Thermal Cycles to refine Grain Structure(O1, 1084, 1095)

Joined
Aug 12, 2002
Messages
2,259
First off, I hope that was a descriptive enough topic, after all the hoopla about that. :)

Anyways, I'm not forging blades yet, still working by stock removal, but I wnat to make the best knife I possibly can to respect my customers.

I have heard mentioned often, including again int hat thread that had the amazing picture of grain structure in one of Ed's knives, mention of thermal cycles to refine grain structure. I am well aware that I will not be able to get the same type of grain patterns/matrices with simple thermal cycles as Ed or other forgers do by combining forging with thermal cycles, but I would like to improve my knives as much as possible.

So anwyays, currently working with O1, and will add the previously mentioned 10xx steels to the line up with my next purchase of steel. So how exactly do y'all go about the thermal cycling. If I understand right, a common practice is three thermal cycles(and I already do triple tempering). What exactly comprises a thermal cycle if you're not forging? Is this just normalzing the steel(which to my understanding is bringing up to non magnetic then letting cool to room temp in the air) before your heat treat and temper cycles? Or do I need to anneal(which again, to my understanding is bring up to non magnetic, then cool to room temp SLOWLY, 50 deg. per hour, by keeping it in something such as sand, ash or vermiculite)? Any help would be greatly appreciated, as I would like to start this process each night for next three nights on two knives I'm working on so I can heat treat and temper them this weekend(or at least start the process). I will be calling around for firebricks this evening, hopefully.

Also, on normalizing, does it matter if I bring up whole piece of steel to nonmagnetic at same time, or as long as all of it's nonmagnetic at some time in process, would what be at least pretty good? My thinking is that since you're not quenching it, just letting it air cool, that it probably wouldn't be quite as crucial to have it all nonmagnetic at one time, which would be easier as I could do that with torch now, even though I don't ahve the firebricks and one brick forge yet(if need be,though, can get it all non magnetic in grill with hairdryer blower. I know, first three knives actually got too hot in that method and so all had very coarse grain structure. This is what I'm trying to avoid, and may actually replace the one of those that I sold, as it was to a good friend).
 
im interested in what the experts say too, especially with respect to O1 steel.

although i have read that O1 and the 10 series steels dont really gain from multiple cycles (not like 5160). tempers maybe, but not quenches.

from what i understand, it is always good to normalize before hardening to make for less of a chance of warping. i dont suppose it would be too crucial whether you got the entire blade up to critical at once, as long as its close. just make sure you dont over heat any of it.
 
Yeha, overheating has been my problem in the past, so I have a pile of rare earth magnets around so I can make sure to test as soon as it comes up to critical.

I dunno, best bet, if I can't find an answer, is to find my own, which I will start doing this summer if we haven't come up with good answers, and possibly even if we have. finishing school will give me time to start just taking my extra scrap pieces of O1(and other steels) and trying thigns to see how it affects grain. So either way, we'll have an answer here eventually. :)
 
OK, putting this down in notes now, so hopefully I'll remember, but I think this is something that could be helpful for more people. Because, while I am not in ANY way diminishing this work, I admire it greatly, while Ed Fowler has learned more about 52100 than anyone on planet I think, and shared every little trick to get best blade possible, it's not as helpful for stock removal makers like me who start with O1 but still want to make as good a blade as possible out of it, with best grain structure, etc. So anwyays, I'll try to find this summer best ways to get O1 working. Course, by then I may have moved on to other steels, but can still do this. Not like it's expensive to pick up O1.
 
Good quesions and thanks for the kind thoughts.

I beleive that multiple flash and full normalizing cycles will refine grain. They may not do as refinement as in conjunction with forging, at least I have not been able to achieve the same level of refinement with normalizing cycles alone.
I strongly bleieve there is room for gaining performance by the above procedures. At least if not heated much above critical (check with a magnet) you will not be hurting anything.

You might add one more experimental blade, and add some light hammer hits to the blade between normalizing cycles.

I suggest the following experiment. Cut two blades from the same bar. Work one as is.
Give the other two flash normalizing cycles, Critical to just below critical, then back up again, do this twice, then a full nromalizing cycle to room temperature in still air. Youi can add more if you wish, they may help. Then anneal, up to 800 degrees or so is plenty, let it cool down in an oven or vermiculite surrounded (but insulated from) heated bar stock to hold the heat as long as possible. When cool go ahead and make a blade same as with the other piece.

Test for edge flex, cut, then break them and look at the grain structure in the blades and you will have an idea what happened.

If the thermal cycles work, you will find the scale from hardening will be easier to remove from the experimental blade.

Note any differences you think you notice while you work the steel, write them down, some day they may mean more than you think.

