I happened to come across your article in Industrial Heating on the subject of retained austenite (“A Discussion of Retained Austenite,” March 2005). There is a statement that says, "Given the opportunity, retained austenite will transform to martensite."  

My question is why does it only transform to martensite? Or does it transform into ferrite/cementite/pearlite/bainite if it is held for a few hours during a temper cycle below A1?  

One of our customers analyzed a large 4320 block (with 0.06%V) in the as-quenched condition and reported 2-19% retained austenite, which we found baffling for such a low-C alloy. Our feeling is that during subsequent tempering at 1180°F (638°C), retained austenite will transform to martensite and then to tempered martensite during post-weld heat treatment (at a temperature just below the tempering temperature). But someone then raised the question on other transformation products, and now we are asking you the question.  

The transformation you refer to (retained austenite to martensite) is situational. Retained austenite will transform due to either temperature or deformation (stress). As you are most likely aware, the martensite reaction (transformation) is an instantaneous shear mechanism. In most instances it will transform due to a temperature drop, such that as the temperature declines, the metastable retained austenite eventually reaches the martensite restart temperature (Ms’) and will continue to transform to martensite until the temperature eventually drops to and below the martensite finish (Mf) temperature. If it transforms due to deformation, this is a martensitic transformation as well.  

Special Note: The martensite restart (or Ms’) temperature is not at the temperature where the initial martensite was formed – it is at a lower temperature – and that temperature differential is a measure of the stability of the retained austenite. The Ms’ is influenced by the parameters of the initial quench and temper.  

For retained austenite to transform to bainite it would have to go through a nucleation and growth reaction (over some period of time). In order to do that, it would have to be raised to and held at a temperature (dependent upon steel type and the carbon level at the position of interest) over some period of time. You can refer to the TTT diagram for the steel type and carbon level of interest to you.   

On the question of whether the retained austenite can go to lower bainite on tempering, I have discussed this with several colleagues and the general consensus is that when retained austenite is tempered (a heat stabilizing treatment) a transformation to a ferrite-carbide mixture may be possible, but not lower bainite per se (bainite is a ferrite-carbide aggregate). The lower-bainite reaction is reported to have a bit of a shear mechanism to it (hence its needle-like structure). My feeling is that it would be more in the vein of what you get when you temper martensite (in this case the BCT martensite wants to go to BCC ferrite plus carbide).   

I believe this is what happens when you use tempering to eliminate retained austenite and "heat stabilize" the microstructure to achieve dimensional stability. Such a temper is higher than normal temperature tempering for common steels. For example, multiple high-temperature tempers are often used on tool steels. With tool steels, they use a standard tempering temperature and repeat the temper two or three times (sometimes even four or more) as the retained austenite is transformed to untempered martensite and then the untempered martensite is transformed to tempered martensite. This is an alternative to cryogenic treatment to transform retained austenite and stabilize the microstructure. Cryogenic treatment is the normal process for carburized components, high-carbon bearing steels, common alloy steels, etc. Heat stabilization is also used on special applications where the temperature in the application will be warm/hot.  

Finally, I don't see a way to have retained austenite transform to pearlite or ferrite as those transformations occur at too high a temperature.