52100 Tempering hardness

BluntCut MetalWorks

Knifemaker / Craftsman / Service Provider
Joined
Apr 28, 2012
Messages
3,462
I broke yet-another 52100 blade into 3 pieces today - a wrist flick, the blade hit a dry 2" pine branch laterally. It was just too brittle, broken like a ceramic blade. Super fine grain, very sharp, etc... however obviously something is off.

Specs: 4" blade 0.093" thick 52100 from Aldo. Evenheat digital oven: normalized 30minutes, 3 martensite grain refinement thermal cycle, 1480F hardened temp, double cryo/LN2, 3x tempered (400F -> 465F -> 515F) 2 hrs each. A 61rc file won't abrade this blade tempered at 515F.

* actual oven temp is +- 20 degrees from panel display. quench: interrupt brine then warm canola (I like ping & crash more than my McMaster-Carr 11secs. Park50 ordered)

Obviously I need to test this broken knife Hrc. I am thinking about sending it to PeterHS to test. I've another blade tempered at 600F and it's still shows sign of micro-chip and 61rc file won't dig into it either. which I speculate/guestimate rc around 62. I looked at 3 tempering charts for 52100, all telling me that my guestimate rc is way too high.

I would appreciate your help/insights :thumbup:
 
............quench: interrupt brine then warm canola (I like ping & crash more than my McMaster-Carr 11secs. Park50 ordered)
......


There is your break. You may not see it at the quench, but I would bet it was there. Use the proper quenchant. 52100 is not a fast quench steel. Quenching faster than recommended will not make the steel any harder.


As to your guess of Rc52 by a file test, they are pretty useless for testing accurately. They dull fast, too, and the test gets very inaccurate after a few uses.

At 515°F, you would have Rc56 under laboratory conditions. I would recommend 52100 be austenitized at 1500°F and quenched in oil. A sub-zero treatment immediately after the quench is beneficial, then tempered twice at 400°F. This should put you about Rc59/60.

If possible, borrow a pyrometer and place the TC in your oven to compare readings. I suspect something is off.

Where I see potential problems is in your current HT regime.
Spheroidize the blade and refine the grain by cycling down - Heat to 1525F, hold for ten minutes, and oil quench, heat to 1350F, hold, and quench. Heat to 1250F and hold for one hour. Air cool to 900F and then quench.
The blade is now ready for final HT. Heat to 1500F and hold for 10 minutes. Quench in medium to fast oil. Cool to room temperature and place in dry ice/LN for ten minutes. Allow to warm up to room temperature and temper twice at 400F.




Doing an excess of multiple quenches to refine the grain and multiple cryo treatments is not going to gain anything, but may damage the blade ( especially if using the wrong quenchant).
Here is why blades should be cycled:
First austenitization - about 25°F above the target and a suitable soak time. This will assure full solution of carbides.
Second austenitization - about 25°F above critical with a suitable soak time. This will assure the finest grain. ( repeating this will gain little or nothing)
Third heat is sub-critical, and as close to 1250-1275°F as you can hold it. 30 minutes to one hour here will spheroidize the steel, putting all the carbides into nice bundles. The steel is nowin as fine grain and stress free condition as you can get.
Final austenitization at the target ( 1500F for 52100), soak for sufficient time to allow the carbides to re-distribute ( 10 minutes is normally all that is needed for a carbon steel).
Quench in the PROPER quenchant for the steel. Using a faster quenchant will not gain any hardness, but may damage the steel by introducing great stress and possibly micro-cracks.
Steels with alloy content and potential retained austenite gain a bit from completing the quench by lowering to around -100°F. This is called sub-zero, not cryo. Cryo does nothing for any but the very high alloy stainless steels. The transformation at sub-zero is fast, and there is no need for a long soak.
What is needed is to get the blade warmed back up to room temperature with no stress ( allow to warm up hanging in still air is best), and then IMMEDIATELY temper. Two tempers of an hour each , with cooling to room temperature quickly between tempers, is the final step in HT.

