Jump to content
Jerrod Miller

52100 Heat Treat

Recommended Posts

I have some 52100 that I am going to be using for a few knives and as I always do before trying something new, I checked the forum here for how to best do it. A Google site search for "52100 quench" gave me plenty of hits, but a lot of disagreement about what to really do. I have decided to take everything with a grain of salt and do the best I can on my own with that knowledge and the TTT diagram. Whatever the results, I will let you know how I heat treated it and what I got for hardness.

 

Source material: 52100, 0.5" diameter rod (spectrometer verified on separate sample from lot)

Forged knife to shape, normalized 2 times (by eye), hand filed profile.

Ran forge between 1550 and 1600 (not the greatest control in this range at the moment, burner geometry doesn't like to hold the flame unless I'm going all-out hot for welding).

Normalized 2 times.

Austenetized then air cooled to magnetic then water quenched (sub-critical quench for grain refinement) 2 times.

Normalized one more time.

Heated and quenched in heated oil (guestimated at 120F). Still soft, well not hard at least.

Heated oile about 20 more degrees, repeated quench. Same results.

Normalized once more.

Heated and water quenched (why not?). Still soft. Checked hardness today at HRC44.

 

Theory: Need to austenitize hotter. Thermocouple may be misreading.

 

I will normalize a couple times to be safe as well as re-train my eye for decalescence and try the hot oil again tonight (if I can). I plan on tempering in a kitchen oven at 300F. Low enough to be safe for poor oven control, but warm enough to relieve some of the worst stresses. Hardness check will come the day after HT.

 

Having problems adding the TTT diagram to this as a jpeg, so here it is a a pdf.

52100_TTT.pdf

Share this post


Link to post
Share on other sites

52100 can be finicky. You really need a controlled oven to get consistently good results with it. If I had an oven to heat treat in, I would normalize one time each at 1600°, 1500° and 1300° to refine the grain and carbides. Cool to black heat between each. If I were to do a normalization without a controlled furnace I would heat it three times until I say the decolesance pass through the entire blade with cooling to a black heat between.

 

To harden I would austinize to about 1450-1475° and soak for 10 minutes. Without a regulated oven you will need to get the steel through decalesance, note the color that it's at, and try to keep it at that color while preventing recalesance while you hold it like that for 10 minutes. 52100 has so much carbon in it that when the steel is quenched it can physically block the austinite form converting to martensite and it will remain retained austinite.

 

I would also at no time water quench 52100, even after cooling to a black heat. You my have read about water quenching on a data sheet but that was no written for things the dimension of knife blades. If you have let the blade cool down to a black heat any phase changes or grain refinement has already taken place and all you are doing is risking a broken blade though, realistically, you'll probably get by with it. I just don't believe in quenching more often than is necessary.

 

You didn't mention any soak prior to quench hardening so I assume that you didn't to any extent. That could be the reason for the soft blade even though you said that you had your forge adjusted to about 1600° You didn't get enough carbon into solution in the austinite before quenching. The other possibility is that you got too much carbon into solution and your blade has been softened by retained austinite.

 

I also saw that you did a multiple quench. You're spinning your wheels when you do that. The second heating prior to quenching wipes out anything from the first quench and with high carbon steels up to about 1% like 52100 you will decrease the toughness of the blade and you expose it to another chance that it might break. Steels in this range form plate martensite upon quenching and where those plates intersect there is a chance of microscopic cracks forming. Microscopic as in you would need a scanning electron microscope to see them. You can't get away from this entirely but you also don't have to increase the amount of these cracks.

 

Doug

Share this post


Link to post
Share on other sites

I set the forge to run at 1700 (starts getting fairly stable at that temp) and did 2 knives. The first knife was one that was last normalized a couple times after forging, the second was the knife from the post above.

 

Should have mentioned this in the first post: Using Canola oil, heated via a hot plate and double boiler. Oil temp judged via index finger, stirred to be even with scrap metal on hand (RR spike partially worked down). All normalizing and quenching is done in a fairly dark shop (a couple lights on in the back so I don't trip over anything).

