After the recent thread on cryogenic processing, I thought I would give a short description of what retained austenite is. Retained austenite cannot be understood without explaining how martensite forms.
First, at room temperature annealed steel we have ferrite and carbides. Ferrite is a crystal structure call Body Centered Cubic (BCC). This is a description of the repeating arrangement of atoms, here primarily iron atoms. Above the austenite transformation temperature the ferrite transforms to austenite. The austenite crystal structure in Face Centered Cubic (FCC). You can see the unit cells for BCC and FCC here (Edit: this image only shows austenite with extra atoms on three of the cube faces, but there are atoms on all six faces)
Unit cells are the smallest possible graphical representation of the crystal structure. For brevity I won't describe the process whereby ferrite transforms to austenite. The image above also has a phase diagram showing temperature vs carbon content. Alpha is ferrite, gamma is austenite, and Fe3C is the cementite, or carbides.
Heating the steel to austenite is called austenitizing, and it the first primary step when hardening steel. Also during this step some (or all) of the carbides dissolve and carbon becomes part of the crystal structure. Because carbon atoms are much smaller than iron atoms, the carbon atoms fit into the gaps in the iron atoms. These small atoms are called interstitial atoms:
Here is a diagram of where we are at the austenitizing temperature:
At this point we have austenite with carbon dissolved into it. If we cooled slowly from here the carbon would have time to diffuse and we would get ferrite and carbides again. If we cool quickly, however, the carbon doesn't have time to diffuse and it is "locked in." Then we have a situation where at low temperature the steel wants to be ferrite but we have carbon trapped into interstitial sites. This leads to the formation of martensite. Martensite is Body Centered Tetragonal (BCT) as opposed to BCC ferrite. The reason it is tetragonal is because the carbon strains the lattice in one direction more than the other which elongates the unit cell. The high strength of martensite comes because of this strained lattice. This can be seen in the following image:
The "Bain model" is FCC austenite with carbon interstitial atoms in it, which then transforms to BCT martensite on cooling.
Because martensite transforms on fast cooling it is a diffusionless transformation. So the transformation occurs through displacement of the existing austenite atoms rather than through a rearrangement of those atoms. So martensite atoms always maintain a relationship with the parent austenite:
Therefore at grain boundaries where the atomic arrangement is no longer regular the martensite can no longer grow. Martensite plates/laths are limited in size by the original grain size of the austenite, where here martensite is the black ovals:
So back to our earlier diagram here is where we are currently after cooling to some temperature below Martensite start (Ms):
Which hopefully has brought us a little closer to explaining retained austenite. These martensite plates grow very rapidly, so their formation is controlled almost entirely by temperature. The lower the temperature below Ms the more martensite we form. The plates form and stop growing when they hit another martensite plate or an austenite grain boundary. Therefore we would expect the largest plates to form first and smaller plates to form later. The more plates that form the smaller are the gaps between plates, these gaps being retained austenite.
So why cryo then? The martensite start (Ms) and martensite finish (Mf) temperatures are related, and those temperatures are controlled by various alloying elements, including carbon and chromium. Higher carbon reduces the Ms and Mf temperatures, until the Mf temperature can be lower than room temperature. You can see this in the following two images:
The first image shows hardness vs carbon content in martensitic steel, with hardness increasing with carbon content. At high carbon, though, you see two diverging lines. One line is the hardness going down and one has hardness going up. The line with hardness still increasing is with the use of cryo to convert retained austenite to martensite. The second image shows retained austenite fraction with carbon. Where you can see that retained austenite increases with increasing carbon. Since kitchen knife steels are high in carbon (just look at White #1) we would expect there to be retained austenite without cryo. Again, this is because the Mf temperature has been reduced below room temperature. Other alloys such as chromium also reduce the Mf temperature.
Here is an awesome video of martensite formation:
Is retained austenite bad? No. However, as you can see from the hardness vs carbon plot above, retained austenite does reduce hardness. That's because austenite is lower strength than martensite. However, austenite is also a much tougher and more ductile phase than martensite, so steels with retained austenite exhibit greater toughness than steels that are fully martensitic.
I hope this all makes sense and it's too long winded.