Sub zero/ temper question

i4Marc

i4Marc

Joined
Oct 19, 2011
Messages
5,034
My understanding is that with carbon steels the first temper cycle is to temper the martensite formed in the quench and convert any retained austenite. The second cycle tempers the newly formed martensite from the first cycle, etc. I will be corrected if I have that wrong but my question is with stainless steels like AEB-L that get a sub zero treatment. Since the sub zero bath is supposed to convert all the RA why is more than one temper cycle necessary......or is it?
 
I think with carbon steels the first temper takes alpha martensite and changes into beta martensite and this will relieve the quench stresses. And the second temper would deal with any RA issues. But why that doesn't happen all at once in one temper cycle.....I have no idea. Multiple cycles also help to release certain microscopic carbides that aid in overall wear resistance, if the steel contains such alloying. Not sure if that helps out any, or if that is exactly right. Good question!

To add to that...Verhoeven says that three things happen during tempering, in addition to the stress relief. 1. Formation of super small carbides that can only be seen by electron microscopes 2. The decomposition of RA into carbides and ferrite 3. Only at really high tempering temps the super small carbides are replaced with cementite.
 
Last edited:
You've got it a little mixed up. You could BTW go right from quench to cryo [Liquid nitrogen temps] .The things that happen are relieve quench stresses, convert RA to martensite . The more complex the alloy the more steps to be taken . If you go cryo you can get small eta-carbides on tempering .Remember the last step must be temper to change any untempered martensite.Each steel is different. Get the best properties of each steel.
 
The use of sub-zero/cryo for stainless steels is to reach the Mf, which is the bottom end of the quench when all austenite is converted to martensite. This is a theoretical point, and in real world quenching, there is almost always some amount of austenite left.
For carbon steel the Mf happens somewhere between room temp and 200F, and thus sub-zero treatments do little or nothing.
For stainless steels Mf happens around -100F. This requires the sub-zero bath. If the quench on stainless steel ends at room temp, the amount of RA will be higher. It isn't enough higher to make the blade a "bad" blade, but with the finish of the quench at -100F or lower there will be more martensite and less austenite.

The temper cycles are for the same reasons whether carbon or stainless. The first temper converts the brittle martensite to tempered martensite, as well as converts any RA into new martensite. The second temper converts the new martensite into tempered martensite. You will get a variety of opinions from knifemakers about how much RA there is, and the value of a second temper...but the overall consensus is that a sub-zero bath is beneficial on any high alloy or stainless steel, and that two tempers are better than one. In some very high alloy steels, a third temper makes a slight improvement. I see more people reference to a single temper, but metallurgically, I feel two are still better. Again, the blade won't be a crappy blade with one temper, and you might need a lab to tell the difference, but if never hurts to err on the side of better.

Sub-zero, at -100F, is what a dry-ice and alcohol bath provides.
Cryo , at -300F, does the job of reaching the Mf, and also make some changes in the carbides. There is disagreement about the benefit of these changes, and about the permanence of them....but the end of the Mf is the major point of cryo.
 
This is a topic I was coincidentally trying to wrap my head around today, rereading Verhoeven's Metallurgy for Bladesmiths. I'm a bit more confused now after reading this thread, so I want to apologize ahead of time if what I'm saying comes off as being contrarian.

On p. 27, Verhoeven says "It is seen that a room temperature quench will begin to lie above the Mf temperature at %C values exceeding around 0.3 to 0.4 %C." The tables on the next page show that even for a hypoeutectoid steel with around 0.4 % carbon the Mf temp is already around -100 C. I know that even if it does have a low Mf temp, a hypoeutectoid steel still isn't going to have a significant amount of RA, but he also says that a 1077 steel will have 6–10% RA after quenching. On p. 156, he says that if one conditions that RA through tempering, it will decompose into a mixture of ferrite and carbides. As far as I understand it (and I know that I DON'T understand it :confused:), the RA only turns into "conditioned" austenite (which then turns into martensite on cooling) when tempering a "highly alloyed tool steel." This would lead me to believe that tempering simple carbon steels will only turn the RA into a ferrite/carbide mixture. I'm sure that's much better than RA, but it still isn't the same as turning it into martensite (which I guess you could only achieve by chilling it to the subzero Mf temp?).

Anyway, I know for a simple carbon steel blade this isn't going to matter much in terms of performance, but I'm just trying to understand the metallurgy of it a bit better.
 
I am not one to get into discussing the validity of Verhoeven's charts and info. But, as I said, there is a fair difference in opinion where the limits fall and what degree they matter.

I would say your final comment is the best attitude to use. Do what you can to assure as good a quench and conversion as possible, temper fully to a suitable toughness and hardness, and test you results.
 
Thank you, Stacy. Verhoeven seems the most accessible of metallurgy texts for bladesmiths so far (which is why it's what I've started with), but there's got to be a danger in relying just on that one text. I definitely don't want to take as scientific fact something which isn't actually agreed on yet!
 
This is obviously a little more complicated than it seems up front. The only way to know for sure is to have a sample tested after following the desired heat treatment procedure. This gets expensive quick. There are quite a few things that could happen, but may not for one reason or another. In low alloy and plain carbon steels, RA can transform into martensite after conditioning during the quench, or it can change to bainite during tempering, which is the "carbide and ferrite" that Verhoeven mentions. Which happens depends on the tempering temperature and time, alloy in question, and previous treatments. All my references mention that RA is apparent after %C goes above about 0.4%, which means Mf is lower than room temperature, maybe. It's really not as simple as A + B = C.
 
