316L, Delta Ferrite & Sigma Phase 1

[generalpatton]

Doing a little investigation and claim upfront to not be an expert but looking for opinions:

MIG welding 10ga 316L SS to 2" sch40 316L SS pipe, .045 316Lmetal core wire. Sheet chemistry below. Assuming there is not excessive delta ferrite present before welding, what is the likelihood that a)additional ferrite is created and b) sigma phase could occur in an estimated time of 5 mins in the zone for sigma phase formation?

I'm sure there details needed that aren't provided, but inquire or please give general statement regarding sigma phase formation.
Thanks.


C .01
SI .28
MN 1.16
P .02
S <.0005
CR 16.62
MO 2.07
NI 10.12
CU .42
AL <.001
B .0013
CO .42
NB .06
SN .006
TI .005
V .11
W .09
FE 68.51

 

[metengr]

For the welding parameters you stated and assuming no excessive heat input, sigma phase in the base metal HAZ would be next to zero because of lack of dwell time with the temperature range to form sigma phase. Ferrite formation is controlled by chemistry of the weld and base metals.

 

[metengr]

Why do you ask? Normally, one would see brittle fractures related to sigma phase formation or confirm sigma phase based on metallographic examination.

 

[generalpatton]

Yes, that's pretty much the issue. There is an issue with transporting a large mass attached to 10ga material, presumably causing fatigue/brittle failure due to vibration and cold working. However, it has been introduced that there is sigma phase formation due to extensive delta ferrite. All of the literature I've come across says 100's to thousands of hours to form, but you indicate elevated temperature may expedite. That's where I lose any perspective. Where does elevated lie? As mentioned, the time at formation temp is likely in minutes. That's where the discrepancy lies. Thanks for the input.

 

[metengr]

The issue is time at temperature exposure. Under conventional welding practices and certainly wall thickness you have in your scenario, I really would be grasping to advise that sigma phase formation will happen. You just don't have the time exposure to 1000 to 1600 deg F for the time for sigma phase to nucleate and grow.

I could buy fatigue crack propagation, especially if the item was not adequately supported versus sigma phase. Proof would be metallographic examination to confirm if sigma phase developed.

 

[generalpatton]

Thanks for your input. Certainly no conclusion here due to limited info, but input appreciated. We are at a metallurgical study difference of opinion at the moment. Go figure. =) Thanks again.

 

[weldstan]

Agree with metengr. Sigma will not be of concern. Geometry induced (high stress concentration)fatigue or overload stress failures are most likely, both of which can readily be designed to mitigate such failures.

 

[Maui]

Sigma phase can form in austenitic stainless steels that contain appreciable amounts of ferrite, as well as in ferritic and duplex stainless grades. It normally forms along the grain boundaries, resulting in chromium depletion of the adjacent ferrite. It is formed by either (1) slow cooling from temperatures of 1900F to 2100F or (2) water quenching from 1900F to 2100F followed by heating at 1040F to 1796F. Thermal exposures at 1560F appear to produce the most pronounced effect. The embrittlement of the steel is most detrimental after cooling to temperatures below approximately 500F. As others have mentioned the precipitation of sigma phase is relatively slow, and does not typically occur in the time frames normally associated with welding. Other mechanisms that may be responsible for the observed cracking should be the focus of your investigation.

Maui

 

[generalpatton]

Thanks. Very good information.

 

[weldstan]

Please understand that sigma can form in fully austenitic steels like 310 & HK40. It is not confined to austenitic ss with delta ferrite.
John Stanley P.E.

 

[Maui]

John, true, but very unlikely in practice.

Generalpatton, here's more than you probably ever wanted to know about sigma phase:

http://www.hindawi.com/journals/isrn/2012/732471/

Maui

 

[generalpatton]

Maui, thank you. That's great info to take in. Thanks to all again.

 

[EdStainless]

In single phase materials it takes awhile to form sigma.
In materials with delta ferrite and the associated non-uniform chemistry sigma will form at the edges (where delta ferrite and grain boundaries intersect) very quickly.
I have done tests with both ferritic stainless grades and duplex.
The samples were under annealed, just enough to get them single phase, but not enough to homogenize the chemistry (they all met spec).
In well annealed samples sigma took hours to form, in the under annealed ones it took less than 10 min.
In some samples it took less than 30 sec.
Prior therm-mechanical history is large factor.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, Plymouth Tube

 

[generalpatton]

At this point the original test was concluded to be invalid, at least the sigma phase theory, and fatigue failure the ultimate root cause.