 

The first knife I normalized 3 or 4 times while watching the recalescence. After feeling I knew where that was I went with a quench cycle: Heat to past where I thought I saw decalescence throughout the blade, pull out just long enough to see recalescence, heat back up, pull out of the forge and hold just above the quenchant until it looks near the recalescence point on the edge, plunge tip down into the hot oil. Cool completely, from oil temp to room temp via water bucket). Put tang in vice and file test. File skated - success!

 

Repeat process for blade from original post. Tried cooling oil by cool water bath of oil holding can. File test seemed good.

 

Tempered both knives in kitchen oven for 1.5 hours at 300 F. Rockwell test today showed that neither tang (near the blade, but back enough for the future gaurd to hide) got hard at 40-45 HRC. The first knife had a reading of 60 on the blade though. The original blade was spotty on hardness.

 

Next try (probably later this week): Bigger quenchant tank with more oil. Get a thermometer to check the quenchant temp. I will redo all the heating listed here for both of these knives to ensure complete uniformity of the hardening. After uniform hardness achieved, temper will be determined via a more controlled oven at work to achieve an even 58-59 HRC. Then a blade with be sacrificed to a severe test (ABS style) with grain size evaluated at the end.

Share this post


Link to post
Share on other sites

I really don't think that your problem is with the quenchant. If you are seeing recalesance while cooling the blade down you are letting carbon go out of solution. Also, if you are not giving a long enough soak you will not be getting enough carbon into solution in the first place. Try dropping Keven Cashen a line and as him. He works with 52100 regularly.

 

Doug

Share this post


Link to post
Share on other sites

Geoff Keys has also worked with 52100 with a blade and I sent him the information about heat treating it yet I can not seem to find it in my files as of yet .

Share this post


Link to post
Share on other sites

The recalescence is only seen during the normalizing cycles. Prior to the quench the blade is soaking at temp (about 1700) for 2-4 minutes. It was already been normalized several times by this point which should have refined the carbides as well as the grains, thereby mitigating the effect of carbides to an extent. Smaller carbides should dissolve quicker too.

 

Spectrometer readings for this specific batch of 52100 listed C=0.95, Cr=1.4. No other carbide formers were present in significant quantities. If we assume ALL POSSIBLE chrome carbides formed: Cr23C6 --> (6C*1.4%Cr)/23Cr = 0.365%C tied up in carbides, leaving about 0.58%C in solution. That is enough to form appreciable martensite in and of itself, and there is no way ALL of the carbon was tied up as carbide.

 

I think this is really starting to be too much of a documented trial and not a "this is how I do it" kind of threaded. It seems to me that it should probably be moved to the "Metallurgy and other enigmas" sub-forum. If a moderator could make that happen I think that would probably be best. Sorry for not putting it there in the first place.

Share this post


Link to post
Share on other sites

Yeah, I don't think your quenchant is the problem....more like temperatures. I know what gives me great results....but you have to be at the right temps.

 

1650 for 10 minutes, cool to black, quench if you like

1550, cool to black, quench if you like

1450, cool to room temp

 

for autstenitizing, 1525 for 10 minutes, immediately into 130F canola oil or similar. You should get 66 HRC out of quench.

 

It sounds as if your austenitizing temperature is not high enough, not long enough. Carbon is coming out of solution, as was mentioned by Doug.

Edited by stuart davenport

Share this post


Link to post
Share on other sites

You got it, Jerrod. Good info, though!

Share this post


Link to post
Share on other sites

Thanks Alan!

I was thinking about this last night and and figured I should mention that my quick calculation should have a bit of a disclaimer. While it shows that I can get martensite without having any soak time at austenitizing temperature, this does not mean that it is going to be the best result. If carbides are not reduced and/or refined then you will not get the optimal properties out of the metal. It is a handy thing to know showing that I did not get martensite therefore indicating that the quench was not adequate. There is a chance that the chrome being tied up in carbides reduces the hardenability too much, but I find this to be highly unlikely. I'll see what data I can come up with for temperature dependent dissolution rates of chromium carbide in steels. That should help determine how hot and how long I need to hold it. I'm hoping to build a bigger quench tank today, so between that and finding the rite temp/time combo I should be in good shape to get it done over the weekend.

