Not logged inAmerican Welding Society Forum
Forum AWS Website Help Search Login
Up Topic Welding Industry / Technical Discussions / NiCu overlay of carbon, dilution problem?
- - By cddolan74 (**) Date 06-07-2014 16:18 Edited 06-16-2014 12:28
Will not be able to post all relevant info, visiting site soon to find out cause. just need suggestions on possibilities of cause and suggested fix.  I don't have specs or WPS.
The issue is solidification cracking (transverse)on the first pass of a two pass overlay  .045" (NiCu-7). My first guess is dilution and  probably get them to weld on the colder side that's allowed by there WPS.  any other possibilities you may know of would help.
thanks
Parent - - By 803056 (*****) Date 06-08-2014 09:42
Do a macroetch on the cross section to determine the amount of dilution with the base metal. Dilution, i.e., depth of fusion should be kept to a minimum. The copper in the NiCu doesn't play well with iron, assuming the base metal is iron based alloy.

Al
Parent - - By js55 (*****) Date 06-09-2014 12:23
Al is correct. Your problem isn't carbon. Its iron.
NiCu's will tolerate no more than ~15% Fe with GMAW.
I'd use another nickel base alloy for the first layer such as NiCrFe. The iron is low enough that dilution is not a problem. But you still must control dilution due to the Cr.
Parent - - By cddolan74 (**) Date 06-09-2014 14:10
Thanks Js55,
not sure if using a different filler is a option, will find out.  I am definetly taking Al's advice on the macroetch and maybe show at differen travel speeds what dilition this customer is getting. 
"NiCu's will tolerate no more than ~15% Fe with GMAW."
most likely getting alot higher than 15%, where did you get this info, would be valuable going in.
thanks
Parent - By js55 (*****) Date 06-09-2014 14:16
I have the info in a nickel based alloy conference presentation on dissimilar welding by Donald Tillack, then consultant to NIDI.
Parent - - By electrode (***) Date 06-09-2014 16:22 Edited 06-09-2014 16:34
I was asking myself 'where' to take this macrosection to reveal the amount of dilution.
Or should be taken more than one (1) macrosection in order to achieve an appropriate average determination?

Anyway. Additionally to js' advice, maybe this helps as well, see attachment.
You may wish to have a closer look at 'page 17' of the document*.

And, maybe also quite delightful and interesting to you:
https://www.youtube.com/watch?list=PLI-jKVKu0QrmAJPlxX95Gig-KAgWWKG6R&v=t2JZYn819DQ.
You may wish to have a closer look at 'part 5' of this series.

Regards.

EDIT:
*) Due to exceeding the size limitations of this forum; please note that I had to delete the 'Brazing' and 'Soldering' chapter from the attached pdf + I had to renew the video link.
Attachment: JoiningOct03.pdf (0B)
Parent - - By js55 (*****) Date 06-09-2014 16:50
Though Al's suggestion about macros is a good one, I prefer to PMI the weld however. Of course, this only works if you have actuals on the base metal and the weld metal. But it can get you pretty close.
If you know what you started with and you know what you have, you know the dilution.
Parent - - By electrode (***) Date 06-09-2014 18:00
"Though Al's suggestion about macros is a good one..."
Yes, as long as there's common understanding that more than one (1) indication is required to properly evaluate.
Sure it was meant this way.
I seem to be nitpicking - again.

However, to 'PMI' the weld sounds interesting.
Is this being accomplished along the z-axis of the weld fusion zone deploying a macrosection?

Regards.
Parent - - By js55 (*****) Date 06-09-2014 18:05
If we assume cracking is initiated in the first layer we have to get down to it. Now, you can either macro to determine that, or reproduce your parameters for the first layer and test your verification pad.
One other thing to think about, if you wish to qualify this to ASME Section IX you should be doing this anyway.
Parent - - By electrode (***) Date 06-09-2014 18:12
I seem to understand.
Thank you.
Parent - - By js55 (*****) Date 06-09-2014 18:42
Corrosion resistant overlay is (as is a lot of welding with dissimilars) basically about dilution control.
Doing something different from one layer to the next is not unusual.
When I worked for an overlay company in the past we would run 309 on the CS substrate and then the 308 as a final if that's what they were looking for.
This is not necessary however. Some believe that 308 on CS will automatically lead to martensite. Not so. It depends upon dilution. We did plenty of straight 308 when we did single layer overlays for example in the ID's of small bore pipe.
Parent - - By electrode (***) Date 06-10-2014 08:54
"Doing something different from one layer to the next is not unusual."
Yes. And as you did already mention. A buffer layer to mitigate fissuring issues in this application appears recommendable.

