During quenching process, in which the metal is rapidly cooled from austenite phase so that a phase transformation (to martensite) would occur. During this transformation, in order for martensite to form, the temperature must past through Ms (martensite start), in which the martensite phase will start to nucleate, until Mf (Martensite finish).
There are many factors that could cause austenite to be retained. The first one is the steel’s carbon concentration. As the concentration of carbon increases, both temperature of Ms and Mf will be lowered (i.e. lower than room temp). When it reaches room temperature, most of it has transformed into martensite structure, but however, some minute amount of austenite will still remain.
Another will be substistutional alloying elements such as manganese. If I recalled correctly, about 5-8% of manganese will cause Ms to fall way below room temperature, in such that a quench to room temp will produce an almost, fully austenitic structure (i.e. Manganese Steel). Almost all alloying elements (for substistutional) lower the Ms temp, accept for some, such as Al, cobalt.
What you can do is to control the alloying concentration. If you can’t, then try to control the quenching temp.
To add to Salvatio's response, you can "supercool" the steel after normal quenching in a dry ice bath (-80 deg F) to assure complete transformation of retained austenite to martensite. This is a common practice with certain grades of tool steel.
We were having an issue with retained austenite causing deformation. To eliminate this we added three identical temper steps each seperated by cryogenic treatment (-100°C or lower.) This reduced the deformation to acceptable levels.
The above responses are good if indeed you need to eliminate RA (Retained Austenite).
In some alloy/applications, RA is actually beneficial if controlled to be between 12-20%. Sorry I cannot provide any references but there is information available and I am not sure if everyone agrees on this since RA under load in service can transform to brittle untemnpered Martensite and of course if it is causing distortion then you must deal with it.
I did some reading on it 20 years ago but my memory fails me. Manfred Suess, co-author of the book "Steel Selection" and founder of Techinimet Corporation, alerted me to the 12-20% beneficial syndrome. Fred and Technimet helped us to solve many metallurgical failure problems.
If you are required to eliminate the RA by specification, customer requirements, or failure analysis, then heed the recommendations of the above respondents. On the other hand, if elimination of RA is what somebody "thinks" needs to happen, you might want to do a little research. Or someone on this forum might give some authoritative references.
Ah-ha, it just dawned on me. The reading I did must have been in Freds' book "Steel Selection." Duh-uh. The book is at work and I will try to remember to look it up on Monday.
Not mentioned is one very important point in the heat treatment of steels, it very important to know what quench rate is required where you will not end up with an undesirable mixed structure upon cooling. This is usually covered in the steel producers heat treating procedures are can be found in a reference to TTT curves. There are horror stories where the proper quench instructions were not followed. The worst offender is not understanding the word immediately.
Stabilization through the elimination of retained austenite became very important in the early years when tool steels started being used for parts other than tools. Some tool steels require a lot of effort to stabilize for use above 250°F. The stabilization was all accomplished by increasing the number of tempering cycles until the advent of sub-zero treatments. A case in point for rotating components where we used D2 it required 9 tempers to allow continuous use at 600°F and excursions to 1000°F with components tolerances in the tenths.
metman,
It has been my experience that on most components that I'm familiar with the retention of 15%-20% austenite would be quite detrimental as it would transform and create metallurgical notches that would be quite detrimental. We have investigated failures where it was determined that origin of the failure of tool steel components was at prior austenitic grain boundries and or site of untempered martensite. The assumption was that this occurred at operating temperature as the material A2 (AH5) was possibly improperly heat treated. This may or may not have been the correct conclusion as our analytical tools at the time were nothing compared to today’s.
unclesyd,
Please note that I said,
"In some alloy/applications..."
and you said,
"...that on most components..."
These two statements are not inconsistent.
You ended by saying,
"...our analytical tools at the time were nothing compared to today’s."
Technimet, now under a different parent corporate name, 20 years ago had a wet chem lab, mechanical testing, salt spray testing, SEM microscopy, and several (4 or more) talented metallurgists.
