Multiple HT

[Stacy E. Apelt - Bladesmith]

In several recent threads the topic of multiple heat treatments and similar processes has come up.After several off forum emails, I thought I would explain the process to elaborate.

The email questions have run on three basic topics-
How many times can you redo the HT?
and
What happens with multiple quenches?
and
What temperature should you forge XXXXX steel at?

The first two are basically the same answer.

When steel is heated above the critical point,or A1. The structure starts to convert to Austenite. The structure will remain Austenite at any temperature above A1 and below the melting point. The grain will grow at a rate and size proportionate to the amount over A1 and the time it stays there. In the case of simple carbon steel, the temperature desired is about 50-100F above A1, and thus most carbon steels are soaked at somewhere around 1450-1500F. Soak times run from 5 to 10 minutes. Add some alloy ingredients, like Cr, Mn, W, Mo, V, and N, and you have to dissolve them,too, so you often need a few more degrees of temp and a longer soak. In the case of stainless steels, with much larger amounts of alloy ingredients ( as much as 30% in some cases), you have to greatly increase the temperature and the soak time. The range for most stainless steels is 1900-2000F and the soak tile is normally about 30 minutes.

When you quench the steel at a fast enough rate ( determined by the quenchant), the austenite supercools as it crosses the critical point, around 1400F, and converts to martensite at the Ms point. This happens around 450F. The treatments for the blade after it cools to room temperature is determined by the specific alloy, but is a process of removing as much retained austenite (RA) as possible and stabilizing the martensite just created. This process is called tempering.

Now what happens if you repeat the process ?????
Nothing but a repeat of the steps explained above. Once the steel heats back above the critical point, the remnants of the previous heat treatment are erased and the game starts over again. The score is nothing to nothing, just like before. If done many times, each is a fresh start. This precludes and damage done to the physical structure of the steel. A crack can not be undone, but large grain, too much pearlite, high RA, etc. can all be reversed and started over, when everything goes into solution.

What do multiple quenches gain?
Nothing magical, but they can be used to refine the grain and distribute the carbides. If you repeat the quench ,say three times for a good number, and reduce the soak temperature by 50F each time, the grain size will be reduced each time. Now, it will be erased the next cycle, but the starting point is smaller each time, and the temperature will be lower, so the carbides won't be affected, but the grain will not grow. You will also have two practice quenches before you do the one that will lock in the fine grain.
Is this some special treatment secret - No. It is a standard practice called Normalization.
If you do a two or three step down cycle normalization, followed by a soak and quench at the proper temperature, you will get the same results every time.

How many times can you repeat this process?
Within reason, as many times as desired. This does not mean that any damage done by forging or improper quenching will be reversed. These things are permanent. Cracks, micro-cracks, decarbed steel, etc., are there for good. I suppose there is a practical limit, but two-three-four quench cycles won't ruin the steel.

What temperature should steel be forged at?
The answer is dependent on the steel type. If the steel has nothing but carbon and iron ( with a few small alloy traces), the temperature is normally above 1600F and below 2200F. This will allow sufficient ease of movement of the steel without much danger of internal dislocation ( cracking). Adding alloy ingredients can raise the temperature that the steel should be forged at. High alloy steels need to be forged much hotter, often between 1800F and 2300F. If you forge these steels ( or any steel) at a lower temperature than optional it may form micro-cracks, or even visible cracks. These can't be repaired or reversed, and the blade is a total loss. Most blades that crack or break in a normal quench have these cracks that were formed in too low temperature forging. Now ,don't get me wrong, forging down to the lower end of the range is a good thing, but if you accidentally ( or deliberately) cross the threshold bad things can happen.

At the end of a forging session, it is good practice to do a few normalization cycles to reduce any large grain that formed.


What folks should take from this long post is that with experience many people have learned to tweak the limits and get good results from outside the box. Newer makers who try and stay within the guidelines will be more likely to get consistent results.

 

[panch0]

Thanks for the explanation Stacy!

 

[Carl_First_Timer]

Thank you. This is what I understood to be that case, and you just confirmed it.

 

[Daniel Fairly Knives]

Thank You Stacy for explaining this in such a clear manner, I finally get it!


Two questions:

1. Most of the information I find about hardening 1095 mentions three normalization cycles but I haven't seen this mentioned for 01 or 5160 even though they are more complex. Why?

2. Am I helping keep my grain more uniform (I'm a stock removal guy) if I don't overheat my steel during the initial stock removal before heat treat? Overheat to a color? Overheat to a sparkler?



I think I know the general answers and I know they can't be oversimplified, just want to hear what you think.

Thanks again for explaining so much here Stacy, I probably use your stickies as much or more than my Boye, Loveless, and McCreight books.

 

[AcridSaint]

In several recent threads the topic of multiple heat treatments and similar processes has come up.After several off forum emails, I thought I would explain the process to elaborate.

The email questions have run on three basic topics-
How many times can you redo the HT?
and
What happens with multiple quenches?
and
What temperature should you forge XXXXX steel at?

The first two are basically the same answer.

When steel is heated above the critical point,or A1. The structure starts to convert to Austenite. The structure will remain Austenite at any temperature above A1 and below the melting point. The grain will grow at a rate and size proportionate to the amount over A1 and the time it stays there. In the case of simple carbon steel, the temperature desired is about 50-100F above A1, and thus most carbon steels are soaked at somewhere around 1450-1500F. Soak times run from 5 to 10 minutes. Add some alloy ingredients, like Cr, Mn, W, Mo, V, and N, and you have to dissolve them,too, so you often need a few more degrees of temp and a longer soak. In the case of stainless steels, with much larger amounts of alloy ingredients ( as much as 30% in some cases), you have to greatly increase the temperature and the soak time. The range for most stainless steels is 1900-2000F and the soak tile is normally about 30 minutes.

