I want to like 52100... but..

Ed Caffrey said:
I gotta share an experience I had about a year ago with 52100. I was experimenting with the number of thermal cycles, and wanted to see what multiples would do to it. I started with 3, then 4, and all the way up to 7, After hardening each, I sent samples off for spectrographing.......those thermal cycled 5-7 times exhibited less and less hardness, until the one done 7X barely hardened at all. The spectrograph results showed that the grain size at 7X thermal cycles was so small that the steel simply would not fully harden. So, from the results of that experiment, it proved that you can make the grain size too small in 52100. As Don said, it has some peculiarities that need to be understood and in some cases considered when working it.

I still stand by my feeling for this steel. Every cutting competition that I've won, has been won with 52100 or its Japanese "twin brother" SUJ-2.
Click to expand...

Ed, How many thermal cycles did you determine were best for 52100?
- Thanks
 
Grain refinement reduces hardenability in carbon steels. Small grain is a good thing though, so speed up the quench? Problem with this, is the high Mn and Cr in 52100 can't usually handle a fast quench, have heard of it being water quenched with success. Like I said, it's a weird bird. Three normalizing cycles seems to be about right for the carbon steels we use. I have quenched W2 5 times and still reached full harness, bet the grain was mighty fine :)
 
Mike Krall said:
Scott,

This is the second time I said what I said and this is the second time you said what you said. I'm serious about my belief, but I'm joking.

Here is a thing I've wondered. There must be a fairly standard HT process in 52100 bearing manufacture... end to end... from starting the balling process to end product. Would there be differences, or not, in HT for bearings and blades? Would you say what you feel the differences, or not, are.

What would you're response be to that question "mete"?

Mike
Click to expand...

How do you remember that kind of stuff? I must be getting old, because I don't remember random comments from different threads like that.

Steel choice is a matter of personal preference. If someone asked why I use 52100, I would tell them because I get a lot of it that either has to be taken to the recycle center or used in my knives. Basically, I use it because it's free, readily available and is a very good knife steel. However, anyone that's been to my shop knows that I have many other steels too. I have 1080, 1084, 15N20, D2, 5160 and W2. My personal favorite steel to use is 5160.

What is really interesting is that most of the bearings being recycled into knives, aren't made of 52100. They are actually A485-1 and A485-2.

52100 won't fully harden through at cross sections greater than about 1/2", thus the use of the other two steels by bearing manufacturers.

As far as your question about the HT for bearings vs. blades. I can tell you that bearing manufacturers do not go through the Ed Fowler method.

Bearing manufacturers also have HT processes that aim for 8% to 20% retained austinite. Retained austinite is a good thing in bearings. It's not a good thing in knives. So it's like comparing apples to oranges.

Three different steels and HT processes that aim to achieve different results between bearings and knives, results in comparisons that aren't relevant.

By the way, I make my larger knives out of 5160 most of the time. I make my smaller knives out of any of the others, expect for 5160. I only use 52100, A485-1 and A485-2 because they are readily available to me.

If I were to intentionally make a personal carry knife for myself, I'd probably use 1080 or 1084 for it, instead of 52100. I don't use O-1 or 1095 at all.
 
Scott,

On remembering and not... everybody is operating with a different data base. Would surprized me more if you did remember... =]

It's nice to learn the differences in bearing material again from you... yeah, you said it before but I'd forgotten.

How do you HT 52100 and the related-others for knife blades, if you don't mind saying? I'm asking because of your background. I know 'mete' was a bearing guy and somewhere I've got his point of reference on the subject saved (not in the traveling data bank, though.)

For single steel blades I use O1 for cutting blades and L6 for tough ones... mix them for the exercise and "the pretty". Most of my damascus is W2/1080/15N20 though I've made a couple of blades of straight W2.

Mike

PS -- What size stock are 1/2" bearings made from?
 
To get back to some of the posts with the answers that I do know....

Forged 52100 is a very different animal from ball bearings. The reason being that ALL ball bearings that are produced from 52100 are machined out of the round stock that 52100 is normally produced in.....generally the desired finished size of bearing is produced from the next 1/4" larger stock. (IE: 1/2" bearings are normally machined from 3/4" stock, as so on)
There are going to be differences (machining/stock removal versus forged) simply because, as with most steels we use for blades, blades were not what the folks who designed the steel(s) had in mind. We have taken the vast majority of steels, intended for other purposes, and adapted them to knife blades. Up until just a few years ago, there were no steels that I can think of that were produced specifically with knife blades in mind (the only one that even comes close was 154CM and that came about from other steels).

OK, that aside, I have settled on multiples of 3 being ideal for 52100.....3X thermal cycles, 3X quench, and 3X temper, etc. I know that some folks are gona scoff at that, and thats OK.....my advice to those folks is to do your own experimenting, do your own spectrographing, and testing.....and I mean actually do it!

