Need help with my heat treating

Jim Moenck

Well-Known Member
One of my New Year’s resolutions was to really work on my heat treating of blades.I know this was something I need to get better at, so I can get consistent results on the metals that I use.
Well, I’ve hit a wall and I need some expert advice. I’d sure appreciate any help and/or insight that you can give me.Here is a synopsis of what I’ve done, and where the problem is.
The material is W2. I bought this quite a few years ago from Don Hanson, so I know it is good stock.It is 2-1/8” round stock. I heated it in my forge and used my press to make it into a manageable size.Finished stock at this point is about 3/16” x 1-1/4”.I forged 6 small blades out of it, from about 2-1/2” to 3” in length. The blades were brought up to just past magnetic and put in a pail of vermiculite to cool overnight.
The blades were ground smooth for heat treat.I have a Paragon furnace with an analog control, yes it is old. The temp setting dial and the thermometer readout do not match, and I have to set the dial higher to get to the desired temperature. I have a digital readout and thermocouple that I tested the furnace with, and I believe the thermometer on the furnace is close to accurate, but I can’t be sure just how close. I heat treated each one just a bit different to try to get the process dialed in.On this first round, I used peanut oil heated to 150 degrees as my quench.Blades were heated differently, from 1425 to 1475, and were soaked for 10 minutes. Then they were tempered for one hour at temps ranging from 350 to 450 degrees.
I took these to a friend’s shop to be tested for hardness. He set the tester using the factory rated test blocks, and checked it again after testing the blades to verify that it was correct.The blades were very inconsistent, and not nearly as hard as I expected; mostly in the 40’sand 50’s RC.What we found strange was the variance in almost every blade.Fromthe ricasso to mid-blade to tip varied from 4 to 16 points RC. Also, blades that were heated to the same temp, quenched and then tempered at increasing temps did not progress the same.I would expect blades treated at 1450/375 to be harder than at 1450/400, and those should be harder than 1450/425 but that was not the case.
I picked up some Parks #50 to see if my quench was the problem. I heated it to 100 degrees, did the same type of regimen expecting better results. I was disappointed again. I still had the big variances within the same blade. I would expect the blade to be within a couple of points at whatever temps I use and would just have to find the “right”combination to get the desired RC. Ireally expected 1460 and a 425 temper to get an even and repeatable RC61 or 62.
Can anyone give me some insight as to what I might be doing wrong? I am changing my Paragon to a digital readout to get a more accurate measure, but I don’t think that is theproblem.
Any thoughts??????
 
Last edited:
First, get your temp controls squared away. Your variation in hardness readings may not be accurate because it is difficult to get an accurate reading on other than a parallel flat section 90° to the probe. Also thin sections can be harder than thick sections, depending on the steel. You may also be testing through decarb. Run some coupon tests after you get the heat control resolved.
 
First, get your temp controls squared away. Your variation in hardness readings may not be accurate because it is difficult to get an accurate reading on other than a parallel flat section 90° to the probe. Also thin sections can be harder than thick sections, depending on the steel. You may also be testing through decarb. Run some coupon tests after you get the heat control resolved.

Thanks for your insight LRB. The test blades were flat and parallel, the test pieces were just profiled. Scale was removed on some of the pieces for a second test, with mostly same result. I think you may be right on about the temp control. I'll also try a thicker piece of steel, say 1/4".
Does the finish on a test piece affect the RC hardness; 120 grit vs. 400?
 
I second what Wick said and would also point out something else.

The material is W2........The blades were brought up to just pastmagnetic and put in a pail of vermiculite to cool overnight.

^^^ This could be another potential issue. W2 is a hypereutectoid steel and when normalizing and thermal cycling, you really don't want to cool those steels any slower than just cooling in still air.
 
Hello Jim, first I must ask if you experienced a glitch in this forums handling of the post you typed? You have many words run together with no spaces, and I have experienced the exact same thing when using I-Explorer at Knifedogs. I type in the post and it looks fine with all the spaces but when it is submitted random spaces are removed throughout my post. If this happened to you, I would like to let administration know about it because so far I am the only person to have the problem.