Share this post


Link to post
Share on other sites

So, everything I have read (not just from today, but what I can remember before my review today too) chromium carbides aren't being dissolved much with any of the practices I have seen listed for knife making. The charts below are pretty neat as they show that heating slower encourages carbide dissolution and the relation between austenitizing temp and resulting carbides. It should be noted that the densities listed in the carbide retention chart will not be the same for 52100 as they were from 420 (martensitic stainless with about 14% Cr). The proportions would hold true though. These came from a paper I found missing the title page, but had the University of Pretoria (in 3 languages, just he letterhead - not the paper). In summary: Heat it "slowly" (about 1 deg F/sec), especially when getting above 1830 F (1000 C). Take it all the way to 2100 F (1150 C), and quench from there. This will give you the least amount of carbides. 1800+ was the lower limit I saw in all my research for when chrome carbides start to dissolve.

 

I am thinking this is not overly practical. For starters, some carbides are not a bad thing. 52100 and other non-400 series stainless steels used in knife making don't have that much chrome in them to begin with, so what carbides are there are not that many. If anyone has some good scientific data on dissolving chrome carbides I would love to see it.

 

Mildly off topic (broader than just with 52100): I am far too skeptical about just taking "this is how I do it and it works" when it comes to things that aren't understood. We know that you have to get to austenetic, the cool it fast enough to get martensite rather than pearlite or bainite. Without knowledge of TTT diagrams I would highly question why some things work in oil and others require water. Fortunately TTT diagrams exist and I can feel good about them. This is especially hard for me when it comes to smiths suggesting soak times for alloys. I need to know why a certain temp and time is given, or it just seems like superstition to me. Right up there with quenching with the blade pointing north. That isn't to say that the smiths are wrong about their suggested times and temps. I just feel the need for a proven reason. I'm not sure if that personality quirk developed as I became an engineer, or that lead me to engineering. Either way it is just the way I am.

 

(Post preview shows the tables coming in with their order reversed, but they are properly labeled.)

Cr dissolution as a function of heating rate.JPG

Cr carbide retention based on austenitization temp.JPG

Share this post


Link to post
Share on other sites

Jerrod,

 

This thread on another forum had some good info, pay special attention to what Kevin Cashen has to say.

 

~Bruce~

Share this post


Link to post
Share on other sites

I also found this (in one of the books on my bookshelf no less, should have started there!); in the section talking about 51XX and 52100: "Since the chromium content of low-alloy steels is less than 2%, the chromium atoms replace iron atoms in Fe3C to produce the complex carbide, (Fe,Cr)3C." Yikes! That means my assumption of Cr23C6 was way off. On average we are probably looking at around 1.5:1 The formula should then be (1C*1.4%Cr)/1.5Cr = .93%C used up in carbides. That is pretty much all of it! Those carbides MUST be taken care of in order to harden the steel.

 

I can find no concrete evidence that a soak is necessary for anything longer than to ensure the temperature has been reached throughout the piece. I have found nothing to contradict it except the conspicuous lack of mention of soak times (with the exception of 1 technical source that specifically mentioned 15 minutes to ensure through temperature conformity). Nothing I have found says that it takes time for carbides to dissolve. I therefore do not feel comfortable making a recommendation to anyone to either soak or not soak.

 

Bruce - Thanks for the link. Kevin's response is right where I was going with my research. To sum it all up for those that don't want to follow the link or try to put together the above info with Kevin's statements in from the link in their heads:

 

52100 HT:

  1. Get the carbides dissolved/refined so that you have carbon in solution.
    1. Get it hot. 2100 F would be great, but at least 2000 F.
    2. Cool it fast, but not a full on water/oil quench. Fan cool, placed between a couple chunks of steel, something like that. Try for bainite.
    3. Repeat as desired (this includes only doing it once)
  2. Proceed with quench and temper cycle.
    1. Austenitize at 1500-1550 F. Your refined carbides will not dissolve at this temperature.
    2. Quench in oil. This should give you up to 67 HRC.
    3. Temper at 350-425 F depending on aim hardness.

