Testing times, temps, normalizing, quenching, etc.

Thanks Rick, that is my theoretical understanding of it.... but in application it just seems like a blade that has gone through the brutality of forging, and then goes directly to a hardening cycle (and from a very high austenizing heat) is going to crack all over the place. I know FOR SURE it will when you let it cool to ambient room temp, and then drive a diamond penetrating point into it on a Rockwell tester. ;)

nb11b- You are absolutely correct about the geometry affecting things. My reason for starting off with test "coupons" or "tiles" is that there's less work involved in developing the initial sample pieces, and they are easy to surface grind and then Rockwell test across the entire piece.

For steels like 52100, 5160, O1, etc. a piece will typically harden pretty darn evenly thoughout, but with steels like W1, W2, and 1095, you can get a test piece that's 66/67C on the edges, and 45C in the center (which is of course why we like those for creating blades with hamon :) ).


I have done quite a bit of tests already, but it's going to take me some time to lay it out here.

My very initial tests were done with a bar of 1084 that I cycled through the forge (1800-1950F) 5 times but did zero forging on, along with some test pieces cut from the parent bar in the as milled condition. I did different normalizing cycles in the Paragon, in the salt, different temps, different rates of cooling, and then austenized in the kiln and in the salt.

A quick note on the results, at this point I'm not seeing much of a difference between the two heat sources other than the kiln causing scale to form and giving just a bit of variation in the Rc tests.

I am very pleasantly surprised to see with 52100, the descending thermal cycles with a quench have given me steel with working conditions that seem very comparable to 52100 that has gone through an overnight spheroidal annealing cycle. :)
 
LRB said:
From what I recall, the high temp with a short soak is to purposely grow the grain to make the random sizes of it more uniform. Then by progressively lower heat cycles, the grain is made smaller and ready for the hardening quench.
Click to expand...

Exactly. So why would you want to go hotter than the temp at which the grain starts to grow? The hotter you go the bigger the grain gets. It seems like the point where the grain grows would be a good highest temp for this. You know the grain grew and should be uniform then progressively lower temps would refine the size.
 
Barry- From my experience, you can throw a blade in a forge at welding heat and blow the grain up... but then thermal cycle it back to the desired range. Not that I'm saying anybody wants to do that, but IME you can.

I think Kevin explained this on his site a lot better than I can try to do (not to mention he obviously understands this stuff at a MUCH higher level than I do!!!).... :)

http://www.cashenblades.com/heattreatment.html
 
If you were going to do something like this, what would you try?
Click to expand...

I would try to follow some sort of experimental design methodology. Not being an expert on phase state transformations in metal, I'll try to keep this more about what I do know.

One of my favorite concepts in experimental design is the latin term ceteris paribus, or "all things equal". So for this kind of test, could all of the sample tiles be cut the same dimensions? Can multiple samples be heated to the same temp, then repeat with a new set of samples of the same materials be heated in a different batch? For example, given nine samples, of three each A B C, (each letter is a different kind of metal), you could heat each group of samples to the same temp. Then do that three times at the same temp but randomize the position in the kiln. What this does is keep as many things equal as possible: same kiln, same temp, same dimensions, same soak time, random positions. At the end, the you would expect that any one of the three samples of material A would have the same hardness and grain structure, regardless of which batch it was heated in. This proves out that the location in the kiln, and the kiln itself, aren't effecting the results. If that convinces you that those things are pretty consistent, it would be safe to start changing one or two of those variables.

So what variables to change? Max temperature? Previously forged? Then look if there any things that can be used as a control. An example of a control would be testing the effects of hand forging by heating three samples of the same material to the same temp, but adding one sample that wasn't forged. This keeps the temp consistent for each sample. After heating and quenching, all three forged samples can be compared to the control. If they are the same harness and grain size, or different, either way you used an experiment to answer the question.

Something else to think about is this concept of testing to the edge of the cliff. From the data sheets I've looked at, it seems like you can increase the temperature up to a point, then the response in the material totally changes (grain blows up). It might be interesting to measure samples where each run has a linear increase in temp: 1000F, 1500F, 2000F, 2500F, etc. It takes some experience, that I don't have, to say how much to increase the temperature for each trial. But knowing where this happens would be another interesting data point.

So I'm really not trying to say how to run this experiment. What I mostly wanted to get across were the general concepts of ceteris paribus, sampling, repeated trials, controls, and test to failure. I've bought enough tool steel, and annealed enough glass, to know how expensive it would be to cut a bunch of samples, and ramp a kiln to a glowing heat. Plus time isn't free. So thanks for reading this, it certainly isn't intended as a critique of your current methods, or anything like that. I just wanted to kick in a couple of ideas to think about. I look forward to the results.

- James B
 
Nick, are you planning to do the same thermal cycles to each type of steel?

Same temperatures and quenchants for instance?

I realize that it would be difficult at best to try to do optimal cycles for each steel, but how do you expect to get meaningful results running the same cycles on different steels? Wouldn't you expect different results?

Maybe I'm too naive, but I don't understand HTing W2 the same way you would 52100.

Like, I wouldn't expect meaningful results on W2 or 1084 using my recipe for 52100, or vice-versa.

I don't know, but it seems like you're taking on more than you can easily manage.

If I could make a suggestion, it would be to have some help.

You can run a test with applicable heats and quenchants for one of the steels, and sub out comparable procedures for other steels to other guys.

For instance, I only work with 52100. Maybe I could do similar procedural steps, but do them with temps, times and quenchants commonly used for 52100.

I don't have to run my own HT, I could run something that would be most acceptable to the guys with the most experience.

IDK. JMHO. I'm simply offering to do some of the leg work for you. I do have all the necessary steel and equipment, and I can follow orders to a T.