I have made some changes in how I HT in recent years, due to spending some time with metallurgists who are vastly more informed, and who have done some blade specific testing. These are the changes you may have noticed over the last couple of years in my posts:
1) I now heat the blade higher on the first cycle. 1525F is my normal first austenitization. You need to reverse any super-spheroidation that may have been placed in the structure by the factory or your earlier processing ( forging and normalizing). This seems especially important for Aldo's 52100, which comes very spheroidized.
2)My austenitization temperature is now at the higher limit of the charts. This seems to get better quality blades. Many used to use the lowest austenitization as they could, thinking that it made harder blades.. There seems no advantage to this, and some disadvantage. If the chart says 1470-1500F, go with the 1500F. Do not go lower than the metallurgical charts recommend.
3)I cycle the steel down in three steps on all carbon steel blades - 200F above critical, just above critical, and subcritical. There seems to be a divided view on whether it is needed to quench or air cool to 900F on the first two, but it seems that either method gets good results.
4)I water quench from the 900F point in normalizing and in cycling. Slower cooling in air, cooling in sand/ashes/vermiculite, or slow oven cooling can make an undesirable type of pearlite.
5)I immediately do the sub-zero treatment as soon as the blade reaches room temp after final quench. Any delay can stabilize the austenite.
6)Cryo isn't the great thing that was originally thought. Sub-zero is very much a good thing. The eta carbides and other changes for true cryo ( -300F) are apparently all reversed in temper. Using LN is fine, but won't appreciably do more than -100F will.
7)I still temper twice, but have been told that in a blade thickness, one hour each temper is sufficient.
8)Tempering should not be delayed any more than is possible. Getting the blade from quench to temper in the shortest practical time is best. Waiting to cool down the HT oven may be too long. It is best to place the freshly quenched blade in a second oven ( kitchen oven or a toaster oven), to hold it at 300F until the HT oven cools for the temper program to start.
 
Stacy, question, I was looking at some data sheets, and one gave two heat treat options, 1475, then fast oil, or 1500 and medium oil. I tried both, and couldn't tell much difference in Rc or performance. Thoughts?
 
Stacy, like you, I have heard that Aldo's 52100 comes in a very spheroidized state and need that normalization cycles before hardening. Kevin says that the stuff works VERY well at an austenizing temp of 1475 because you end up with virtually no RA.
 
Stacy/bladsmth, thanks so much for your generosity with time; expertise and patience:thumbup:

............quench: interrupt brine then warm canola (I like ping & crash more than my McMaster-Carr 11secs. Park50 ordered)
......


There is your break. You may not see it at the quench, but I would bet it was there. Use the proper quenchant. 52100 is not a fast quench steel. Quenching faster than recommended will not make the steel any harder.
I grind post-ht, where in this case, I tried to remove most surface cracks. Pic of this completed grind blade, where breakage may have originated from the surface crack but with adequate toughness the blade shouldn't snapped like a ceramic knife.
IMG_0982.jpg

As to your guess of Rc52 by a file test, they are pretty useless for testing accurately. They dull fast, too, and the test gets very inaccurate after a few uses.

At 515°F, you would have Rc56 under laboratory conditions. I would recommend 52100 be austenitized at 1500°F and quenched in oil. A sub-zero treatment immediately after the quench is beneficial, then tempered twice at 400°F. This should put you about Rc59/60.
My guess was/is rc62 because I just dulled the edge of my small square jeweler file tried to too hard to cut this blade. I used bm940 s30v rc58 & spyderco stretch zdp198 rc66 as baseline for comparison. I also compare these experimental knives edge retention with my kitchen knives in skd-11 rc64, sg-2 rc63, v2 rc63, blue #1 rc62.

If possible, borrow a pyrometer and place the TC in your oven to compare readings. I suspect something is off.[/B]
I used an oven thermometer to calibrate 350F, 400F, 450F what under the tool wrap stainless steel blanket. I will buy a pyro-thermometer (lol - I book marked a long time but never pull the trigger).
IMG_0985.JPG

Where I see potential problems is in your current HT regime.
Spheroidize the blade and refine the grain by cycling down - Heat to 1525F, hold for ten minutes, and oil quench, heat to 1350F, hold, and quench. Heat to 1250F and hold for one hour. Air cool to 900F and then quench.
The blade is now ready for final HT. Heat to 1500F and hold for 10 minutes. Quench in medium to fast oil. Cool to room temperature and place in dry ice/LN for ten minutes. Allow to warm up to room temperature and temper twice at 400F.