Now on to your problems. I have seen some of these issues before and in the past it was difficult to help the folks experiencing problems because they had picked up some of these specific techniques from people they trusted and all efforts to help was contradicted by the other advisers. So while you may get feedback from friends saying that I am over-thinking this or making things too complicated, I do need to know that you trust that I may know what I am talking about; a few people who have had some success do. Many of these processes are entirely interconnected parts of a whole system, patching together Joe’s anneal, John’s quench, and Bill’s tempering is almost always a recipe for chaos rather than success. So with that said, here we go…

The blades were brought up to just pastmagnetic and put in a pail of vermiculite to cool overnight.
Don’t do this with your steel, this one is that simple- don’t do it. You can do this with 5160, 1075, 1080 or any other steel that has less than .80% carbon, but most W2 is hypereutectoid (above .80%) and this treatment with result in segregated carbide that will make machining and later heat treatments more difficult than it has to be. For hypereutectoid steels you can spheroidize, or simply go from normalization if you are just grinding with no machining. Spheroidizing consists of heating the steel to dull red but keeping the magnet sticking the entire time and allowing it to lose color and the repeating. Or, since you have an oven, heat to around 1275F-1300F and soak for 45 mines and then shut the oven off.

On this first round, I used peanut oil heatedto 150 degrees as my quench.
You did the correct thing in ditching the peanut oil and 150F was too warm for most oils.

Blades wereheated differently, from 1425 to 1475, and were soaked for 10 minutes.
If things are working the best temp is around 1475F. There is no point in going less than 1450F and you don’t want to go to 1500F or higher. So I recommend 1475F and sticking with it. Make it your constant while you change your other parameters, which you want to do one at a time to effectively troubleshoot your system. I would also stick with 10 minutes for now as you work on other things.

350F to 450F is quite a range but for your size of stock tells me a little bit about the knives you want to make and I would stick with the 350F to 400F range until I work out the kinks. If you nail the heat treatment with this steel you will get around 60-62 HRC at 375F-400F.

Where were the HRC readings done? On sections thicker than 3/16” I would expect to see a range from 36-48 HRC but on thinner areas there will be better numbers, and this is if you did it all correctly, W2 is shallow hardening. More importantly, never take a reading on a bevel, it must be on a flat level surface or the reading is not good, no matter how much you like it the test is lying to you. Also the surface should be prepped by light grinding/polishing to around 400X for the most accurate testing; this will not only eliminate decarb, it will give more consistent penetration of the test. I always take 5 readings and average them to get my number.

I picked up some Parks #50 to see if my quench was theproblem.
Excellent! This is the single best thing you could have done to work out your heat treatment system, and the main thing that told me I could write a page or two to help you. Not because Parks is the greatest, bestest thing ever (there are other good quench oils), but because it is expensive and not as easy to procure as peanut oil, this demonstrates how serious you really are about getting things worked out. But the biggest factor is that this product is a constant among your variables. It was designed to do this task repeatedly so you won’t have to chase those results around, and every bucket will be the same, so somebody else working with it can compare and understand your results. The fact that buying the Parks didn’t solve all of your problems is not a bad thing, it allows us to focus on the other areas and not worry about the quench!

What are your post forging treatments before the anneal? A good normalizing schedule should be in order. Heat to 1600-1700F, very evenly, and air cool. You can follow this up with other heats at 1500F or so to refine things but that first one is important to refine the carbide structures in the steel and set things up for success in the later heat treatments.

Follow this up with a fine spheroidal annealing, but if you still have final hardness issues try just normalizing from 1500F. You will now have an internal structure of very fine and evenly dispersed spheroidal carbides, or, if you just normalize, a network of very fine pearlite. Either of these will be the easiest structure to put back into solution for the final hardening.