I bleieve that if you add some multiple quench hardening cycles you will find improvement also. while experimenting you may just as well answer some more questions. I have never seen multiple quench hardening cycles fail to encourage better cut and tough. Again three seems to be the key.

I commend you for seeking knowledge by participating, all to many seek ready answers and tend to ignore knowledge for the sake of expedience.

Good Luck!
 
Shortly before being advised to multiple quench I stumbled into it very much by accident. I had quenched my blade, of O1, and learned that I had left it in the quenching oil too long. I went through the HT again. The edge retension of that O1 blade is remarkable and much better than another O1 I had done to the same specifications but with only one quench. I now do three quenches before the cryo (I snap temper before cryo) and tempers. I did not realize why that blade turned out so well until someone (wish I could remember who) advised all us here to triple quench. Then it clicked with me.

Roger
 
Ok I`ll have a kick at the cat, since pretty much all I use is O-1, which I`m sure I`ve got it down to pass my journeyman test.
I forge O-1 into knives from round bar, I start by heating in my forge to bright yellow and forging to shape once I have about 50% forged to shape, I forge "cooler" and "cooler" untill my finished blade was forged at a dull red. I then continue to heat to just above dull red to say red 5 times letting cool enough to handle by hand between heating cycles. I light my smithy with only one 60 watt bulb so as to keep it dark enough to see my colors properly. once done normalizing I heat my blade to just beyond non magnetic, make sure you heat slowly as O-1 seem to change to non magnetic quicky. If you see orange you went too far, just heat till non magnetic then back in the forge for a count of 5 then into vermiculite...depends on how hot you run your forge for this, I personally won`t run welding temp when doing this. The inside color of the forge must be no more than orange.
When heat treating I pre warm canola oil to 135F in my 3 gallon quench tank, I then proceed to heat up the blade, again taking care to heat slowly in an orange color forge temp, checking frquently with a magnet. I heat the tang side first as it is usually thicker, and once I`ve attained a red color, I turn around and begin heating the blade and tang. Once you`ve reached nonmagnetic, put it back in forge for a count of 5, then as fast as I can I quench in the 135F canola oil ( full quench or differential ) I leave the blade/s in the oil till I can fish them out by hand, then they go immediately into a prewarmed modified toster oven for 2, 1 hour cycles at 365F letting cool to room temp between cycles.
This procedure gives me VERY fine structure as the boys at Ed Caffreys hammerin seen and also give a hardness of between 57 to 59 rockwell.

Enjoy
 
I highly recommend the 3+3 HT and Draw, with 10 series steels and 0-1. Thermal cycling the steel reduces the size of the grain structure=stronger edge and blade. I have done the heating, quenching,
and breaking tests. I have started to place my blades in the freezer for 12 hours, helps to further the conversion to martinsite. Some will say, you do not have to do all this to get a good blade, but I think it is worthwhile to get a couple of more points of hardness.
:)
Ken (wwjd)
 
I agree. I've been doing the normallize, triple quench, triple temper, then in the freezer method (just outside the shop door, lately) on 1095 & 1084 as well as 5160. I've gotten outstanding results from the knives I've tested and some really great feedback from some of my customers on the cutting ablility and toughness.
 
well heck, now I have more input than I have time to try all of it. :)

Thanks for the tips everyone, I guess what I'll start doing is trying these methods and variations as I finish knives(which isn't often due to lack fo time), and then come summer, I'll take your input, what I've learned tryign it myself, and get a big chyunk of steel, all from same batch of possible, and just keep trying different variations.

So thanks again, and in a few months I'll try to write up what I find out.
 
I don't feel three times is enough. Also, non-magentic isn't quite critical...it's usually about 1-200 degrees past non-magnetic.

In my moten salt, I normalize as follows:

1600, then down to black
1550, then down to black
1525, "
1500, "
1475, "
1450, "
1425, "
1400, " X's three.

If you really want to see fine grain structure break a drill bit and look inside. I haven't seen many makers achieving that type of grain...but with the above cycles you can. :)

Like Ed said, just keep notes and try everything that crosses your mind. Testing is addictive...sometimes too much so :)

Keep us posted,
Nick
 
Hrmm, was planning on using these dull drill bits to practice sharpening bits by hand, but that's a good idea, gives me something to shoot for.

Man, that's a lot of normalizing cycles, and more than I can do until I at least get propane setup for it, preferably a salt bath sometime. Sadly, have to also keep track of the time and cost involved, as I have yet to produce stuff that I can charge much for. We'll see. As y'all said, and I agree, just always keep experiementing.
 
Take a good quality worn out file and break it and observe the grain. It will be super fine and pearly grey. This piece of file will be your control sample.

Take that same file and soak it at temperatures around the upper end of the forging range for a minute or two and quench it. Break again and you will have outrageous grain growth, highly visible crystaline structures.

