Help with 1084 heat treat.....

What I would do is make stepped coupons with each step wide enough to take HRC readings something a little less thatn 1/8" thick and then step up a couple of times to your .20" . What I actually do is make a wedge shaped (flat ground blade cross section) that is around .062" widening to .312" at the top and around 1.5" wide and then cross section it to Rockwell the end and determine depth of hardness. If going with the steps I would narrow your piece to 1.25".

Kevin,

Is this something close to what you had in mind? Serviceable at least?





They are 1.25" x 2.1" (minus the 1/2" handle), the first step is about .108" thick, the second step is about .140", the third step is about .175" and (full thickness) the final step measures .201". My little ole mill drill is not that precise. I should have milled them using the Y axis where I have less runout.

BTW thanks for the tip in your ABS grinding video about using a dowel rod to apply pressure to the grind, so much better control. Not great yet mind you, but better, I'm working on great!


Thanks....


Joe aka JaXXas
 
Hi Joe,
please be sure to remove the decarburation from the heat treated sample before taking any reading.
Heat treating a 1084 specimen i had as quenched variable numbers from 62 to 64,5 hrc and those had recieved just a little hand sanding after the HT. The same samples after being well lapped on the diamond bench stone (with a bit of water) read 65,8 - 66 hrc, every time, all over the surface... consider that i am using PCB antiscale compound, it offer some protection but at normalization temperatures a bit of decarb still happens.
Decarburation it's subtle and almost invisible to the untrained eye. On a polished surface could look like a faint haze and you can test those places with your rockwell and see the differences.
 
Last edited:
....Another thing to be sure during your testings.
Be sure that the bottom surface of your sample (contact to the rockwell anvil) it's lapped flat, without any bow. Your number will be lower if the sample will flex under the brale.
That's also why you can't reverse the sample after reading and read the bottom too. The previous indentations, now on the bottom will skew your readings....the bottom of the sample should be lapped flat. Good readings require good preparation
 
Hi Joe,
please be sure to remove the decarburation from the heat treated sample before taking any reading.
Heat treating a 1084 specimen i had as quenched variable numbers from 62 to 64,5 hrc and those had recieved just a little hand sanding after the HT. The same samples after being well lapped on the diamond bench stone (with a bit of water) read 65,8 - 66 hrc, every time, all over the surface... consider that i am using PCB antiscale compound, it offer some protection but at normalization temperatures a bit of decarb still happens.
Decarburation it's subtle and almost invisible to the untrained eye. On a polished surface could look like a faint haze and you can test those places with your rockwell and see the differences.

....Another thing to be sure during your testings.
Be sure that the bottom surface of your sample (contact to the rockwell anvil) it's lapped flat, without any bow. Your number will be lower if the sample will flex under the brale.
That's also why you can't reverse the sample after reading and read the bottom too. The previous indentations, now on the bottom will skew your readings....the bottom of the sample should be lapped flat. Good readings require good preparation


Good information!

I've been grinding before testing, using 120 grit on the first pass and then 220 grit for a final pass. I thought I had remove all the decarb but maybe not. I will pay more attention in the future for sure. I didn't realize about not testing the bottom after testing the top side. Good to know! I will work more on my sample prep post Rockwell in the future. I have some learning to go yet for keeping these things flat, a surface grinder would be nice!

Thank you Stezann


Joe
 
Hi Joe,
please be sure to remove the decarburation from the heat treated sample before taking any reading.
Heat treating a 1084 specimen i had as quenched variable numbers from 62 to 64,5 hrc and those had recieved just a little hand sanding after the HT. The same samples after being well lapped on the diamond bench stone (with a bit of water) read 65,8 - 66 hrc, every time, all over the surface... consider that i am using PCB antiscale compound, it offer some protection but at normalization temperatures a bit of decarb still happens.
Decarburation it's subtle and almost invisible to the untrained eye. On a polished surface could look like a faint haze and you can test those places with your rockwell and see the differences.


Question about your PCB antiscale compound, how do you apply it? I've heard that you heat the blade up and sprinkle it on, what are the specifics of application? The ATP-641 I'm using works OK, but similar to what you are saying it kinda loses it at normalization temps and it is slow to apply.

Thanks Stezann,

Joe
 
OK, so I received my Parks 50 quenchant. It has possibly helped. Here are the results of my last 4 tests.

1500 F, 7 minute soak, Parks 50


1500 F, 19 minute soak. Parks 50


1515 F, 10 min soak, Parks 50


1500 F, 5 min soak, brine quench


Overall I'm getting to closer to RC numbers that I think I should have but it seems I need a higher temperature. The 1515 F (#3) test is seemingly the best I have tested. I think the furnace temperature is off!? Even the brine test wasn't quite up to snuff. For my next test I plan to try 1530 F with a 10 minute soak.