My neighbors keep calling the cops on me for excessive noise. Every time the cops show up I have to explain to them that noise is just my head exploding again from trying to understand this metallurgy stuff. Is there a copy of "Metallurgy for Dummies" available? Seriously, I've read and re-read the first four chapters of Verhoeven about five times.
 
That's why we're all here. Read one chapter at a time. It's kinda Metallurgy for the Non-Metallurgist, or I think it may have been published under that name at one time. If not, he wrote another book by that name. Take one chapter at a time and make sure you grasp it before moving on. Ask lots of questions.
 
When I do subzero (dry ice+acetone or MEK) I always leave the blades until all the dry ice is gone, it lasts usually an hour, but I read somewhere I do not recall that the change in the structure is instantaneaous the moment it reaches the point of conversion, is this true or not? Speaking only about knife blades thickenesses of course.


Pablo
 
Transformation from austenite to martensite in any particular crystal in the lattice happens almost instantaneously. The entire piece of steel may take a while as it drops from +400F to -100F. Once it reaches -100F, all transformation is finished. This is why there is no real need to keep the blade at -100F for more time than needed to bring the entire blade to that temperature. 10-15 minutes in the DI bath should be fine. Since the DI is not re-usable, you might as well let the blades sit unless there are more blades to quench. No harm comes from leaving them for a reasonable time. The DI bath should be followed by two temper cycles as soon as possible.
The drop from austenite to -100F and then through the tempers should be as much a continuous process as possible. The process should graph on a Time/Temperature chart as a smooth series of curves with as few flat spots as possible.

The carbide changes at cryo temp, -300F, happen over time, and one to six hours at cryo temp are the norm.
 
What's the general consensus on the effect of time v temperature when tempering?

As I understand it when tempering most of the effect is instantaneous when the desired temperature is reached, but most heat-treat guides recommend long soaks (2 hours usually), I guess in order to make sure that thick cross sections come up to temp?

Is there any harm in doing quicker tempers when dealing with medium/high alloy steels? I know a lot of traditional smiths will 'flash temper' simple steels with seemingly no harm done.
 
As with pretty much everything else in metal working, temperature has a much greater effect than time. Take a look at some aging treatments for steels. These replace tempering in parts requiring maximum dimensional stability. They usually involve temperatures below 300 F, but require times between 25 and 100 hours long. The same hardness can be reached by tempering at 300-400 in just a couple of hours.

I don't think I'd call the effects instantaneous, but they happen pretty quickly. I've seen very accomplished heat treaters recommend 2 tempers of just 1/2 hour each for simple steels. Remember that tempering is a diffusion driven process, and is highly dependent on temperature. The thing is when tempering, the distances the atoms have to diffuse if very small, so it doesn't take as long as it could.
 
Formation of martensite is a speed of sound , fast crystal change . There are videos of this available .They show an austenite grain then as the temperature drops the typical acicular martensite takes over the grain. If you set it up properly you can hear it ! Twinning is the same .Take a 1/4" zinc rod and bend it near your ear .You'll hear the twinns form !
Tempering however is a diffusion type of change, taking time . Martensite is a stretched cube and as the carbon comes out of the crystal the long dimension shrinks .This is a time and temperature dependent process.

"flash Temper " ? What happens I think is that the structure is not fully martensitic so you don't get any bad problems . It's still a poor practice as you really don't know what structure you have . If you have a problem it may ocurr later while your customer is using the knife !!
 
Formation of martensite is a speed of sound , fast crystal change . There are videos of this available .They show an austenite grain then as the temperature drops the typical acicular martensite takes over the grain. If you set it up properly you can hear it ! Twinning is the same .Take a 1/4" zinc rod and bend it near your ear .You'll hear the twinns form !
Tempering however is a diffusion type of change, taking time . Martensite is a stretched cube and as the carbon comes out of the crystal the long dimension shrinks .This is a time and temperature dependent process.

"flash Temper " ? What happens I think is that the structure is not fully martensitic so you don't get any bad problems . It's still a poor practice as you really don't know what structure you have . If you have a problem it may ocurr later while your customer is using the knife !!
 
Thanks for the responses guys, makes sense!

I love being able to quickly fact-check my knowledge like this. Really helps solidify what I've learnt from all the various sources.
 
As far as time vs temperature in tempering, the time is much weaker than the effect of temperature. I use what I call "the rule of ten".

( these are not exact, and only for example. Each steel has its own data)
If a 1 hour temper at 400F was giving a Rc60 result
A ten hour temper at 390F would probably yield the same
If you forgot the blade in the 400F oven overnight, it would probably only drop to Rc 59.
A ten degree increase to 410F in a one hour temper would also give Rc 59.
 
mete said:
"flash Temper " ? What happens I think is that the structure is not fully martensitic so you don't get any bad problems . It's still a poor practice as you really don't know what structure you have . If you have a problem it may ocurr later while your customer is using the knife !!
Click to expand...

Your comment made me remember of this video of an old knifemaker doing the temper, its weird how our memory works :)

https://www.youtube.com/watch?v=zpeyhC-UIFg#t=520


Pablo
 
peu said:
Your comment made me remember of this video of an old knifemaker doing the temper, its weird how our memory works :)

https://www.youtube.com/watch?v=zpeyhC-UIFg#t=520


Pablo
Click to expand...

Great video Pablo! That's exactly what I was referring to!

If you watch Murray Carter's videos as well, he still flash tempers his blades. He heats them slowly at the mouth of the forge until water sizzles off them in just the right way, then quenches them to stop the temper. Interesting the methods that smiths developed to deal with their lack of precise temperature control back in the day!
 
You must log in or register to reply here.
Back
Top