Sigma phase

Date 07-25-2007 00:42

I disagree with your statement to some degree. When ferrite containing austenitic stainless steels (304 and 308) metals are exposed to temperatures 1050F to 1700CF, sigma is most likely to form. Higher chromium and moly contents favor higher rates of sigma. For example, with the 6% Mo grade, 254 SMO, sigma will form in less than a minute in the 800C-900C range. Therefore, the heat input must be carefully monitored. Temperature excursions and cooling rates must be designed to avoid the precipitation of these undesirable phases. But, the higher alloyed pure austenitic alloys like the 321 will take longer to form sigma within the kenitic temperatures. Because sigma is an equilibrium phase, it also forms from austenite but generally requires much longer times at higher temperatures relative to nucleation from ferrite. In essence, steels in the ferrite forming-moly bearing chemistry are subject to sigma moreso than the other grades. Higher ferrite contents tend to accelerate transformation and to lower the temperature where transformation occurs. The pure austenitic steels with no moly (321) tend to avoid sigma longer than the ferrite bearing grades.

Chuck

By prasad

Date 07-25-2007 19:01

Hi Chuck,

Ok i have done some research of my own . Sigma has been described as an intermetallic compound that forms in some chromium containing high alloy steels when they are exposed for 24 hrs or longer in the temperature range of about 1000 to 1700 F . Sigma phase forms most readily from the ferrite although the phase can form in a wholly austenitic structure. Increasing total alloy content and the presence of ferritic forming elements like columbium , molybdenum and titanium increase the susceptibility to sigma formation . For these reasons sigma is not likely to be found in SS 301, 302 , 304 or 308 and is more likely to be found in SS 309, 321 ,347 , 310 ,316 , and 317.

By chuck meadows (***)

Date 07-25-2007 23:16

Yes, sigma is an intermetallic compound, but it is a real chemical compound, rather like a carbide, except that it forms from two metals, like chromium and molybdenum. I can guarantee you that steels, like the 316 you mentioned, will form sigma in a LOT less time than 24 hours. As a matter of fact, this 316 will form sigma in about 20 MINUTES at 1450 degrees F. A 317 will form sigma in about 10 minutes at 1650F. A 254 SMO, a highly alloyed pure austenitic material will form sigma in less than a minute at this same 1650 F. To say 24 hours is a guideline for sigma formation at 1000-1700 is a misnomer. The higher the temperature, the quicker the sigma formation. Sigma definitely forms quicker in the moly-bearing steels, whether it is an austenitic or pure austenitic. Depending on the chemistry of the steel and the temperature exposure time is what determines sigma formation. The kenitics are what they are. Depending on the kinetics, a 304 or a 308 can form sigma just as quickly as a 309 or any of the other grades you mentioned. At some higher temperatures, sigma will form in a matter of minutes while it takes other grades much longer times at lower temperatures. The stabilizing agents like columbium and titanium and niobium will have greater effects on preventing sensitization than sigma. These stabilizing agents are normally such a lower addition to the steel that they have minimal effect on sigma. It is the moly that has the greatest effect, with or without stabilizers.

By 803056

Date 07-26-2007 01:48

A great explanation.

Best regards - Al

By js55

Date 07-26-2007 13:53

A few other points if I may. One of the greatest problems when concerning yourself with sigma is not so much the formation of it as an absolute and immediately upon procedure qualification, but the fact that the nucleation kinetics are more difficult to overcome (at a given temperature) than the growth kinetics. Once the nucleation 'mountain' so to speak (or bump in the ground with some alloys and at some temps) is overcome-once the lattice is established- growth can proceed rather easily depending of course upon the temperature.
What does this mean? Most often (though there are certainly exceptions that merit concern especially as Chuck made clear with higher alloys) sigma (and let us not forget its bretheren chi-which is often overlooked yet maintaining very similar responses mechanically and chemically) in its initial formation will effect immediate properties very little. There is work out there verifying that as much as 1 or 2 volume percent may not effect tensiles or bends. At lesser volumes it will generally be almost undetectible mechanically and quite often even through corrosive testing. In even lesser volume percents it can even go undetected visually through photomics. Especially if the exact proper etching agent is not used. But, once the nucleation has taken place the growth is facilitated more easily, and depending upon how close the service temp is to critical temps, can reduce service life considerably. Therefore, even if immediate post qualification testing reveals no problems, the long term effects can be quickly manifest.
In my opinion, the thinking behind testing for sigma in some alloys is actually quite similar to corrosive testing as a whole. Short term tests with an eye towards long term effects.
Good discussion.

By prasad

Date 07-26-2007 14:56

Good discussion . Tks for the responses

By js55

Date 07-30-2007 15:23

I did a quick reveiw of an article I had handy, and I need to add something and make a correction of myself. The diagram you are looking for is called TTP diagram (Time Temperature Precipitation). They manifest as C-curves.
The other thing I wish to add is that carbide precipitation can effect the nucleation and growth of sigma as well, since carbon helps to stabilize austenite as opposed to sigma. The idea being a carbon depleted austenite microsturcture will find an equilibrium in sigma much easier, if I may put it crudely.
So my thought is, even if sigme is not formed during the welding regime, if excessive carbides are, they may lead to accelerated sigma formation in service.
How much carbide precipitation is too much?
I don't know. Not sure that question even has an answer given the huge diversity of service regimes. Just don't let the SS get too hot, even if its a fully autenitic like 310.