Again, I haven't done this (yet) so these are not my actual results. I can tell you that my research is leading me to want to do 2-3 carbide treatment cycles. Mainly because I do not know what was done to the steel before I got it and it doesn't take a lot of time to run through it. Until I really get into a groove with this material and get a good oven for tempering I will continue to throw the blade in the kitchen oven at 300 right after quench to help protect it until I can hardness check it. After I get a hardness reading I will temper at a higher temp and recheck hardness to see if I got what was needed.

Share this post


Link to post
Share on other sites

If dissolving carbides was all there was to it then of course you would want to heat the steel up to 1800-2000° and get the job done. However, that's not all that there is to it. If you heat steel that hot then you are going to have a lot of large grain and a loss of toughness and increased brittleness. Many makers recommend that you quench from under 1500° to prevent grain growth and to offset the lower temperature effect on carbon solution in the austinite then you will have to increase the time at temperature. According to Verhoeven the upper limit is 1560°. One of the problems with reading IT diagrams and data sheets is that they are almost never made taking into account things the dimensions or purpose of knives. They are for larger objects probably over with a minimum thickness of 1". Another problem with diagrams and books is that they make certain assumptions like that you are aware what the proper soak time is fore the steel

 

Another problem, that I pointed out earlier, with strongly hyperauticoid steels such as 52100 is putting too much carbon into solution in the austinite and then trapping it in the austinite matrix and not allowing the austinite to slip into a body centered tetrahedron with carbon trapped in it (martnesite). It will remain a face centered square (retained austinite) That will cause problems up front and down the road.

 

Read the post from Kevin Cashen that Bruce supplied you a link to. Slower than air cooling can create problems with carbided. Cooling to form bainite is possible with isothermic quenching. Just remember that what is converted to other products of quenching will be unavailable to convert to martensite. I've done this. I quenched 52100 that had been soaked at above decalesance for 10 minutes in oil at 400° for 4 hours. I got a blade that would hold an edge chopping through a 2X4 twice and bent very close to 90° before it broke. I haven't been able to afford to ship it out to a lab for testing yet.

 

If you really want to get the straight skinny on heat treating 52100, or much anything about heat treating, contact Kevin Cashen. He is well versed in metallurgy and has a lab to do his own testing, including microscopic. You can count on what he has to say to not be superstition. You can probably find a link to him hear or go to his site Hype Free Blades Forum. You can probably find a good discussion on the heat treating of 52100 there. I'm sure it's been covered there.

 

Doug

Share this post


Link to post
Share on other sites

"I can find no concrete evidence that a soak is necessary for anything longer than to ensure the temperature has been reached throughout the piece."

 

The following is from a thread on BladeForums titled Working the three steel types by Kevin Cashen. It is a highly informative thread and I suggest reading it through several times.

 

 

Remember with hypereutectoids it is better to go a little lower and stay there longer. Let me explain why- your soak temperature will determine how much carbon you will put into solution, your time will determine how evenly it will be distributed. If you have an entire grain (incredibly small by our standards, yet a very, very large space by a carbon atoms standards) to fill evenly, you may get the carbon you want in solution really quick with your temperature but now you have to move it though the grains to equalize the concentrations. You can move it quicker by bumping up the heat but this will automatically also put more carbon into solution than you may want.

 

~Bruce~

Share this post


Link to post
Share on other sites

If dissolving carbides was all there was to it then of course you would want to heat the steel up to 1800-2000° and get the job done. However, that's not all that there is to it. If you heat steel that hot then you are going to have a lot of large grain and a loss of toughness and increased brittleness. Many makers recommend that you quench from under 1500° to prevent grain growth and to offset the lower temperature effect on carbon solution in the austinite then you will have to increase the time at temperature. According to Verhoeven the upper limit is 1560°. One of the problems with reading IT diagrams and data sheets is that they are almost never made taking into account things the dimensions or purpose of knives. They are for larger objects probably over with a minimum thickness of 1". Another problem with diagrams and books is that they make certain assumptions like that you are aware what the proper soak time is fore the steel

 

Another problem, that I pointed out earlier, with strongly hyperauticoid steels such as 52100 is putting too much carbon into solution in the austinite and then trapping it in the austinite matrix and not allowing the austinite to slip into a body centered tetrahedron with carbon trapped in it (martnesite). It will remain a face centered square (retained austinite) That will cause problems up front and down the road.