All of that said I would definitely appreciate you more knowledgable folks weighing in on my tests. Again I think maybe my oven temps are off. I have looked into an additional pyrometer. Equipment suggestions and methods of installation would be appreciated. What is the best way and the best place to install another pyrometer? I hate the thought of drilling my new oven, but if I must I must!

Edited to add, the thickness for #3 is .110", .141", .160" and .2" from top to bottom in the picture.

TIA for any help!

Joe
 
Last edited:
Nice report!!
Yes i sprinkle the antiscale on the hot piece and then i put back it in the furnace.
Now the compound turns "glassy" and somewhat "bulky", so i remove the piece from the oven and "spread" the compound like butter in order to get an uniform layer, leaving a little excess building up at the edges....where the decarb has more angle of attack, so to speak.
Yes, at higher temperatures it offers less protection, at least i've had this experience. Of course with protection you get less decarb than without, regardless the temperature, so i use it also for normalization.

If you have the chance to move an hand held probe inside your furnace you will find different temperatures in different places, so it is better having your sensor as close as possible to the blade, but avoid contaminating an unshielded TC point with the antiscale compound.
 
Thanks Stezann!

I have a bag of the PCB received in the last package from USAknifemaker, I'll have to give it a try. Though I'm not sure it sounds any less involved. I assume you might use a pan or plate of steel to bake the PCB on in the kiln? Or is it OK on the ceramic? What kind of temperature do you do the bake at?


On the temperature of my Evenheat kiln, I called Evenheat today, described my adventures in heat treating and they are going to send me a new TAP controller. They say there was apparently a few that made it out the door that had some temp control issues and they think it sounds like mine probably has issues. He said the errors they had seen were in the 30-80 degrees F range which is much larger than the 'standard' error. I still have a few days until it gets here so I will still try my next test at 1530 F out of curiosity. I'm hopeful!
 
I use a gas oven, and it has the bottom made of firebrick. My blade rests on a steel holder so i have no problems...i would definitely not let the PCB touch the ceramic wool liner.
The powder will stick to the blade when it's hot (the blade is hot). insert blade in the oven, retrieve it as soon as it's starting to glow and sprinkle the powder, than back into the oven. You don't have to bake the powder in the oven before!
 
OK guys, my latest test and other developments. Honestly I'm not sure where I stand, your interpretations are sincerely appreciated. Note that I have not yet received my controller replacement from Evenheat!

So here is my latest test, 1084 steel, 1530 F, 10 minute soak, quench in Parks 50. To me it seems I'm there. Overall it is the best test I've made yet. The RC numbers look decent, probably within the accuracy of the tester. Don't know that I can do better though I may try a slightly higher temp with a longer tong handle.



Here's the break, it broke into a number of pieces.....looks fairly fine grained, though I'm not yet sure what it should look like.




So here's the rub, my latest hypothesis is that my oven temps are off (low). So I bought a 1500 F degree Tempilstik and it melted right at 1500 F degrees according to the oven. The templilstik is rated at 1% accuracy, so 15 degrees. Maybe I should try to melt it at 1485 F degrees? Maybe I'm close enough? Maybe the degree of accuracy of the tempilstik and the oven works out that 1530 F degrees is actually 1500 F? I guess I'm not really sure if my temp controller is off or not!?


To my way of thinking the austenitizing temp for 1084 steel is between 1475 - 1500 degrees F. Possibly my oven is off 40-50 degrees? The templistik test throws my hypothesis out the window, the real world results seems to confirm my oven temps are off.

Your thoughts please!
 
Last edited:
It certainly look like your steel is through hardening from the reading that you got. The slight amount of chromium in the 1084 may have helped. It's kind of hard to see the grain in the broken pieces, especially at 3AM, but it might be just a little on the course side. Get a good quality file and break it and compare the grain structure. It will give a good comparison for judging grain in broken steel. Even if it's a little course I've seen far worse. You might want to see what happens with a soak at 1500° or maybe cut the soak back to 5 minutes.

Doug
 
Ideally, you should not be able to see any "dots" when checking for grain. Agreed.....those samples look to have sort of large grain. Not blown up, but not ideal. Like Doug mentioned, a broken file or drill bit will give you what you are after. It should look just like a smooth gray surface, without any texture whatsoever. I call it a "velvety" look. I've done my own "grain size" tests on Aldo's 1084 and W2. Without normalizing (From 1650 then cycling down), the samples looked like yours. After the normalizing and cycling, they looked like they should. Just a wall of gray without any texture.
 