Doing an excess of multiple quenches to refine the grain and multiple cryo treatments is not going to gain anything, but may damage the blade ( especially if using the wrong quenchant).
Here is why blades should be cycled:
First austenitization - about 25°F above the target and a suitable soak time. This will assure full solution of carbides.
Second austenitization - about 25°F above critical with a suitable soak time. This will assure the finest grain. ( repeating this will gain little or nothing)
Third heat is sub-critical, and as close to 1250-1275°F as you can hold it. 30 minutes to one hour here will spheroidize the steel, putting all the carbides into nice bundles. The steel is nowin as fine grain and stress free condition as you can get.
Final austenitization at the target ( 1500F for 52100), soak for sufficient time to allow the carbides to re-distribute ( 10 minutes is normally all that is needed for a carbon steel).
Quench in the PROPER quenchant for the steel. Using a faster quenchant will not gain any hardness, but may damage the steel by introducing great stress and possibly micro-cracks.
Steels with alloy content and potential retained austenite gain a bit from completing the quench by lowering to around -100°F. This is called sub-zero, not cryo. Cryo does nothing for any but the very high alloy stainless steels. The transformation at sub-zero is fast, and there is no need for a long soak.
What is needed is to get the blade warmed back up to room temperature with no stress ( allow to warm up hanging in still air is best), and then IMMEDIATELY temper. Two tempers of an hour each , with cooling to room temperature quickly between tempers, is the final step in HT.

I have made some changes in how I HT in recent years, due to spending some time with metallurgists who are vastly more informed, and who have done some blade specific testing. These are the changes you may have noticed over the last couple of years in my posts:
1) I now heat the blade higher on the first cycle. 1525F is my normal first austenitization. You need to reverse any super-spheroidation that may have been placed in the structure by the factory or your earlier processing ( forging and normalizing). This seems especially important for Aldo's 52100, which comes very spheroidized.
2)My austenitization temperature is now at the higher limit of the charts. This seems to get better quality blades. Many used to use the lowest austenitization as they could, thinking that it made harder blades.. There seems no advantage to this, and some disadvantage. If the chart says 1470-1500F, go with the 1500F. Do not go lower than the metallurgical charts recommend.
3)I cycle the steel down in three steps on all carbon steel blades - 200F above critical, just above critical, and subcritical. There seems to be a divided view on whether it is needed to quench or air cool to 900F on the first two, but it seems that either method gets good results.
4)I water quench from the 900F point in normalizing and in cycling. Slower cooling in air, cooling in sand/ashes/vermiculite, or slow oven cooling can make an undesirable type of pearlite.
5)I immediately do the sub-zero treatment as soon as the blade reaches room temp after final quench. Any delay can stabilize the austenite.
6)Cryo isn't the great thing that was originally thought. Sub-zero is very much a good thing. The eta carbides and other changes for true cryo ( -300F) are apparently all reversed in temper. Using LN is fine, but won't appreciably do more than -100F will.
7)I still temper twice, but have been told that in a blade thickness, one hour each temper is sufficient.
8)Tempering should not be delayed any more than is possible. Getting the blade from quench to temper in the shortest practical time is best. Waiting to cool down the HT oven may be too long. It is best to place the freshly quenched blade in a second oven ( kitchen oven or a toaster oven), to hold it at 300F until the HT oven cools for the temper program to start.
:thumbup: I drawed on your & other's wisdom & knowledge for my earlier batches of experimental 52100 knives. Their performance in my informal wear-resistance test was exceptional - 2nd best just behind k390 steel.
http://www.bladeforums.com/forums/s...-Wear-resistance-test-for-12-different-steels

although I don't know much about metallurgy, curiosity satisfaction worth pushing the limits. If I can get a sense for where the trade off between hardness/toughness/carbide_vs_martensite, then success.

To add cryo to my ht tinkering, I bought a 20L dewar. This knife is #4th in this experimental batch of 16 knives. First cryo is 10 minutes either after quench and or after quick snap temper at 315F. 2nd cryo is deep soak 12-24hrs to precipitate some eta carbide (maybe some or none). Oh I almost forget to mention - previously I can rough grind 2 knives with a 2x42 60 grit ceramic belt. But for these knives, it took 2 belts to grind 1 knife. I frugal with belts (they ain't free in the hobby world), so I use it until completely dull off the drive wheel.

