Sparks when GTAW Duplex Stainless

Date 12-17-2008 21:23 Edited 12-18-2008 03:25

We're doing some procedure development on 2003 duplex stainless with 2209 wire. Shielding gas is 92 Ar - 3 He - 2 N per manufacture recommendation. We're doing plate tests right now, 1/8" and 5/16" and we're seeing the root goes in clean, but when we go to fill out the joint, there are tiny little sparks the shoot from the pool. As the joint is filled, the more the sparks happen. Interpass cleaning is stainless wire brush. I and one other guy I talked to have seen similar stuff on hot wire GTA of 625 but it was very intermittent. Any thoughts on what causes this?

By js55

Date 12-17-2008 22:32

Yes.
N in the shielding gas will do that.
The general recommendation has been to limit N to root pass welding.

By defaced (**)

Date 12-18-2008 03:28

Since this is a duplex, what other tricks can be played to promote austenite since that is the reason for the N addition to the shielding gas? In test plates welded with Ar shielding, the root and the cover passes showed increased ferrite (>60%), while the fill passes showed acceptable ferrite/austenite ratio.

By Kix

Date 12-18-2008 13:31

Use an Argon/Nitrogen mix for the purge on the backside and pure argon through your torch for shielding the otherside. I don't see why you need to have Helium in the mix besides to fill the vendors pockets. We used a 95/5 Argon/Nitrogen mix for purging only. If you have the $$$ to use the trimix on the root then use it because I think you might like it better, but it can be done with the 95/5. You're using the GTAW process right?

By js55

Date 12-18-2008 13:52

Duplex solidifies as ferrite and then transforms to austenite through a diffusion process. The proper heat input for the best balance is a window. If ferrite is too high slow the cooling rate down (generally by increasing heat input-either through higher volts/amps or slower travel speed) and give the diffusion more time to take place. Duplex is not an austenitic. It is not hot crack sensitive like austenitics.

By js55

Date 12-18-2008 13:54

Also, adding N to your shielding gas may not help you. If your cooling rate is too fast the diffusion process doesn't have time and you end up with chromium nitrides. The Cr will segregate to the ferrite side and if the N doesn't have time to diffuse it will combine with the Cr to make the nitrides and your impact toughness will suffer if volume percents are high enough.

By defaced (**)

Date 03-12-2009 16:47

An update. We ran a test matrix from 0 to 2% N2 and found that at 1% N2 - 99% Ar we had % ferrite in the range of 40 to 50 with two different welders and three thicknesses. Above 1% N2 in the gas and we saw the sparks, below we didn't see sparks. I also did some quick and dirty SEM/EDS work on the spots left by the sparks and found they're slag. I don't have the sheet handy with the EDS results, but Ca, Al, O2 seem to stick out in my mind as the main components of the spots.