I appreciate your input and cautionary comments. Well put. My intent is to help us optimize material selection/processing and therefore intend to ferret out the appropriate reference(s). I think it is fair to say that eliminating austenite and apropriately tempering the subsequent martensite is the safe approach generally speaking assuming correct quench rate. However, generally safe is not necessarily optimum.
For example; for a number of years, it was safe practice to avoid or minimize Boron in certain steel alloys which was causing dramatic failures. Modern metallurgical analysis solved this problem allowing use of less expensive B steels to replace more costly strong alloying elements for depth of hardening in heavy sections "in SOME applications" since quantity/availability must also be considered. Catapillar put this to good use as a result of advice by Fred Suess when he worked for Caterpillar before he started Technimet. His book "Steel Selection" is all about optimizing selection/processing and based on expert experience and analysis. Another example which is in this book is how to calculate required carburized case depth rather than go for an excess depth which can be costly since time vs depth is eponential and excess depth can actually be unsafe.
I am not trying to market his book just very impressed with it's practical approach.
One application in which 15-25% retained austenite is considered beneficial is in the carburised case of some gears. It is believed that this allows the "bedding in" of the mating teeth to the gear profiles.
if you can't do much to eliminate the RA, why don't you try to make it a manganese steel (10-14% Mn, 1-1.4 C), if it's still the the mech properties required range.This steel of autenitic structure would work hardens and has an very high initial strength. With these combined, ti would provide tough, hard, adrasion-resistant metal.
In response to MOB1's comment - it depends on the tooth geometry - in regular involute profile gear teeth 5% retained austenite is now considered to be the optimum.
Geoffriong, what grade of steel are you using? How, exactly, are you heat treating it? What type of furnace are you using, and what size and geometry is the part? What is your austenitizing temperature, and how long is the part held at temperature? How are you quenching it? What is the type of quenchant, and if you know, what is the quench rate? How many tempers are you performing, and what is the tempering temperature? How long is the part held at temperature during each temper?
What evidence do you have that retained austenite is the culprit? We can only help you if we have enough information. Please provide some additional information so that we do not need to speculate.
Maui,
Actually, no speculation is needed and the group has answered my question perfectly. My question was more for general knowledge than for an actual process. Thanks again to those that posted info!
MOB1 and carburize. Thank you. I was thinking it was gear teeth but could not remember for sure.
I looked in "Steel Selection" by Kern and Suess but it was not there. I made a search on Google using key words "Retained Austenite benificial." One of the links which I have not had time to finish reading gives railroad bearings as an example with a very complete description of the subject of retained austenite(RA). This would likely be a good link to put into an RA FAQ. It has phases color coded in an Fe-FeC diagram for better understanding of non-metallurgists among other helps.
Here is a pertinent excerpt from the link: forums/0001.htm
Some of the confusion regarding the subject of retained austenite in a railroad bearing results from the misapplication of knowledge developed in other industries. The toolmaking industry does not regard retained austenite favorably. Retained austenite is recognized as a cause for many premature failures of expensive tools and fixtures. Retained austenite’s low hardness is also incompatible with an application that demands the maximum attainable hardness to resist wear. The mere presence of retained austenite in a tool steel is generally regarded as a sign of improper heat treatment.
The gear industry has a more favorable view of retained austenite. While some of the same mechanisms that effect tooling applications also effect gears, there are some major differences. Tools and dies are predominately manufactured using through-hardened steel and as a result have high hardness combined with low impact strength. Gears are typically made from case-hardened steel that has high impact strength. Where most tools fail by wear or fracture, most gear failures are the result of spalling on the teeth. Spalling (FIGURE 9) occurs when the surface of a metal component is subjected to repeated cyclic loads. A crack will form and grow until a small portion of the surface breaks loose, damaging the surface and adding debris to the system (for greater detail on the subject of spalling, see Brenco Tech Forum 94-1). The gear industry has learned that having the proper amount of retained austenite in a gear tooth can delay spalling damage by suppressing crack growth. While railroad bearings experience a different form of loading and fatigue than gears do, the principle that retained austenite inhibits crack growth is common to both industries.
