Tempering Process

J

jawilder

Joined
Jun 27, 2006
Messages
2,331
When you are in the tempering process and the blade goes into the oven at ### degrees and allow it to cool to still air room temp.... do you simply turn the oven off and leave the blade in the oven to cool very slowly, or do you remove the blade and set it outside the oven?

Also, what exactly does the normalizing process consist of? I believe it's a process of heating to critical and cooling in the air with the blade pointing up. How many times do you do this, and when ie after anneal but before grinding or after grinding and before hardening. Why do you put the blade tip up or am I mistaken on this?
 
I hear most people air cooling after tempering. You are correct about the normalizing, it is usually done three times after forging is done. I do not know a reason for pointing tip up.
 
Temp differs depending on the steel being used. You can get the data sheets on pretty much any steel online and it will tell you at what temp and how many times.
If it says temper twice at a given temp and a given time, I personally temper for the given time and take it out and allow to reach room temp then place it back in and when the given time is up I turn the oven off and allow the knife and oven to return to room temp. (with the door open) This would not be wise if you have small children in the house.
Dido what Keith H said.
 
I do it either way, afik there is no difference as long as you let the blade cool to room temp or below.

If you're stock removing, it doesn't really matter when you normalize, but if you're grinding hot I guess it couldn't hurt to wait until the blade is ground. For most carbon steels, the normalizing process is how you describe. It doesn't matter which way the tip points in my opinion.
 
Tempering is all about time at the set temperature, after you have achieved the desires results it doesn't matter much how you cool it. Say for your steel that 400F will give you 61.5 HRC after 2 hours, if you then drop to 398F it doesn't really matter unless you extend the time out to many more hours. Within the two hours you are working temperature is the determining factor and the highest temp you go to will determine the outcome. If you want to leave the blades where they are at there will be no detrimental effects, but unless your oven takes many, many hours to cool there will be no benefit to it either, so you could also just take them out to cool in ambient air. Or you could even quench them in water and get right back to work if you want, I am certain there are some who will read this line and gasp, but it really doesn't matter how fast or slow you cool things if all has been done correctly. The only problem you may have is if you were to lay a 400F blade flat on your anvil on one side, then there could be some distortion issues, and thus it is best to allow them to cool evenly if it is a slow cool. Which kind of brings me to the normalizing thing...

The reason many folks put the blade point up in normalizing is to achieve an even cooling. I put the blade point up but on and angle that faces more of the edge upward, this way the heat of the spine will tend to convect up and around the thinner edge ant tip and give a Little aid in uneven cooling gradients. The most important thing about normalizing is that everything is done evenly. The blade must be heated evenly and it must be cooled as evenly as possible in order to gain the true intended benefits of normalizing. Thus any operation that involves tossing the steel on the floor, anvil or bench is not really normalizing, or it is really lousy normalizing if it is.
 
To give you an idea of the tempering process and how it works, it's time and temperature dependent. Higher temperature for shorter time...vs...lower temperature for a longer time achieve similar results (within a certain temperature range, depending on steel type).

For example: The following list will give similar Rockwell hardnesses from tempering on the exact same piece of steel...this is not an exact time and temperature comparison for any particular steel, but just to illustrate how temperature and time at a given temperature are interwoven during tempering..The list isn't a chart that you should follow, but just an example of how temperature and time affect each other.

400 degrees F. for 1 hour
is approximately equal to
350 degrees F. for 2 hours
is approximately equal to
300 degrees F. for 4 hours
is approximately equal to
250 degrees F. for 8 hours
is approximately equal to
200 degrees F. for 16 hours

As I said, this isn't an exact chart for any particular steel and is not meant to be followed, but instead shows the general relationship of Temperature vs. Time in the Tempering process. You should use the data sheets online for particular steel types and follow their recommendations for Time and Temperature for tempering. If you wanted, however, you could take different samples of the same steel and test them (at the temps and times in the chart) and check hardnesses after tempering to find out how closely time and temperature are related for your particular steel choice. This would be a really cool test to run, just for the education aspects of it!
 
