Hello Gary, and I do not have a good answer for you. I think this is a question for our resident metallurgical guru, Mr. Cashen. Good to see you on here.

Brion

I think I like the title of “poor slob who has made so many more mistakes than others that he may be able to tell you what not to do” even more.

Gary your question is a good one and it seems quite logical it tempering is viewed in a vacuum. However tempering is actually a VERY complex process, perhaps one of the most complex when you look at the steel on the inside. If all you wish to do it relax the stress of hardening and impart some toughness, which is the most common and obvious view of tempering, then all you really need is 1 cycle for 2 hours and you would be done.

But we do not live in a perfect world and many, many things happen in the process if heating hardened steel to 350F or above.

First, in an oven it takes a very long time to heat the steel to around 400F, much longer than it would take to heat it to 1400F. So realize that most of time of the first hour is just to get the entire blade to temp. and you need a lot of time for the changes in the steel to occur at this temperature, while they happen in seconds at 1500F they take hours to happen at 400F. So this is the reason for the standard recommendation of 2 hour tempers.

But to the heart of your question- in a single temper cycle changes occur in the steel both from the heat and from the cooling. No quenching operation is perfect in getting 100% transformation to the hard stuff you want in your blade, and when things make it to room temperature without transferring to the hard stuff it tends to be trapped in limbo. When the steel “relaxes” a bit in tempering further transformations can occur when the blade cools again, leaving microscopic zones inside that are again fresh and untempered. So the second temper just sort of finishes the process.

I typically do at least three tempering cycles on my blades because I like to walk in the exact Rockwell hardness I want, so I start low and bump up the temp until I nail it. I have a log book next to my Rockwell tester where I record the information on every test that is ever done in my shop and I have noticed an undeniable trend in the numbers with just about any tempered steel I have tested. The first series of readings right after hardening will be in range that can deviate as much as 1.5 points HRC, but after the first temper it drops to a deviation of 1 HRC, after the second tempering the deviation is smaller. By a third tempering cycle (there is no significance the number 3 BTW) the deviation is normally less than ½ point Rockwell. This all illustrates a definite evening out and normalizing of the overall strength of the blade by the retempering process.

There is some debate as to whether one needs to increase the temperature to get these gains or if the same temp will work. My observations show that even the same temperature will help in evening things out as it zaps anything converted by the previous cycle.

P.S. It is great to see a question and conversation in this forum!

Thanks, Kevin. What you explained makes a lot of sense. I appreciate you taking the time to explain to a non-metalurgist. I like to understand why the procedure works rather than just accepting that it does.

You didn't say if your tests were with a simple 10xx steel or with a higher alloy. Is this going to make much difference or will an alloy react close enough to a simple high carbon steel that it's not worth worrying about the difference?

Gary

Thank you Kevin for the easily understood explanation. I am still going to say guru. Makes me realize how much I still have to learn.

Brion

|quoted:

Thanks, Kevin. What you explained makes a lot of sense. I appreciate you taking the time to explain to a non-metalurgist. I like to understand why the procedure works rather than just accepting that it does.

You didn't say if your tests were with a simple 10xx steel or with a higher alloy. Is this going to make much difference or will an alloy react close enough to a simple high carbon steel that it's not worth worrying about the difference?

Gary

I do testing and heat treating for a large number of other people (although I am not in the business of HT'ing for others), so my results are from alloys that cover the entire gamut. While the differences are much more noticeable than you would expect for the simpler steels they become even more so, or even problematic, with increasing alloy content.

There are other subtle mechanisms at work but the most glaring issue is obviously retained austenite. I suspect you have a grip on what I am talking about Gary, but for those who want more details on the austenite thing I would direct you to the article I wrote for the American Bladesmith that has been included on this site here: The Ups and Downs of Austenite

Anything that distorts the atomic matrix of steel has the potential to stabilize austenite, nickel and chromium are the two worst culprits when in greater concentrations, but the one that we need to be aware of is good old carbon. In steels above .8% C we need to be constantly on our guard against overheating, just because a steel has extra carbon does not mean we want to use that carbon in hardening because over .8% and you will not only max out hardness, you will actually start seeing a drop in hardness if it is used in the soak. Recommended temperature for 1084 is 1500F while for 1095 it is 1475F and this is the reason why. You can watch the hardness plummet on 52100 for every degree you overheat it due to austenite that will not convert just by quenching.

The heat treatment of steel is simply a matter of redistribution of carbon, and that is what goes on in tempering. From 250F to 350F there is no real noticeable difference in hardness because the carbon movement is so slight, but there is a subtle and progressive relaxation of the atomic distortion that is responsible for all the stress. At 375F the noticeable drop in hardness can begin with the largest leaps happening from 375F to 400F. Not that this is where you stop but where you may get a drop from 65 HRC to 61HRC, a loss of 3-4 points in just 25 degrees, while later increases in temperature may yield around 1 point per 25 degree increase.

From 400F to 500F you get the greatest amount of ultra-fine tempering carbide precipitation (not observable with any optical equipment), and it is this carbon movement that effects things like retained austenite. As the austenite loses carbon it becomes much less stable and either converts to other structures at temp or will transform to martensite on cooling. Although it seems to be a bit of folksy voodoo, this is why I quench from tempering as well, aside from getting back to work quicker. But if you work with alloys that can have enough retained austenite to require sub zero treatments, it is important not to go above 375F for tempering before that treatment as you will lessen the effectiveness of the colder conversion. It is, however, very important that you then temper any cold treated blade again or you could have unpredictable embrittlement problems.

Another range to be aware of is from 450F to 550F where the tempering carbides can collect in areas and cause embrittlement. Not all steels suffer the same from this and you need to study up on your alloy of choice. L6 has a significant dip in toughness in this range while still continuing to lose hardness, while O1 merely has a leveled out plateau in toughness gains. The correct term for this is “tempered martensite embrittlement”, not “tempering embrittlement”, which occurs in richer alloys at a much higher range (around 700F). Again these embrittlement issues may be alleviated to some degree by quick cooling from the temper.

After all the superfine carbide precipitation, when the temperature increases you will get conglomeration into larger carbides that are visible with optical equipment and a much greater degree of softening until the steel is annealed again.

I went ahead and wrote this tome to give an idea of how complex the tempering process really can be, even though we bladesmiths often treat it as a simple afterthought following the more dramatic heat treatments. But in what goes on inside the steel, tempering is actually much more involved than hardening.

|quoted:

Thank you Kevin for the easily understood explanation. I am still going to say guru. Makes me realize how much I still have to learn.

Brion

You are welcome, but if you still have more to learn, well that never ends. I was the smartest back when I followed a simple recipe of heat till the magnet lets go and dunk in oil, and was making "the best knives in the world". Every new book I open, every new test I do, shows me how much I have to learn. I am now about the most ignorant I have been in my career with the expectation of becoming even more so as my questions pile up exponentially. The most ignorant people of all are the ones foolish enough to say it is simple and they have it all figured out.

I'm among the readers here who are grateful for the knowledge you generously share Kevin. That info on tempering was very enlightening to say the least and I might add, explains a lot of things I've observed.

Kevin,

Thank you. Each time that you expound on topics like this, I feel like I take a step toward understanding what is going on. I appreciate your time.

Gary