Don't take pieces of the puzzle from a whole bunch of different sources and hope to come up with a completed picture...it doesn't work....I tried that, all it got me was more confused and frustrated. When I finally decided to separate the "wheat from the chaff", and started at ground zero, on my own, things became very clear. It did take some effort on my part, and yes it did cost me in dollars and time, but the reward was being secure in my mind that what I currently do with 52100 does work, and works well. That doesn't mean that it won't change int he future...this is an ever evolving art, and I'm just one face in the crowd. I certainly don't know everything there is to know about 52100, nor would I ever be arrogant enough to think I ever will. BUT! I know enough to be able to separate what is fact for me, versus a lot of hog wash that some folks will shovel at ya.
Scientific proof? Thats kind of an ambiguous thing to me. Often times it seems that those who tout that term are more interested in propping themselves up, than they are at getting to the heart of a matter.
Not sure if you can call it "Scientific" or not, but I have all the proof I need in the file folder of specrograph results, and the 6 notebooks full of notes that I've accumulated over the years on this steel.

I guess what I'm getting at is this: Take people's words ONLY as a starting point (yes, even mine). Get yourself a working theory, and then prove or disprove that theory. Those theories you prove, then become fact for you, those you disprove become fiction. If you can't solve the whole puzzle, work on a piece at a time, and then start putting the pieces together. With some time and effort, you'll solve the puzzle, and in the end you will realize that its not really the finish line thats driving you... its the journey that got you there.
 
Am I missing something? Everyone is trying to get the best out of this steel, is there not existing data available for it? I would imagine, after a couple normalization cycles at decreasing temperatures each time, because of the high carbon and high alloy content, it would need a moderatley high Austenitizing temperature say 1550F, and a good bit of time at that temperature for everything to come into play nicely, then quench. All this multiple quenching and stuff seems to be roundabout ways of trying to NOT do it properly, please someone correct me if I am wrong.
 
Sam Salvati said:
Am I missing something? Everyone is trying to get the best out of this steel, is there not existing data available for it? I would imagine, after a couple normalization cycles at decreasing temperatures each time, because of the high carbon and high alloy content, it would need a moderatley high Austenitizing temperature say 1550F, and a good bit of time at that temperature for everything to come into play nicely, then quench. All this multiple quenching and stuff seems to be roundabout ways of trying to NOT do it properly, please someone correct me if I am wrong.
Click to expand...


That's the page I was on, but apparently we're not even looking in the right book.

I will find out this weekend what is up.

I always do 3 normalize any way, so I'll do a couple identical blades up to the normalizing then quench one 3 times after the edge becomes non-magnetic and the other I will soak 10 mins at 25F lower than the last normalize cycle. then snap the tips off and continue to temper, sharpen and test both blades.
 
Define "trying to NOT do it properly".

If you don't have a controlled HT oven, or are fixated on heating just the edge with a torch, all of these gyrations may well be the best path.

If you do have a controlled HT oven I agree with Sam and Matt.
 
Sorry if it seems like I muddied the waters. That wasn't the intent. For me the "established data" was the starting point. All the rest came seeking to go beyond that. In my opinion 52100 is "OK" using those standards, but there is much more to be achieved with the steel...thats where I was coming from. Doing it "properly" is a matter of perspective. If you just want to create a blade to "established standards" thats fine. If you want to go beyond those standards it requires some time and effort on your part to discover how to get there. That was the whole point of that long post.
 
Dan Pierson said:
Define "trying to NOT do it properly".

If you don't have a controlled HT oven, or are fixated on heating just the edge with a torch, all of these gyrations may well be the best path.

If you do have a controlled HT oven I agree with Sam and Matt.
Click to expand...

My advice if you don't have control, get some 1084 :D You can go to all this trouble trying to eek out all the performance aspects of a steel that would kick ass if you just heat treated it right, when you can bang some 1084 in the forge, go non magnetic and quench. You just don't recommend a square peg for a round hole.

Ed, understand I meant no disrespect.
 
Thanks for your post Ed.

A basic scientific approach would be to measure the Rockwell C hardness, and then graph how your HRC changes to smaller grain sizes. The anecdotes youve shared while interesting, dont provide any data to back up your claims or give a deep understanding.

I try to learn what I can from the tool and die industry because its a billion dollar global industry that has been the subject of many rigorous tests and research. It is known that it becomes more challenging to achieve good martensite formation in smaller grain sized tool steels but there is papers that has showed research about changing quench methods and quenchants to still have a nice hard product with next to no retained austenite.

The benefits to greater toughness in particular at the same HRC values should be something that knifemakers strive to achieve from smaller grain sizes.
 
Ed Caffrey said:
Sorry if it seems like I muddied the waters. That wasn't the intent. For me the "established data" was the starting point. All the rest came seeking to go beyond that. In my opinion 52100 is "OK" using those standards, but there is much more to be achieved with the steel...thats where I was coming from. Doing it "properly" is a matter of perspective. If you just want to create a blade to "established standards" thats fine. If you want to go beyond those standards it requires some time and effort on your part to discover how to get there. That was the whole point of that long post.
Click to expand...

Ed,

I don't know from muddied waters. I'm more interested in what you are doing. Would you say what "3 thermal cycles, 3 quenches, and 3 tempers" is in your process?

Mike
 
Last edited:
Ed, you mentioned CPM154

"Up until just a few years ago, there were no steels that I can think of that were produced specifically with knife blades in mind (the only one that even comes close was 154CM and that came about from other steels)."