 

Read the post from Kevin Cashen that Bruce supplied you a link to. Slower than air cooling can create problems with carbided. Cooling to form bainite is possible with isothermic quenching. Just remember that what is converted to other products of quenching will be unavailable to convert to martensite. I've done this. I quenched 52100 that had been soaked at above decalesance for 10 minutes in oil at 400° for 4 hours. I got a blade that would hold an edge chopping through a 2X4 twice and bent very close to 90° before it broke. I haven't been able to afford to ship it out to a lab for testing yet.

 

If you really want to get the straight skinny on heat treating 52100, or much anything about heat treating, contact Kevin Cashen. He is well versed in metallurgy and has a lab to do his own testing, including microscopic. You can count on what he has to say to not be superstition. You can probably find a link to him hear or go to his site Hype Free Blades Forum. You can probably find a good discussion on the heat treating of 52100 there. I'm sure it's been covered there.

 

Doug

 

Doug - I'll try to address things in order that you posted them.

1) Dissolving carbides is not everything, true. But if you notice I mentioned to cool from this state fairly fast and aim for bainite. Cooling it fast from such a high temperature will do quite a lot for you in terms of refining grains. Upon further thought though, this should probably be followed up with at least 1 lower temp normalize cycles. Anything from 1475-1650 would probably be fine, going with quicker cooling rates with the higher temps. Still not quenching though. For those following along we had a nice little discussion about grain growth here: http://www.bladesmithsforum.com/index.php?showtopic=27264

2) I still can't come up with a good reason to ever increase time at temp (outside of tempering) other than just long enough to ensure the metal is entirely up to temp. Kevin Cashen mentions to allow for carbon to diffuse evenly, but I would think that all should happen during forging. It will surely happen as much as necessary if you go up to 2000 F. I also like to think that commercially produced steel should be assumed to not need any homogenization treatments.

3) IT diagrams are made with the intent to be valid regardless of section size. CCT diagrams are better about this as they show what happens for a given cooling rate (which part size affects greatly).

4) The cooling to bainite and subsequent heating to lower austenetic temps will encourage some, but not complete, carbide formation. If one was to quench and achieve martensite from the really high temps then the carbon would be in solution and there could be a problem with austenite if you don't get it cool enough (sub-zero). Reheating from this condition would be ideal for refined grains and even carbide distribution. It may be risky from a thermal stress perspective though.

5) Clearly I read that link. I thanked Bruce for link and mentioned that the content reflected right in line with where I was heading. I also just read the entirety of the second link he posted.

6) I am a trained metallurgist too. I am currently sitting about 6 feet away from my microscope. I fully intend to follow up my testing with metallographic analyses. I am very fortunate to have access to this stuff at work.

 

 

"I can find no concrete evidence that a soak is necessary for anything longer than to ensure the temperature has been reached throughout the piece."

 

The following is from a thread on BladeForums titled Working the three steel types by Kevin Cashen. It is a highly informative thread and I suggest reading it through several times.

 

 

~Bruce~

 

Bruce,

I just added that link to the Link Collection thread in the Beginners Place. I am very much of the same mindset with Kevin on understanding the core concepts of what is going on, not just following a recipe. Though see note 2 above. Also Kevin doesn't really mention it, but A2 in the Fe-C phase diagram is the curie point for ferrite. Everything past the eutectoid (hypereutectioid) is non-magnetic above A1, which becomes A1,2,3. Not everyone seems to notice that but it is the case. This means the magnet test works for eutectoid and hypereutectoid steels as a measure of "the austenite you have is capable of forming martensite if you cool it quick enough and cold enough". It just doesn't tell you how much carbide is still going to be around and how stable that austenite is, so you may have a heck of a time reaching Ms.