I am confused by my experimental data. Someone please fish me out of this rabbit hole or better yet, send down a flash light:)
 
Quenching a THIN blade of deep hardening steel in a "violent" quench medium like water is where you will see that kind of behavior in my limited experience. i had that happen with Parks #50 and a thin kitchen knife blade made from Cru Forge V.
I broke yet-another 52100 blade into 3 pieces today - a wrist flick, the blade hit a dry 2" pine branch laterally. It was just too brittle, broken like a ceramic blade. Super fine grain, very sharp, etc... however obviously something is off.

Specs: 4" blade 0.093" thick 52100 from Aldo. Evenheat digital oven: normalized 30minutes, 3 martensite grain refinement thermal cycle, 1480F hardened temp, double cryo/LN2, 3x tempered (400F -> 465F -> 515F) 2 hrs each. A 61rc file won't abrade this blade tempered at 515F.

* actual oven temp is +- 20 degrees from panel display. quench: interrupt brine then warm canola (I like ping & crash more than my McMaster-Carr 11secs. Park50 ordered)

Obviously I need to test this broken knife Hrc. I am thinking about sending it to PeterHS to test. I've another blade tempered at 600F and it's still shows sign of micro-chip and 61rc file won't dig into it either. which I speculate/guestimate rc around 62. I looked at 3 tempering charts for 52100, all telling me that my guestimate rc is way too high.

I would appreciate your help/insights :thumbup:
 
Hi Joe, thanks & I agree that 'violent' transformation would result in dimensional tears especially in the vulnerable edge thin cross section. I grind post-ht, so I've full thickness to mitigate edge tear/crack at quench.

Quenching a THIN blade of deep hardening steel in a "violent" quench medium like water is where you will see that kind of behavior in my limited experience. i had that happen with Parks #50 and a thin kitchen knife blade made from Cru Forge V.

edit: add pic for blade tempered 600F and whittled dry oak.
IMG_0986.jpg
 
Last edited:
We use Kevin C's recipe of 1475° at a 15 minute soak, quenched in 11 second oil..I broke a blade in parks 50 and wont do that again..52100 dosnt need parks 50 to get fully hard..You can get close to 67rc with the 1475° aust. temp..Temper @ 450° for what we use..
 
To add to what Phillip said, when Kevin was messing around with 52100, he ended up using medium speed oil for the quench instead of the low temp salts that he normally uses on O1 and L6. You have to figure that if he abandoned the more stable quench and auto tempering characteristics of salt in this case, there had to be som significant benefit to using AAA instead. My recollection is that he and others have said that while there is a lot of "mysticism" out there about the heat treating of 52100, it, like D2, is one steel where you can arguably get a better result at least for knives by going off the page a bit from the standard industry heat treatment recommendation, which was designed for bearings, not cutting tools.
 
I had a discussion with another well known smith, and he thought that 52100 from Aldo's stock might work well with a 30 min soak without the grain refinement cycles. (He also said that if Kevin Cashen recommended something else than listen to Kevin!) I will test this week or next ( I ordered a new kiln today!!!!:thumbup::D) once I can control the soak and temps properly. A 30min soak in a forge is a disaster waiting to happen. I have quenched 52100 with fast oil from 1475 and got Rc66 without any problems with brittleness or chipping after tempering.
 
:thumbup:. I also have done many 52100 blades with 11sec oil and 125F canola oil. Most turned out well, except a few severely warped blades when experimental quench at 1650F (1st thermal). I look forward to try Park50 - ping or not to ping, that's the caution.

We use Kevin C's recipe of 1475° at a 15 minute soak, quenched in 11 second oil..I broke a blade in parks 50 and wont do that again..52100 dosnt need parks 50 to get fully hard..You can get close to 67rc with the 1475° aust. temp..Temper @ 450° for what we use..

heheh, If Kevin is messing around, which could mean I paint blindfolded with a shovel and hope to produce a Monet... painful but I accept:cool:

I water quenched to test whether it's possible to achieve compressive hardening (on top of martensitic) by dimensional collapse from surface inward. otoh, ok to LOL about now.

To add to what Phillip said, when Kevin was messing around with 52100, he ended up using medium speed oil for the quench instead of the low temp salts that he normally uses on O1 and L6. You have to figure that if he abandoned the more stable quench and auto tempering characteristics of salt in this case, there had to be som significant benefit to using AAA instead. My recollection is that he and others have said that while there is a lot of "mysticism" out there about the heat treating of 52100, it, like D2, is one steel where you can arguably get a better result at least for knives by going off the page a bit from the standard industry heat treatment recommendation, which was designed for bearings, not cutting tools.