What about O1? O1 was and is used in industry for cutting blades, seems like it was a good fit for making into knives.
 
Sam its best not to generalise like that. Blanking knives are also made out of S5. Also some blanking knives in the tool inudstry are made from Micro Melt Maxamet Alloy. There is a world of difference between S5 and Maxamet which can be hardned to HRC 70.

The real answer is that tool steels have for decades been specifically designed and used in knife applications. Blanking, shredding, chipping etcetc
 
Here's my perspective.... The reason I do all of this testing and experimenting in my shop, using the mentioned "established standards" as a starting point is simply because NONE of the established data is based on using this steel, or any of the others for that matter, to create the items/uses we are. It has always been my thought pattern to take the next logical step(s), as it relates to what I am trying to achieve with the given steel.

Concerned "established standards".... if you've ever looked at/used a heat treating manual, look closely at the first couple of pages or the inside cover page. You will find a "key" which usually states something like "all of the data in this book was achieved utilizing as 1" cubic section of the indicated material", or something of that nature. What thats telling you is that they used a 1"x1"x1" piece of the steel type to acquire the data. Now I don't know about you folks, but I have NEVER made a knife blade that is 1" thick, and there is a difference! I gotta run to an appt. but will continue this a bit later!

OK, back from my appt. To continue with where I was going. The "industry standards" are based on the broadest spectrum of uses for the indicated steel(s). In other words, if you treat the steel according to those specifications, it SHOULD be OK for MOST applications. Thats just not good enough for me. I'm using the steel for a very specific function, but within that function there are several distinct characteristics that I desire. I want not only excellent cutting ability, but also the toughness and durability, but just as important is the ability of the end user to be able to easily resharpen the blade. Its going to be up to the individual maker to determine the characteristics that they desire, and strive for those. What I desire may not be the same as you, but the methods that may seem the long way around to you, are whats required to achieve what I desire. Its always a give an take situation with blades...you give up a little of one thing to acquire a little of another. The trick is finding the balance point where all the aspects/attributes are where YOU want them.

Since I came back to complete this post, I read the two after it. Thanks for the support Mitch, I think you understand where I'm going with this. Sam: Try what you mentioned in your post! Seriously, thats how we learn. If it works out where you achieve the overall package you desire in a finished blade, then you've solved the riddle of 52100 for yourself.

OK, I gotta finish this up and get to work. I'll describe how I came about the 3x thing with 52100. I already talked about the thermal cycling in a previous post, so I won't repeat that here. The 3x quench can be either an edge or full quench, what the spectrographs showed is that no hardness change takes place, but the the 2nd and 3rd quenches show successively reducing amounts of retained austinite, as well as refinement of the grain. (NOTE: My method is to quench the blade, and allow it to cool IN the oil, to approx. room temp before doing a 2nd or 3rd quench) What I found odd about it at first was that after the 2nd and 3rd quench a file would cut the steel more easily, which made me believe the steel was getting softer. But there was no Rc hardness changes...the testing proved that. Now that could simply be that the surface was decarbed and cut more easily, but even after a clean-up grinding, it still held true. It stood to reason with me that if I was setting things up in that manner, then the 3x tempering would be of benefit too. It was only a guess at first, but in the end the testing proved that guess to be the correct one. The gentleman who does the testing for me has a way more technical vocabulary than I, and his input was, in my translation, that due to the specific amounts of Cr in both 52100 and 5160, these method did indeed improve this particular steel for knife blades over a basic "heat, quench, temper" method.
Some may call it "Voodoo", or whatever they want, but its proven that it creates a MUCH better blade from 52100, than one that has been treated to industry standards. OK, gotta get to work, will check back this evening!
 
Last edited:
Sam, excessive grain growth occurs with 52100 when you soak it at the recommended "industry standard" temp. prior to hardening.
Guys like Ed Caffrey and Ed Fowler have developed ways to optimize 52100 for knife blades by "setting up" the steel with sub-critical normalizations prior to the quench so that the soak can be avoided.
All of us are lucky to have Ed Caffrey share his knowledge with us; he has spent years developing his 52100 and knows what he is talking about.
- Mitch
 
Last edited:
Then why not drop the temperature, and increase the time? It seems that would solve the problem. Or maybe go to a higher temperature for a shorter amount of time?
 
Sam Salvati said:
Then why not drop the temperature, and increase the time? It seems that would solve the problem. Or maybe go to a higher temperature for a shorter amount of time?
Click to expand...
Sam, I'm just finishing up my studying with 5160, and just about ready to start working with 52100, but from what I've gotten from these guys that have spent a lot of time and energy working to optimize the performance of 52100, is that the steel is very quirky; that as soon as you hit critical for ANY amount of time, excessive grain growth occurs. YET, you still have to get it to critical prior to the quench, or it won't get hard.
To avoid hitting critical when doing their sub-critical normalizations, some guys use magnets; the critical temp. of 52100 is something like 50 degrees F above non-magnetic.
Sam, these guys didn't spend a lot of time developing 52100 because they didn't have anything better to do. :)
 
You must log in or register to reply here.
Back
Top