Try 1475F for 10 minutes and then full quench into the Parks, warmed to 80F-100F, keeping the entire blade below the surface and moving in a cutting motion for agitation, until it is hand warm. It would be great for troubleshooting to get some as-quenched HRC numbers but you don’t want to take too long before the temper. As-quenched should be 65-66HRC.

Temper from 375F to 400F for the HRC range I mentioned before. If the hardness is still too high you can always temper again above 400F but you can go back if you are too low.

One final note- a lot of folks who use this steel are playing hamons and most people need to realize that many aspects of hamon formation are actually in opposition to normal hardening techniques, so folks need to know to include that bit of information when troubleshooting because if there is clay or edge quenching involved all bets are off and the variables rise almost to infinity. People really need to be able to fully harden their chosen steel consistently before adding these variables to the system.
 
Last edited:
Kevin, thanks so much for your input. I was hoping you would chime in and help me out. I will study each part of your recommendations and report back. This is a lot of info and again, thanks for all your pages of info. I was on your website earlier today and found much more info that I need to study.
I'm not sure about my original post though. I typed it on word, then copy - pasted to the forum here. The font seemed to increase in size, which surprised me but I didn't notice words running together. I'll take another look though.

And thanks to everyone who helped me out here. You guys are the best.
 
Kevin, you are correct. There were quite a few words run together, and I didn't see it until you just mentioned it.
 
Great info Kevin. Thanks for taking the time.

I have a simple (I think) quick question.

When doing a true normalizing, like 1600-1700 F. , would you soak that for any time at that temp? Or just bring it evenly to that temp then air cool?

Specifically, I do the normalizing immediately after forging in my forge so it's by eye and magnet. I bring the steel up evenly but there is no real soak time. After I do my rough grinding, then I will use a more controlled heat, usually my salts to do some thermal cycling prior to quench.

My question was, is doing it by eye in the forge with little to no soak sufficient to to truly normalize things?

Sorry, not meant to hijack the thread. It just seemed relevant.
 
Don’t do this with your steel, this one is that simple- don’t do it. You can do this with 5160, 1075, 1080 or any other steel that has less than .80% carbon, but most W2 is hypereutectoid (above .80%) and this treatment with result in segregated carbide that will make machining and later heat treatments more difficult than it has to be. For hypereutectoid steels you can spheroidize, or simply go from normalization if you are just grinding with no machining. Spheroidizing consists of heating the steel to dull red but keeping the magnet sticking the entire time and allowing it to lose color and the repeating. Or, since you have an oven, heat to around 1275F-1300F and soak for 45 mines and then shut the oven off.

I have a question. With W2, I have put blades in stainless foil, and cycled with my kiln through 1650f 10min, 800f 10 min, 1550f 10min, 800f 10min, 1450f 10min, 800f 10min, then 1200f 45min, then cool to room temp. I have let the oven cycle on its own, so cooling between cycles would be slower than air cooling. Will this cause problems? Currently, performance has been very similar to my 52100, which I air cool between temps. Should I be air cooling W2 as well?

I have played with a series of austentizing temps with my W2, and with my kiln, my oil, my soak, 1460f resulted in the highest, most consistent Rc#s, 67-68. I am using Aldo's W2, which is almost identical to Don's W2 in composition. I am not disagreeing with your 1475f recommendation, to clarify, but just noting the importance of knowing your equipment. Thermocouples might be off a few degrees, and this is the reading my kiln gives me for my best results.

Thank you for sharing your knowledge again.
 
First of all, I want to say THANK YOU to you Kevin for your help. I am a lot happier today, thanks to you.

I read and re-read Kevin's post, as well as the info that is a sticky with links that is at the top of this forum. I'd forgotten that that was posted some time ago, and there is a lot of good info there. Here is a quick run down from the last few days.

I made 5 test pieces; two biscuits and 3 blades that I had ready made up. I normalized each one separately by heating them to 1700 degrees, then air cooled them. Then I dropped the temp to 1500 degrees, brought them up to temp, and again air cooled them. Next is where I screwed up. I spherodized them at 1375 instead of 1275 like I was instructed, and I didn't realize it until today when I was going over my notes. They were brought up to heat, soaked for 45 minutes and then I shut the kiln off with them inside and left to cool.