However, in regards to 'dilution' it would be interesting to know whether reducing the fusion depth alone would overcome the problem's root cause. I may be wrong, but in my understanding the latter might be addressed maybe to the metallurgy itself.

Do convective weld pool effects play a major part here rather? If so, then fusion depth reduction may lead to some measurable positive effects. But I suspect that somehow you will always end up in a detrimental metallurgical iron-copper reaction, independent of how "much" (whatever that means) the fusion depth is reduced.

In the broadest sense the older ones among us might be reminded of these SAW overlay relaxation cracking issues in the 1970's mainly related to ASTM A 508 / Cl.2 Grade steels. Whatever you did, finally you ended up with a Heat Affected Zone (HAZ), a (more or less) coarsened grain zone and - in regard to the overlapping between 2 weld beads - a detrimental region of HAZ discontinuity. All this - and certainly more (which I honestly forgot) - in combination made it almost impracticable to predict cracking occurrence.
Parent - - By js55 (*****) Date 06-10-2014 12:16
electrode,
Reducing fusion depth will of course reduce dilution. But given the alloy you are combining I doubt you will see consistent satisfactory results.
Think about the math. Fe is around what, 98%, of carbon steel.
This means your dilution is going to have be reduced to less than 16% (and lower with processes other than GMAW-I don't know why). This is tough. Tough in the sense that with standard fusion processes such as the big 5, GM, GT, SM, FC, and SA you are beginning to get to a consistent bonding problem. If you have to weld in this region you need to think in terms of either auto-welding with pulsing or even better, strip cladding, where in my experience you can get dilution under 15% and 10% respectively and still get good fusion.
If you do a lot of overlays I would recommend taking a look at strip cladding. I know a very good company that deals with this. But if you don't the big 5 are fine (some better than others) but you have to make other accommodations, like buffer layers.

Convective weld pool effects will homogenize the dilution. Stuff moves around fast when liquid. But the available alloy is the available alloy regardless of convection.
Parent - - By electrode (***) Date 06-10-2014 13:16
js,
Yes, at the end of the day it is as ever.
You may have to find some sensitive balance between proper metallurgical bonding and metallurgically driven drawbacks.
It would be fine to learn on how the OP could finally tackle his particular problem.
Welding remains fascinating!
Thanks again.
Parent - - By js55 (*****) Date 06-10-2014 13:47
Yup. Science mixed with art.
With a little bit of luck tossed in.
Oh, and the phases of the moon as well.
Parent - By electrode (***) Date 06-10-2014 15:06
"...and the phases of the moon as well."
That impressively precisely says it and leads us - as I seem to notice at this stage - to Mr Lawrence' "chocolate covered doughnuts" again. :)
- - By cddolan74 (**) Date 06-13-2014 12:16
Thanks guys for your comments. Travel speed at this point corrects the cracking issue. I have in the back of my mind your suggestions of a different filler for the first pass.
Js55,
"I have the info in a nickel based alloy conference presentation on dissimilar welding by Donald Tillack, then consultant to NIDI"
I have orderd some papers by Tillack on this subject, I couldnt immediatly find anything available on his discussion you mentioned.
If I can,  I will later post the results I find on dilution %.
thanks again for your responses
Parent - By Metarinka (****) Date 06-13-2014 23:13
I spent a good deal of time on a very similar issue. They went one step further and wanted to weld NiCu directly to SS 304... with no filler.    It is true that around 12-15% Fe you'll start to see cracking as it is not soluble with copper, if you look on the met cross sections you'll tend to see islands of copper floating in the nickel-ferrous alloy, with cracks coming form the interface. Anything that can lower dilution, such as less heat input, faster travel speed etc tends to help.  Alternatively you can butter the weld with a Nickel based alloy which is soluble with copper and iron.    Finally the obvious culprit is to remove Iron if possible or switch filler metal chemistry.  This is not always an option.

Good luck, it tends to be a great alloy to weld if it stays far away from Iron.
Up Topic Welding Industry / Technical Discussions / NiCu overlay of carbon, dilution problem?

Powered by mwForum 2.29.2 © 1999-2013 Markus Wichitill