Tempering question

i4Marc

i4Marc

Joined
Oct 19, 2011
Messages
5,052
I have a tempering question for the steel gurus.

It seems that we tend to chase the highest as quenched HRC numbers. Why? If I want my knife to net, say, 60HRC why does it matter whether I get 63 out of the quench or 66 as long as it's higher than the number I'm tempering to?

What happens in the steel of a blade that is tempered to 60HRC from 66 vs. a blade tempered to 60 from 61.5?
 
The higher the as-quenched hardness, the more complete the transformation to martensite, and the more perfect the carbides. This will still be the case when tempered down to Rc63.

Lets put it in terms most men can understand. One woman is a ten, another is an 8. If all you need to be happy in life is a 5 who can cook, does it matter which you pick?
 
Steel in the annealed condition has twice the carbide volume as is found in hardened steel. The higher the austenitizing temperature the more carbide dissolves in solution, even in high alloy steels, there is a point where all or most carbides dissolve.

Carbon and alloy from dissolved carbides goes into solution forming martensite upon quenching. Quenching from too high aus temps creates retained austenite and quenching from too low a temp will result in under hardening.

Quenching from the high end of a steels aus range will produce a finished steel with more abrasion resistance, quenching from the low end of the aus temp will produce higher toughness.

A steel like 52100 for knives seems to do best when quenched from the middle of the suggested aus temp range.

I am speaking in general terms here of course, every steel behaves differently and there is an optimum hardening and tempering temperature for every steel and it’s application.

Hoss
 
Good info. Thank you very much.

I ran some coupons today which brought up a few questions. A couple of them are now answered.

My other question is about best coupon heat treating protocol. I made a few coupons from 1/8" 1084 and a few from 1/4" 1084. The blade I have ready for HT is 3/16" 1084. Is it a good idea to test samples that are the same thickness as the blade stock or use 1/8" coupons because it's closer to the thin cross section of the blade that will harden the most?

The reason I ask is because of my HT sample results. See below.

At 1490˚/5 min., sample #2 looks good. That is an 1/8" thick coupon. However, the same procedure ( with a few more minutes soak due to the thicker coupon ) run on the 1/4" sample was a bit disappointing. The blade stock is thicker than 1/8" but the blade cross section by the cutting edge is thin of course. I would think that the thinner coupon is more representative of what will happen at the most important part of the blade.

What is the best way to approach this?

FJrjmQN.jpg
 
If you do a statistical analysis on the numbers (analyze with the student t-test function in Excel, t.test, with two tailed unequal variance) I bet there is no significant difference (less than 5% chance they are different) between the as quenched hardness numbers between 1/8" and 1/4" steel. The average for your 1/4" values is being thrown off by the one outlying hardness value of 54. Your notes show that coupon number five also had one reading that seems an error where you wrote 53.6/63.8.

Also the manganese in 1084 should made it deep hardening.
 
The boxes that have two numbers were, as you guessed, where I had an outlying value. The second sample was taken next to the first and used in calculating the average.
 
Grain growth and plate martensite are also potential problems from too high austenitizing. And if the tempering temperature required becomes too high than you can get into the tempered martensite embrittlement range where you are no longer improving toughness but actually lowering it. On the flip side, if your hardness is lower because your quench is insufficient than you can get carbide formation at grain boundaries which you don’t want.

As long as the tempering temperature is sufficient, lower austenitizing is usually better, unless it is stainless.
 
Last edited:
Did you account for the increased decarb depth on the 1/4 due to the increased soak time?
 
I agree I’m not sure there is a difference between your 1/8” and 1/4”. You might want to stick with 1480F though.
 
Can someone address best approach to coupon thickness choice? Should the sample be the same as the blade stock or always thin to approximate the zone by the cutting edge?

If higher as quenched hardness means more complete transformation to martensite, better carbide condition and better wear resistance why might I want to go with a lower aus temp? That is to say, if I'm getting the highest as quenched hardness at 1490˚ why should I choose 1480˚? It's not like 1490˚ is unusually high for 1084 so I wouldn't think RA would be much of a concern. Would an extra temper cycle or a dip in a sub zero bath help to deal with any possible RA issues?

Also, when I surface grind a sample before hardness testing, how far is deep enough to remove the decarb layer?
 
Thanks Hoss but it's not about the 1084. I'm trying to understand process and procedure so I can apply to all the steels I use. And of course I want to make the very best product I can. I know you understand that.
 
An option is to make a coupon with a stepped feature.
You want a parallel surface for the hardness tester points, but maybe you could have a stepped coupon with the 3/16" on one end and stepped down to a 1/8" at the other end.
I can send a sketch if this doesn't make sense.
 
Still kinda looking for some guidance on whether or not coupon thickness matters, as well as roughly how deep will the decarb layer be so I'll have a better idea how much to surface grind preparing for the hardness testing?
 
Marc, you might consider using ASTM guidelines, I believe it's something like a clear, reflective surface is required.
And the min thickness is specified somewhere. I don't have access to ASTM specs anymore but you should be able to research this.

From a practical bladesmith perspective and end user consideration, what matters most?
I would use coupons that I finish forge to, then grind them down with parallel surfaces to the end blade thickness requirements and then do the test.
 
You are just starting to see that everything is related in metallurgy. No one factor alone will determine the outcome. It is way more complex that just steel type and HT temp.
Some things that all go into the stew are:
Steel internal and external condition before the HT
Austenitization pre-heat and rate
Austenitization temp
Austenitization time
Quench speed and quenchant type
Direction of steel in the oven ... it should always align with the polar N-S axis ( just kidding)
Sub-zero treatment or cryo ( when indicated)
Tempering method, time, rates, etc.


All that said, in most carbon steels used for knives the parameters are pretty simple. As Hoss said, don't overthink it. The accepted temps, rates, etc. are all well researched and available. Stick with those and you will get a good HT.

Reading some good books on the metallurgy of steel will help a lot. If you read German, Roman Landes' book, Messerklingen und Stahl is supposed to be great. I wish he would put it out in English.
 
Thanks Stacy. If you look at my test chart you'll see I'm trying to take into account as many of those conditions and parameters that I can fit on a page, and some that I can't. As I said before, the 1500/400 recipe for 1084 is fine if I was interested in making "pretty good" knives. My questions aren't limited to 1084. Like the rest of you who want to make the best knives you can with the equipment you have, there is a way of going about it. I'm trying to develop a reliable, repeatable procedure that takes into account as many of the little things as I can handle at my stage of development as a knifemaker so I know I'm putting out the best product I can. Although I get your point and certainly appreciate it, I'm sure you and Hoss didn't get where you are by not overthinking it. I'll keep doing some research. Unfortunately I don't read German. Verhoeven's text is hard enough to get through in English.
 
You must log in or register to reply here.
Back
Top