Next I surface ground all of them to .141 - .142 thick. I wanted to take out any variable I could. Then I set my kiln for 1475 and my Parks #50 oil for 80 degrees. Each piece was heated separately, the first one not going into the kiln until it was up to 1475. Again, I wanted to keep that a constant. After a 10 minute soak, they were quenched in the Parks. I was surprised that the oil did not go up in temperature a lot. I was able to keep it within 10 degrees or so for all 5 blades by putting it on the (cold) concrete or heating it, or just setting it on a piece of cardboard.

I tempered 3 pieces yesterday, and 2 this morning so there was a bit of time difference there and I'm not sure what effect that might have caused. Each piece was tempered at a different temp, since I wanted to really see how much different each would be, and where I needed to adjust. Here is the result of those tests. Each piece has 3 readings. The biscuits were front, middle and back. The blades were at the ricasso, middle of the blade and near the tip.

#1. Tempered at 350 for 1 hour. RC 64.3 69.5 63.5
#2. Tempered at 375 for 1 hour. RC 59.7 61.8 63.5
#3. (Blade) Tempered at 400 for 1 hour RC 63 61.7 62.6
#4. (Blade) Tempered at 425 for 1 hour RC 61.7 61.5 62
#5. (Blade) Tempered at 450 for 1 hour RC 61 61 60.6
and on the back of this one RC 60.5 60.9 61.7

I can't tell you how happy I am. And I owe this to Kevin, so thank you Kevin. I'm in the process of converting my kiln to digital so I can get more precise temps. My previous processes were sloppy, and I paid for it. I realize how critical it is to keep everything accurate every time. I don't know what effect the 100 degree higher temp on the spherodization had, but I'll watch it closer next time and check the difference.

I still have a long way to go, but I think I'm on the right path now.
 
[FONT=&quot]Jim, I can’t tell you how happy your update and kind words made me. Note the interesting and very classic trend in the HRC numbers from tempering. I have always observed that the most dramatic drop from 65hrc to 63-63.5HRC to occur in the initial stages of any tempering operation, and then after that you struggle for every extra point you shave off with these hypereutectoid steels. I believe the initial transition from the body centered tetragonal alpha martensite to the more stable body centered cubic to be responsible for this initial change with slower precipitation of tempering carbides accounting for the subsequent more gradual lowering of HRC.[/FONT]

[FONT=&quot] [/FONT]
[FONT=&quot]Way to go with the test coupons! This is exactly how I approach this type of troubleshooting. To be honest you have your ducks in such an effective row, you probably would have had this worked out in short order even without my input, but this way everybody gets to benefit from your work and findings. [/FONT]
 
Kevin, again I thank you. I may, or may not have figured it out on my own. I know that asking for help made it a lot quicker. I have always been reluctant to ask for help, always try to figure it out for myself. I trusted you, and you didn't let me down.
What is your take on my (small) screw-up on spherodizing temp; 1375 vs. the 1275 you recommended? Will it have any factor in the next batch, or how did it affect this batch?

You were spot on with your endorsement of using a good quench oil, the Parks #50 made me a believer. If found out it even came from your shop years ago, and that makes it a little bit special for me.

Again, thanks. I'm indebted to you. I hope I get to meet you face to face someday.
 
Agreed on the Parks 50. 1095 and W2 take on a new life with that stuff! I have tested it on every carbon steel I play with. 1080+, 1084, 1095, W2, O1, O7, Blue, White, 52100, 15n20, CFV. Some of those steels don't need the fast quench, and I usually use warm canola for CFV, 52100. But it made a big difference when heat treating shallow hardening steels for sure!!!
 
I've got most of those steels laying on my bench, and I'll give each one of them a test eventually. Parks is definitely part of my regimen now.
 
Back
Top