Thermal cycle 3x (just over critical and let cool to black heat) and then heat again and quench on the rising heat, paying close attention to your heat and quench and break, you should discover, if you didn't overheat the steel on the quenching heat, that the grain is as fine or finer than the original control sample. If it isn't then you probaly overheated the steel. You can repeat the previous steps numerous times to nail your heats and when to quench. I discovered, using this method that I have been overheating most of my blades a bit when quenching. Not enough to make an unuseable knife, they still performed very well, but enough that they didn't perform as well as I'd want.
 
Lately, I've been mulling this subject over ad nauseum. Thank you etp777 for asking.
The idea of breaking a drill bit or file is something that didn't occur to me. Very useful idea guys.;)


All the best,
Mike U.
 
Happy to help, but don't thank me misque, thank all these people who gave such great input.
 
Nick! You just gave me the biggest Eureka! of the week! Hooray!

Okay, so what's all the shouting about?

For several weeks I've been having a major crisis of heat treating, resulting in some blades that just did not perform... We will all agree this is the ultimate bad news for a knifemaker, news that could end a career. My confidence was at an all-time low. There's no point in this if your knives won't cut like crazy; that's what it's all about! You can buy lots of pretty knives, finding performing knives is a different story entirely.

Quick Aside: Always Test Every Blade Before You Put Any Finish Work In It! Always! This learned the hard way.

Anyway, I'd heard so much about the dangers of overheating steel, and I suspected that I have done so. My response was to start under-heating it. I'd get right to non-magnetic and quickly quench, making darn sure I didn't overheat the steel. Which meant, I'm pretty sure now, I didn't harden anything...

I made a major knife that wouldn't cut paper. I realized I had to get this fixed - or quit thinking of myself as a knifemaker. Something that won't cut is not a knife, even if it looks like one.

After eliminating as many of the variables in my own heat treating setup as I could I did a few small tests and slowly was beginning to conclude that I wasn't heating enough. Testing that idea is in fact on my agenda this weekend.

Nick gave me the vindication I needed, I'll push the heat just a bit more. I'll try Little Hen Knives' 5-beat delay first and go from there. I think my Reil forge runs about orange when I'm heat treating, and I think I've found a good place in there, out of the direct blast of the burner. I've been trying to heat more slowly too.

All the suggestions here tell me one thing: That I'm going to have to do some real scientific testing of my O1 - and any other steel I adopt. Which really is going to be tough, as I just don't have a lot of knife making time... Every sacrificed blade is several hours not dedicated to a beautiful one. That will be hard to overcome, but performance comes first. A pretty blade that won't cut is asinine! And an almost unbearable embarrasment, even if you do discover it all by yourself... :o

Thanks to all, good thread!
 
Dave,

It's great to see you so excited--we all know that feeling!

One more good idea, since it seems you are not using salts to austinitize, is to put a pipe inside your forge. I got this idea off of a Tai Goo video, but later found out that Tom Ferry and a couple others were using it as well. Get the forge running at a dull orange (of course, this depends on ambient light, but....) and put a pipe large enough to hold your blade inside the forge. It you have a cap on the far end of the pipe and slightly crimp the front end you will get some very nice benefits: extremely even heat; no direct flame on work; almost no oxidization, etc. Don Fogg gave me another good idea: put some sand in the far end of the pipe. You can shield the tip in the sand and pull it out at the appropriate time so it doesn't come up to heat before the rest of the blade. With this sort of set up, you can confidently count those extra 5 seconds because you know the piece won't be going from critical to 1,800F in a matter of seconds. It is also a great set up for the thermal cycles prior to forging.

John
 
Hey John-

Good addition there. I meant to put that in my post last night and just got ahead of myself.

The pipe in the forge method is what I sometimes use on my clay-coated blades. It's much easier to get an even heat on the blade with the pipe, and it's also easier to not overheat the tip with the sand in there.

Neat thread guys,
Nick
 
Hey that's a great idea, thanks John. I don't have any pipe laying around here but I think I'll wait until I do to try to fix my warped blade. Good stuff!

Dave
 
Thanks for all the input everyone. This thread turned out to be a lot more informative than I expected, esp. after it went a day or so with no input.

I just finished printing it out though, and maybe in a few years I'll have some info on O1 even slightly approaching the info Ed has made available on 52100.

Gonna start sometime in next week though(had two contracting jobs comein, and for knives I'm currently making, money I get in on those doesn't even approach what I make doing computer contracting) :) Think I'll start with two flash normalzing, one regular normalize, triple quench, triple temper, and see what that gives me. Then will continue from there.Going by walmart tonight and will pick up an extra notebook to start seeing what I can do
 
Back
Top