 

 

The tricky part to all of this really is knowing how much carbide exists, and actually controlling it. That in turn gives you your carbon content which is what you really want and the rest is relatively straight forward from there. Assuming one fully grasps everything Kevin lays out in the most recent link Bruce posted. We just have to keep in mind that the iron-carbon phase diagram does not strictly apply once Cr is added, so being able to find the A1 and Acm difficult.

Share this post


Link to post
Share on other sites

Jerrod, do it your own way. We've tried to help you here but all you do is to get defensive. If you would go to Hype Free Blade Forum, do a search for 52100, and scroll down to where it deals with heat treating 52100 and look around for Kevin Cashen's posts you will come up with the heat treatment that he uses and why. IT diagrams and data sheets only take you so far and there is a difference in dealing with thin sections as below 1/4" there is little difference between cooling the surface and the middle of the steel.

 

Jerrod, I'm not a metallurgist but If enough people tell me that I'm using the wrong process, I think that I would listen. Contact Kevin Cashen, as I said he is well versed in metallurgy and a one of the other moderators at that site is a metalurgist so both can discuss this at your level.

 

Doug

Share this post


Link to post
Share on other sites

Doug, my apologies if I am coming across as defensive. My intent is to explain my reasoning. People say "why re-invent the wheel?" My response is that I like the pneumatic tires on my car better than the original stone/wood wheels. Well, having not tried the originals I am assuming that the pneumatic tires are better, but I think that is a safe assumption. At this point the process I have laid out is easier for me to do and I cannot find anything to definitively say not just that it will not work, but why. I would MUCH rather destroy 100 knife blanks and learn what is going on than follow a heat treat recipe without knowing what variations (intentional or not) do to the process. If all that comes from my trials here is proof of why something shouldn't be done (such as maybe I am wrong in the amount of grain refinement that I will get from quick cooling from approximately 2000 F) then I do not think the efforts will be a waste of time and energy.

  • Like 1

Share this post


Link to post
Share on other sites

Ok, Jerrod, I can't give you the answer you need the soak, but I have gone back to my books. Even with that I doubt that I can discuss this on your level so again, let me recommend that you drop Kevin a line. He is very approachable and he's probably the one who I rely on most for information though he does have opinions that some may not agree with. So let's just say no blood, no foul. I've go to admire a man who would destroy knives to prove a point or learn something new but don't you think 1000 is a little over the top? :P

 

Doug

Share this post


Link to post
Share on other sites

Well you gotta break a blank every now and then. Not just to check grain size or appease shop gremlins, but to show the rest of your stock that you mean business. :angry: (That is my "tough guy" face intimidating the rest of my stock pile). Let's just hope it doesn't come to breaking that many. Maybe 10 though. If I can't see any real improvement in methods and understanding after that I may have to go back to 1080, work to 1095, then come back to 52100 later. I am just thankful that I get to use some nice tools at work to be able to do all this research and development. Without it there is no way I would be able to get so involved in expanding my knowledge about the alloys and heat treat. When I get done with the first round of tests and post the data here I will probably drop Kevin a line with a link to this thread to see if he wants to weigh in on it all. Unfortunately things came up and it looks like no heat treating this weekend. :(

Share this post


Link to post
Share on other sites

Sam - Thanks for those links. I haven't had the time to go through all of the material yet, but it is VERY nice to know that the folks at Northrop got 100% carbide dissolution at only 1800. I think to play it safe I will still overshoot by 50 degrees since I don't trust my thermocouple as much as I would like. I can't wait to get a chance to fire the forge up. I really should work on my burner before I do this though. I am running a venturi and I think I need to adjust my flare and/or gas orifice as I am getting some back-pressure shooting the flame back into my pipe. This seems to only happen with lower pressures and I want to have better temp control in the 1500-1700 range so I should use a smaller orifice to keep higher pressures and less volume. I am still hoping to get to heat treating sometime this week.

Share this post


Link to post
Share on other sites

Wow.

Quite a bit of trouble just to nail down a heat treat on the steel ain't it?

 

:0)

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

×
×
  • Create New...