Last edited:
These are some very helpful responces. Let me throw in another question, how does the cryo process work and when is it used?
 
What we KNOW about cryo is that it reduces the amount of retained austenite.Other claims have not been substantiated. It is part of the HT process .The more complex steels benefit most from cryo and it will add 1-2 HRc points of hardness. When the quenched steel cools to room temperature it is then put into liquid nitrogen .It is then brought to room temperature and tempered. Some temper before cryo but temperatures above 300 F will stabilize the RA and reduce conversion to martensite.The 300 F temper is called "snap temper".
 
Scott,
I know you work in metals for a living, but are those figures realistic at all? If the range from top to bottom was 50-75 degrees, I might be inclined to agree, but as far as my metallurgy understands it, 400F for one hour and 200F for 16 hours are no where in the same temper range. I would be interested in how those numbers were arrived at.

(not picking an argument, just a little confused over the disparity)

Stacy
 
mete, can you tell us how "home-cryo" with dry-ice and acetone or kerosene, which I've read gets you about -100 degrees, compares to liquid nitrogen cryo which gets to -300 degrees?
 
bladsmth said:
Scott,
I know you work in metals for a living, but are those figures realistic at all? If the range from top to bottom was 50-75 degrees, I might be inclined to agree, but as far as my metallurgy understands it, 400F for one hour and 200F for 16 hours are no where in the same temper range. I would be interested in how those numbers were arrived at.

(not picking an argument, just a little confused over the disparity)

Stacy
Click to expand...

My companies literature shows that heating a bearing up for 4 hours at 300 degrees F. doesn't affect the tempering of the steel significantly, but if you go beyond the 4 hours it begins to affect the tempering. We also recommend that you not go beyond 8 hours at 250 degrees F. or 16 hours at 200 degrees F. for the same reasons. These times were arrived at through testing samples at the three temperatures and seeing how long they could be held at those temperatures before the metallurgy began to experience changes in the tempering.

I asked our metallurgists about this and was told that the tempering affect (temperature vs. time) was more or less equivalent for the three different temper cycles mentioned but may be different for each different steel type. However, there is always a correlation of higher temperature needs less time than lower temperature needing more time, regardless of the steel type. I also asked about higher temperatures and was told that the same pattern can be followed upward to about 375 degrees F for 52100 steel. The reason for the 375 degree limit is because our bearings are tempered at about 375 degrees. Once a person goes beyond that 375 degrees, then they are going past our tempering temperature. If I get a call that someone heated a bearing beyond our recommended 300 degree maximum, the first things I ask them is what temperature did it go to and for how long. Usually they say that they had their heater set at 350 degrees instead of 250 degrees and that it was only at 350 degrees for a matter of minutes, to which I reply, go ahead and heat it back up to 250 degrees and install it. It wasn't at the higher temperature long enough to adversely affect the steel.

Now, for "our" knives, we're starting our tempering from scratch (which means the 375 degree upper limit no longer applies and.........), with fully hardened and untempered martinsitic steel. However, the time vs. temperature scale still holds true, as explained above. A 400 degree F. temperature would need less time than a 300 degree termperature would need to get a particular temper. If you temper your knives twice at 400 degrees F. for 2 hours each cycle, then you could temper at 300 degrees F. twice also. You'd just need to hold it at 300 degrees F. for a significantly longer time each cycle to get the same tempering effects. How long you'd need to hold it at 300 degrees F., to attain the temper you get with the aforementioned 400 degrees for 2 cycles of 2 hours is not known to me. That would have to be arrived at through testing of your particular steel.

The actual temperatures and times may be different for each different steel type, but the scale is somewhat accurate. My point was not to give someone a guideline for tempering knives, but instead to show that there is a correlation between time and temperature. Each steel type will have it's own temperatures and time scales.

So as to clarify, the whole point of my post was to show that higher temperatures and less time at that temperature can yield the same results as lower temperatures and longer times at that lower temperature.
 