This seems like a worthy cause, that I'm willing to donate time and energy to, but I don't want to be a buttinski either.

It would be meaningful to me, because I could do things you determine on my own steel. I'd just be re-testing 52100 by a different set of guidelines.

Maybe I could test more thoroughly, because I'd only be testing one type of steel. I'm quite sure that Aldo would hotfoot a piece of bar stock to me for next to nothing. I've got lots of good 52100 already carefully forged down from large stock.

And I have liquid nitrogen, just in case you want to see it's effect.

Up to you.
 
Whoa Mark, I think you're getting your cart ahead of my horse! ;)

I certainly agree, there is no "one size fits all" heat treating recipe. Wouldn't that be nice.....?!?!? :)

I apologize if I somehow made anyone think that was what I was after here.

The offer for help with this from you guys is much appreciated! :) Like I told Mark, I'm horrible about asking for help with anything I do... and since this thread was just a shot from the hip... I don't have anything planned out well enough to know what to ask for help with at this point.

As per the liquid nitrogen--- I'd certainly be interested in seeing the effects another maker sees with it. I have it in my shop and am a pretty firm believer in using it for air hardening steel. But for steels like 52100, it's been my experience, if the steel was carefully thermal cycled and then soaked at austenizing heat, there will be a miniscule amount of retained austenite. On the 52100 I purchased from Ray Kirk, after a 10+ minute soak at 1475F and quench in AAA, I very consistently hit 66/67 Rockwell--- so I haven't actually given a 52100 blade a cold bath for quite a few years. That probably means it would be good to mess around with it again. :)

FWIW- according to the Heat Treater's Guide, the top as quenched Rockwell for 52100 will be 66. I thought maybe my tester was off, but I have checked it many times with the test blocks and it has always been right on the money with them.... *shrugs shoulders*

Thanks again guys. :)
 
I wouldn't expect any blade smiths would ask for help, even in the face of daunting testing.

That's fine, Nick. I totally understand, especially if you don't have your procedures thought through.

I know you're busy. Just wanted to offer a hand.
 
Last edited:
Nick, you appear to be using the lower austenizing temp for 52100. I do believe that Kevin C. also gets 67 with salt and oil quench. Does the Heat Treaters Guide call for the 1550F or so austenizing cycle?
NickWheeler said:
Whoa Mark, I think you're getting your cart ahead of my horse! ;)

I certainly agree, there is no "one size fits all" heat treating recipe. Wouldn't that be nice.....?!?!? :)

I apologize if I somehow made anyone think that was what I was after here.

The offer for help with this from you guys is much appreciated! :) Like I told Mark, I'm horrible about asking for help with anything I do... and since this thread was just a shot from the hip... I don't have anything planned out well enough to know what to ask for help with at this point.

As per the liquid nitrogen--- I'd certainly be interested in seeing the effects another maker sees with it. I have it in my shop and am a pretty firm believer in using it for air hardening steel. But for steels like 52100, it's been my experience, if the steel was carefully thermal cycled and then soaked at austenizing heat, there will be a miniscule amount of retained austenite. On the 52100 I purchased from Ray Kirk, after a 10+ minute soak at 1475F and quench in AAA, I very consistently hit 66/67 Rockwell--- so I haven't actually given a 52100 blade a cold bath for quite a few years. That probably means it would be good to mess around with it again. :)

FWIW- according to the Heat Treater's Guide, the top as quenched Rockwell for 52100 will be 66. I thought maybe my tester was off, but I have checked it many times with the test blocks and it has always been right on the money with them.... *shrugs shoulders*

Thanks again guys. :)
Click to expand...
 
Barry Clodfelter said:
How about at what temp the grain begins to grow? Heat to 1500 and quench then break to see the grain. Then heat another piece to 1550, quench and break, another at 1600, quench and break. You get the idea. Stop when the grain looks larger than the prior one. Wouldn't this give you a good idea what the highest heat for normalization that you could get away with?
Click to expand...

An Excellent idea. I've always wanted to conduct a test like that myself. I second Barry's great idea ^5!
 
yes, but according to a fair number of people, you will get a rather different product at 1550 compared to 1475.
chad2 said:
Kevins website states 1475 to 1550 hardening heat
Click to expand...
 
Kevin is giving industry standards along with his own personal experience for heat treating. You have to find out what works for you.
 
jdm61 said:
yes, but according to a fair number of people, you will get a rather different product at 1550 compared to 1475.
Click to expand...

every kiln is diff. for the most part for knife makers first you want to stay in the lower temps of the hardening range, second manufacturer specs are for the most part for cross sections of 1 inch and above, third you may want to go higher in temp for knives (IF) the oil you are using is not fast enough for the steel you are using.

there are many other factors as to why people get different hardnesses from the same steel. for one example it could be that they are getting to the oil to slow making a higher temp more benificial to them.

but again in almost all of these cases they are still growing grain when going that high especialy with 52100.

the only way to check what is best for you is to first, make the time between the kiln to the oil less then a second, second heat your kiln to 1450 and let cycle for 30 mins. place blade in kiln soak for 5-10 mins, depending on thickness and quench then check hardness if it skates a file and breaks easy you are at a good starting temp. raise your temp about 15 degrees letting the kiln cycle for 30 mins again quench and now you are checking for grain growth keep gowing up 15 degrees at a time untill you notice grain growth from the first broken sample. the temp before the grain growth is your critical temp. for me this is 1490

just remember grain growth is more related to temperature then time. this is another reason why you want a low temp because for thicker knives it may need to soak a little longer. and if you are in the lower ranges of the hardening temp you wont have to worry as much on how long it needs to sit.
 
Last edited:
You must log in or register to reply here.
Back
Top