apart than the differing components) so that even a bead on component weld has dilution. Even an autogenous single component spot weld has dilution. This may be accurate in an absolutist sense but is of no value whatsoever to welding.
And this doesn't even address the issue of how to determine the resulting dilution. Or even the absolute chemistry of the base metal. There is always error. Most methods are rather rough and inaccurate but are practical and useful. Even in measuring chemistry with say a PMI machine there will be error. Even with wet chems. So what value is a consideration of microscopic variations in dilution? Any dilution from that standpoint would be irrelevant to mechanical properties.
One might argue what about corrosion properties? The answer is twofold. First, if a design is so close that it relies on a determination of microscopic dilution perhaps the design engineer needs a new occupation. And second, I would be far more concerned about phenomena such as segregation that I would dilution.
Parent - - By Lawrence (*****) Date 01-17-2017 19:05
46 & Js

Good discussion of terms and behaviors.

Those of us sitting in the stands are learning something :)

Thanks
Parent - - By js55 (*****) Date 01-17-2017 20:12
So am I.
Parent - By 46.00 (****) Date 01-17-2017 22:24
As am I.
- - By 803056 (*****) Date 01-12-2017 22:04
Looking at the three primary alloying constituents of Inconel 600; Chrome 14 - 17%,  Iron 6 - 10%, Nickel 72% min.
Considering the boiling point of each of the elements Chrome 4840 deg F., Iron 5182 deg F., Nickel 5275 deg F.

Considering the temperature of the weld pool is probably between 3000 and 4000 deg F.

There would be little chance the Cr, Fe, or Ni would be vaporized. Whether the Cr could be oxidized, possibly if the protective shielding gas was insufficient. 

Best regards - Al
Parent - - By fschweighardt (***) Date 01-14-2017 02:43
Pretend it's ice
You pour some water  in the tray and you don't like how the cube solidified.
Remelt the whole cube and try again
You might get another bad cube, but none of the defects from the original ice cube will be present.
Parent - - By 46.00 (****) Date 01-16-2017 22:23
Yes - butt the ice cube will be smaller, try it and see!
Parent - - By fschweighardt (***) Date 01-16-2017 23:33
Not sure why it would be smaller, if i freeze, remelt, and freeze any given unit of water it should all still be there at the end
Parent - By 46.00 (****) Date 01-17-2017 22:29
No, it will not trust me. You may wish to research this.
Parent - - By js55 (*****) Date 01-17-2017 14:02
It doesn't matter if the ice cube is smaller. Its an analogy and is irrelevant to any consideration of dilution.
Parent - - By fschweighardt (***) Date 01-17-2017 22:42
It WAS an analogy, evidently not a stellar one, but still not sure why some water disappears,  if we are still talking about water
Parent - - By 46.00 (****) Date 01-17-2017 23:22
What happens when you leave a kettle to boil? Eventually the kettle will boil dry. This phenomena will happen even without bringing the water to absolute boiling point.
Parent - - By fschweighardt (***) Date 01-17-2017 23:26
Ah, i would only bring the water to 33 deg before refreezing, losses would be infinitesimal, i suspect
Parent - By 46.00 (****) Date 01-18-2017 00:09
Ah, you would be wrong I suspect, try that with a welding arc - the force is strong with you.
Parent - - By 46.00 (****) Date 01-18-2017 00:58
Dilution is the essence of being less than before.
Parent - By fschweighardt (***) Date 01-18-2017 01:27
Well, that's new information for me.
Parent - - By Kix (****) Date 01-24-2017 15:49
Maybe there is a bit of argon dilution in my autogenous weld. ehhh?
Parent - - By 803056 (*****) Date 01-25-2017 02:19
Its the argon that makes "TIG" welders so cranky.

Al
Parent - - By welderbrent (*****) Date 01-25-2017 13:07
They really need to stop breathing that stuff, it just doesn't do the same for you as Helium.

BB
Parent - - By Lawrence (*****) Date 01-25-2017 20:31
I worked at a place in the 80's that made the new guys stand on their heads at the end of the shift to empty the argon out of their lungs.
Parent - By 803056 (*****) Date 01-25-2017 23:29
The question is, "Did it work?"

Al
Up Topic Welding Industry / Technical Discussions / Remelting HAZ during rework of inconel 600 tubing

Powered by mwForum 2.29.2 © 1999-2013 Markus Wichitill