Last edited:
James , the -100 F just transforms less RA than the -300 F cryo.The -100 has been around for a long time.We used to call it 'subzero' treatment.
Scott , in the 'old days' a 400 F temper was very common in most industries .It's still very automatic for me to recommend a 400 F temper.Industrially the time considerations can be significant, especially if a continuous furnace is used.
 
mete, if I understand correctly, that means -100 is good, but -300 is better? Is it safe to presume that -100 is still worth doing, in terms of better hardness without brittleness?
If so, I'll search for the "homestyle cryo" threads, I'm sure it's been addressed before. I just ordered a furnace so I can learn to do my own HT, I'd like to be able to do my own cryo as well. The whole point for me is avoiding the cost/lag time of shipping blades out, while providing excellent quality.
 
This is a great discussion on tempering. I'd like to toss one more question out there.

How important is it to immediately temper a blade after quench? What adverse effects would you find if you had to wait one hour, or to stretch it even farther, a whole day after quench?

For now I'm using fairly simple steels (1080, 5160, 15N20) but I'd be interested in how this applies to more complex alloys as well.

Walter

PS Ok, just spent some time reading the tutorial in the newbies sticky. I'm not sure I fully understood all of it though.
 
Last edited:
Thanks Scott,
While I was aware that there was a time/temperature correlation, I did not know that it was that severe. My understanding was that the relationship was heavily weighted toward the temperature, and not so much toward the time factors.The information I had read ( admittedly, a while back) lead me to believe that a two hour temper at 350F would not be significantly affected if the time ran over by even as much as double. Your data shows that it would be the equivalent of a 50 degree overage in the temperature control. Thanks for the info.

I realize that you were just pointing out that there was such a relationship, and were not being specific.

If I get a long day with no previous commitments, it would be a nice test to HT five pieces of 1084 all at the same time. Do a Rc test and record the results of A through E. Temper for 1,2,4,8,16 hours, removing one sample at each data point. The final Rc of each test piece would establish a base line chart for the time/temp/hardness relationship. I realize that it would need to be repeated at least three times to establish a reliable set of statistics, but the initial test would be interesting to see.

Anyone with an accurate programmable oven want to burn up 16 hours of electricity? The more results, the better the final data.

Stacy
 
Byork, the stresses from the martensite transformation are very high .They can and have cracked blades that haven't been tempered immediately ! Many forum members use a toaster oven .Put a brick or piece of ceramic in it, calibrate with an oven thermometer.
James, RA is not all evil as it's tougher than martensite. The lower the temperature the more RA you remove, but you'll never remove all of it ! As far as importance , the more complex alloys the more appropriate the cryo.So for 1084 I wouldn't bother but for a CPM154 I would. Always remember that each steel has it's own characteristics, learn what they are.Always keep in mind that the best way to control RA is an accurate hardening temperature.
 
bladsmth said:
Thanks Scott,
While I was aware that there was a time/temperature correlation, I did not know that it was that severe. My understanding was that the relationship was heavily weighted toward the temperature, and not so much toward the time factors.The information I had read ( admittedly, a while back) lead me to believe that a two hour temper at 350F would not be significantly affected if the time ran over by even as much as double. Your data shows that it would be the equivalent of a 50 degree overage in the temperature control. Thanks for the info.

I realize that you were just pointing out that there was such a relationship, and were not being specific.

If I get a long day with no previous commitments, it would be a nice test to HT five pieces of 1084 all at the same time. Do a Rc test and record the results of A through E. Temper for 1,2,4,8,16 hours, removing one sample at each data point. The final Rc of each test piece would establish a base line chart for the time/temp/hardness relationship. I realize that it would need to be repeated at least three times to establish a reliable set of statistics, but the initial test would be interesting to see.

Anyone with an accurate programmable oven want to burn up 16 hours of electricity? The more results, the better the final data.

Stacy
Click to expand...

I think it would also be of value to repeat your suggested test at 375 degrees, 400 degrees and 425 degrees to see how much the results are influenced by temperature control, as well. I would expect that after doing this testing at all four of the temperatures, you'd find multiple combinations that result in nearly identical results, with only insignificant differences.
 
You must log in or register to reply here.
Back
Top