I've tried/tested normalize aldo 52100 using a forge. Heat to ~2000F hold 1 minutes. Heat to ~1700F hold 3 minutes. After this, I thermal cycle & harden. I got good result too.
I had a discussion with another well known smith, and he thought that 52100 from Aldo's stock might work well with a 30 min soak without the grain refinement cycles. (He also said that if Kevin Cashen recommended something else than listen to Kevin!) I will test this week or next ( I ordered a new kiln today!!!!:thumbup::D) once I can control the soak and temps properly. A 30min soak in a forge is a disaster waiting to happen. I have quenched 52100 with fast oil from 1475 and got Rc66 without any problems with brittleness or chipping after tempering.

I just re-tempered a couple 52100 blades from previous batch (no water quench & cryo) to 515F. A file has no problem abrading these blades, i.e. RC in mid 50's. My oven reading at 400F still agrees with an oven thermometer.

How can this batch 600F tempered blade whittled oak for 15 minutes w/o instant dull/edge-roll? I certainly understand that all you have go by is my blah blah w/o evidence. Please continue to provide insight & guidance & humor-the-process.

Much appreciated!
 
I water quenched to test whether it's possible to achieve compressive hardening (on top of martensitic) by dimensional collapse from surface inward. otoh, ok to LOL about now.

why?
what are the success criteria? How would you test compared to a more standard heat-treat recipe?
 
A soak CAN be done i a forge if you have precise temperature control like with a Fogg drum forge. You can also soak in a regular forge if you use something like a black iron pipe muffle, but you have to be able to hold the temps tight. I saw John White sue his small Foog forge and the variance was under 5 degrees for the entire 10 minutes he was soaking W2 at 1425 or so after it stabilized.
I had a discussion with another well known smith, and he thought that 52100 from Aldo's stock might work well with a 30 min soak without the grain refinement cycles. (He also said that if Kevin Cashen recommended something else than listen to Kevin!) I will test this week or next ( I ordered a new kiln today!!!!:thumbup::D) once I can control the soak and temps properly. A 30min soak in a forge is a disaster waiting to happen. I have quenched 52100 with fast oil from 1475 and got Rc66 without any problems with brittleness or chipping after tempering.
 
Success = high performance 52100 blades - compare against baseline ht blades by such as Peter, Paul Bos, etc.

Informally I can test experimental vs baselines from me & external/commercial sources.
wear resistance: cut sand & denim rope & abrasive mixed rubber slab.
edge retention: cut rope, cardboard, whittle wood and other repetitive tasks.
strength & toughness: chop, bend/flex, notch.
nice to have: SEM - etch to show grain and allow to calc carbide density.
why?
what are the success criteria? How would you test compared to a more standard heat-treat recipe?

Mete - it is about time for you to end my nonsense/mis-guided metallurgical thoughts/ideas:foot:
 
Last edited:
I was able to get a 15 min consistent temp in my forge, but the chamber was not big enough for a muffle pipe. I debated between a new forge or a kiln. and the kiln won.
 
yeah, that will solve your problem.
I was able to get a 15 min consistent temp in my forge, but the chamber was not big enough for a muffle pipe. I debated between a new forge or a kiln. and the kiln won.
 
I do the cycling steps on all carbon steels.

I use very little 52100. Nothing wrong with it, just that I use other steels more often. 15N20, W2, and Hitachi paper steels get used on most of my kitchen carbon blades. 1084 and 1070 for most forged camp and field carbon blades. Most all other knives are stainless, with CPM-S35VN being my go-to stainless steel. I have a batch of AEBL blades working to compare.
 
A soak CAN be done i a forge if you have precise temperature control like with a Fogg drum forge. You can also soak in a regular forge if you use something like a black iron pipe muffle, but you have to be able to hold the temps tight. I saw John White sue his small Foog forge and the variance was under 5 degrees for the entire 10 minutes he was soaking W2 at 1425 or so after it stabilized.

I used a muffle and a thermocouple and could hold temps with 3° +/- for 20 minutes no problem. I tuned the forge with the gas flow and homemade baffles. It took time for sure but it worked very well. Ive got pics somewhere if anyone wants to see them?
 
Back
Top