It certainly look like your steel is through hardening from the reading that you got. The slight amount of chromium in the 1084 may have helped. It's kind of hard to see the grain in the broken pieces, especially at 3AM, but it might be just a little on the course side. Get a good quality file and break it and compare the grain structure. It will give a good comparison for judging grain in broken steel. Even if it's a little course I've seen far worse. You might want to see what happens with a soak at 1500° or maybe cut the soak back to 5 minutes.

Doug

Thanks Doug!

Though it kills me to do it I'll break a file! :( Are you saying a soak at 1500 F and then only a 5 min soak at 1530 F?
 
For example, with his 1084 this is my procedure: Normalize to dissolve carbides...1600°F for 10 minutes, allow to cool in still air. Carbides taken care of...check. Grain growth...maybe. Let's thermal cycle to nucleate new aus grains. 1550°F, air cool. 1500°F air cool. 1450°F, air cool. Now you have a steel that is absolutely ready to harden, and should be able to reach max RC out of quench now. Soak at 1500°F for no more than 10 minutes, quench. Should read 65HRC or better.

Ideally, you should not be able to see any "dots" when checking for grain. Agreed.....those samples look to have sort of large grain. Not blown up, but not ideal. Like Doug mentioned, a broken file or drill bit will give you what you are after. It should look just like a smooth gray surface, without any texture whatsoever. I call it a "velvety" look. I've done my own "grain size" tests on Aldo's 1084 and W2. Without normalizing (From 1650 then cycling down), the samples looked like yours. After the normalizing and cycling, they looked like they should. Just a wall of gray without any texture.


Thanks Samuraistuart!

I have a test in process using your normalization process. Going to run it twice, once at your numbers and then once again at 50 F degrees higher cause I'm thinking my oven temp is off! I'll definitely post up the results.

A drill bit, even better, I have many broken drill bits! Thanks again, it is very nice to have a 'description' of what I'm looking for on the breaks!
 
No, I'm saying to reduce the austenizing temperature to 1500° for 10 minutes and see what that does to the grain size. If the grain still seems to be a little large cut the soak to 5 minutes. The reason is that temperature has a bigger effect than time.

Doug
 
More test results...

First up the 1650 degree F normalization test, Rockwell isn't where it should be, and the breaks look a bit grainy.

1650 degree F normalization procedure....




Second the 1600 degree F normalization test. Rockwell is way low but the breaks look much better to me.

1600 degrees F normalization procedure...




Third test I did, No normalization just looking to see if I could get the Rockwell numbers up, breaks are grainy.

1550 degree F heat treat, no normalization.




The 1600 deg normalization sample definitely seems to have the 'best' grain, but the RC testing was very low. The 1650 normalization sample was the best RC but the grain wasn't good. The 1550 hardness test didn't come up to snuff on the RC and of course it's grain wasn't good.

So I received the new controller from Evenheat today so I need to get that installed and I'll see what kind of results I get with that.


Thanks for your inputs!
 
I'm wondering if the larger grain on the first test gave you better hardenability thus a higher HRc. The finer grain on the middle test may have decreased hardenability and lowered you HRc. Also is that 47.5 HRc reading on that coupon in the last test where you were holding it during the quench?

Doug
 
I'm wondering if the larger grain on the first test gave you better hardenability thus a higher HRc. The finer grain on the middle test may have decreased hardenability and lowered you HRc. Also is that 47.5 HRc reading on that coupon in the last test where you were holding it during the quench?

Doug

As you will see in my next post I'm not sure I trust any of my previous testing. But I'm a believer in proper grain size and that will be my next big push. Sorry I forgot to label the tong area on the last coupon, I have adopted a firm commitment to use tongs to grip the ID notches. So no the 47.5 HRc wasn't due too tong effect.


Thanks Doug!
 
OK, I recieved my new kiln controller yesterday, installed it tonight and pushed through a quick test. I think the results are somewhat remarkable considering all the tests I've made. As I noted in a earlier post I became convinced that my kiln temps were off and Evenheat tech support agreed. Kudos to Jim @ Evenheat, a short probing discussion and 4-5 (business) days later I have a new kiln controller. Great customer support!

So here is my new controller test...


The breaks....


Maybe not perfect but the best HRc I have achieved so far at the 'correct' temperature. I think the grain is terrible but then it hasn't been normalized. Hopefully from this point it's 'just' a matter of proper normalization for grain size and tweaking the soak times for hardening.

TIA for any and all input, advice, criticisms and discussion!
 
Back
Top