Grinding / finishing of welds 1

[geesamand]

My company purchases welded components in low carbon or 304 or 316 stainless steel. The welding is usually continuous fillet welds ranging from .19 to .5 nominal size.

Occasionally a customer wants the welds "ground smooth". This creates troublesome expectations because I've not seen a technical definition of "ground smooth".
Expectations such as:
- must the toe of the weld be blended into the base metal
- how much grinding undercut is permitted into the base metal
- is 100% cleanup required over the surface of the weld
- if passivated, should all chemical spots and marks get polished out
- is heat discoloration permitted
- what surface finish (Ra roughness / grinding grit size)
- what tests, if any define acceptability

Generally vendors believe it should be a quick pass with the power grinder and customers believe it should look like a medical implant. We all need clearer expectations.

Are there standards or other methods to communicate these expectations to/from customers and pass them consistently to our fab vendors? Suggestions please.

 

[HakiNoir]

This was a headache for me at some point of time!Particularly if you have a customer who does not have a good knowledge of Metallurgy.
Unless you are working for pharmaceutical or some nuclear jobs, heat tints, chemical spots are more than ok based on unwritten industry common practices! You can find some in SSPC standards (SP1 - SP11). If you dont have those, try SSPC-SP COM which has a summary of ISO, NACE and SSPC on surface preparation. I ended up writing my own spec with lots of example pictures plus hours of discussion with customer to convince them and clear the path!

In terms of welding, based on the welding code there might be some restriction. Blending the weld toe is a good practice for the parts faced fatigue loading and sometimes to enhance that, you would be better to TIG weld the transition. Some of these are in API RPs and ASTM welding handbook vol.?. But unless specified in the PO, its hard to justify that.

My suggestion: you need to have a customized spec in place, as there is not a comprehensive one in the industry as far as I know. Basically we couldn't answer this simple question with a quantified response: "How much clean, is clean??"

Hope this helps. Good luck!

 

[geesamand]

Hello HakiNoir, very helpful response.

I had some ideas in another section but nothing appeared comprehensive or designed for this perspective.

http://www.eng-tips.com/viewthread.cfm?qid=415254

Our company in fact does some joints that are welded and finished and used in a fatigue loaded environment. Those are fully machined and don't generate concerns from customers. Then we have many other components that are welded without fatigue loading. The other components are where customers and vendors and we vary in expectations.

We have company standards for machining, finishing, inspection, and acceptance of the fatigue case. We don't have a standard for all of the other situations, and I think creating such a standard is much broader than the fatigue standard.

Regards,

David

 

[MikeHalloran]

Heat tint is just really bad.
It should be removed completely by pickling or by machining or by grinding.



Mike Halloran
Pembroke Pines, FL, USA

 

[SwinnyGG]

If this issue is causing you that much headache, maybe you should come up with a standard of your own, complete with photographic examples, for distribution to customers. Allow them to choose, and upcharge them accordingly.

 

[HakiNoir]

@geesamand Your welcome. I agree with you and @ jgKRI on creating a standard as I did it myself as well. Thats also recommended in SSPC SP COM. Make sure you add "Accepted" and "Rejected" pictures both in that. However, if you happen to face an inspector which couldn't distinguish between mill discoloration/marks and rust you would come back to the first step!

@Mike I am not agree with that. It's all depends on your application! The logical acceptance criteria in oil/gas, nuclear and pharmaceutical industries are different. You dont need to add cost where its not necessary.

 

[MikeHalloran]

Let me rephrase that.
Leave the heat tint, if you don't mind mysterious cracking in service.


Mike Halloran
Pembroke Pines, FL, USA

 

[geesamand]

If this issue is causing you that much headache, maybe you should come up with a standard of your own, complete with photographic examples, for distribution to customers. Allow them to choose, and upcharge them accordingly.

It should be well understood that 3rd party / existing standards offer much higher value than a standard I write and attempt to educate my customers about. This issue of weld "quality" most often arises after the component has arrived and someone at the customer site decides they don't like it. These are customers who hardly read their approval drawings and proposals, so getting them to read a multi-page company standard is not likely to happen. A perfect solution would involve me putting down "all unmachined welds meet std xxx grade b" on the proposals and general arrangements, so that they can't claim we're choosing our quality ex post facto and they cannot argue with the content of the standard.