When you quench the steel at a fast enough rate ( determined by the quenchant), the austenite supercools as it crosses the critical point, around 1400F, and converts to martensite at the Ms point. This happens around 450F. The treatments for the blade after it cools to room temperature is determined by the specific alloy, but is a process of removing as much retained austenite (RA) as possible and stabilizing the martensite just created. This process is called tempering.

Now what happens if you repeat the process ?????/
Nothing but a repeat of the steps explained above. Once the steel heats back above the critical point, the remnants of the previous heat treatment are erased and the game starts over again. The score is nothing to nothing, just like before. If done many times, each is a fresh start. This precludes and damage done to the physical structure of the steel. A crack can not be undone, but large grain, too much pearlite, high RA, etc. can all be reversed and started over, when everything goes into solution.

What do multiple quenches gain?
Nothing magical, but they can be used to refine the grain and distribute the carbides. If you repeat the quench ,say three times for a good number, and reduce the soak temperature by 50F each time, the grain size will be reduced each time. Now, it will be erased the next cycle, but the starting point is smaller each time, and the temperature will be lower, so the carbides won't be affected, but the grain will not grow. You will also have two practice quenches before you do the one that will lock in the fine grain.
Is this some special treatment secret - No. It is a standard practice called Normalization.
If you do a two or three step down cycle normalization, followed by a soak and quench at the proper temperature, you will get the same results every time.

How many times can you repeat this process?
Within reason, as many times as desired. This does not mean that any damage done by forging or improper quenching will be reversed. These things are permanent. Cracks, micro-cracks, decarbed steel, etc., are there for good. I suppose there is a practical limit, but two-three-four quench cycles won't ruin the steel.

What temperature should steel be forged at/
The answer is dependent on the steel type. If the steel has nothing but carbon and iron ( with a few small alloy traces), the temperature is normally above 1600F and below 2200F. This will allow sufficient ease of movement of the steel without much danger of internal dislocation ( cracking). Adding alloy ingredients can raise the temperature that the steel should be forged at. High alloy steels need to be forged much hotter, often between 1800F and 2300F. If you forge these steels ( or any steel) at a lower temperature than optional it may form micro-cracks, or even visible cracks. These can't be repaired or reversed, and the blade is a total loss. Most blades that crack or break in a normal quench have these cracks that were formed in too low temperature forging. Now ,don't get me wrong, forging down to the lower end of the range is a good thing, but if you accidentally ( or deliberately) cross the threshold bad things can happen.

What folks should take from this long post is that with experience many people have learned to tweak the limits and get good results from outside the box. Newer makers who try and stay within the guidelines and will be more likely to get consistent results.

As to grain growth - Generally our soak temperatures are above Ac1 and below Ac3, to the best of my understanding, so we can't simply say you'll experience grain growth at any time/temp above A1. We can, however, safely say that one will experience grain growth at temperatures above Ac3.

I have slight disagreement with the statement that once you reach austenitization temperature that previous grain growth is erased and that you have started with a clean slate. If you have grown the grain from previous heat treatments and continue to overheat the steel in subsequent heat treatments, you a very real possibility of compounding the growth. This is particularly relevant to steels which have higher alloying contents and thus require more time to go into solution.

Of course, I'm not a materials science expert or professional heat treater, I'm just a guy on the Internet, so take my comments with that in mind.

 

[Stacy E. Apelt - Bladesmith]

The grain growth is reduced in the quench, not in the heating. That is the purpose of the lowering cycles ion normalization.

As I said, the amount over Ac1, and how long it stays there, is what determines the amount of grain growth. In a proper soak at target, grain growth is of no consequence. At forging temps, it is more of an issue. Thus the post forging normalization cycles.

 

[Stacy E. Apelt - Bladesmith]

Daniel,
1095 needs to be fine grain and fully stress relieved to avoid problems in quenching. Higher alloy steels, like O-1, 5160, and 52100 are of less concern, but will still benefit from the normalization cycles.

Unless the steel heats up above 1400F while grinding and sanding ( not likely), you have nothing to worry about in changing the grain. Since you have no idea what the internal condition of the steel actually is when you get it, a normalization before the quench is advisable. This assumes that the steel is not a high alloy or air hardening steel.The HT for stainless and high alloy steels usually contains a stress relief soak at a temp below target.

 

[LRB]

Thank You Stacy for explaining this in such a clear manner, I finally get it!


Two questions:

1. Most of the information I find about hardening 1095 mentions three normalization cycles but I haven't seen this mentioned for 01 or 5160 even though they are more complex. Why?

2. Am I helping keep my grain more uniform (I'm a stock removal guy) if I don't overheat my steel during the initial stock removal before heat treat? Overheat to a color? Overheat to a sparkler?



I think I know the general answers and I know they can't be oversimplified, just want to hear what you think.

Thanks again for explaining so much here Stacy, I probably use your stickies as much or more than my Boye, Loveless, and McCreight books.

As long as you are doing stock removal with 01, the grain is already as good as it gets. No normalizing is necessary. Heat colors shy of red while grinding 01 will not affect the grain size, but will induce stresses in the steel. These are removed by soaking at 1250° 30 minutes to an hour before bringing the heat up to your austenitization soak temp. This step removes stresses and lowers the chances of warpage. I don't know about 5160, but any steel that has been forged needs to be normalized to equalize and reduce grain size.

 

[Nathan the Machinist]

On some steels, D2 and HSS for example, multiple HT cause profound grain growth if not "reset" first. The carbides that fix the grain boundaries on the way up the first time fail to do so the second time and the grain size can increase by an order of magnitude.

 

[Stacy E. Apelt - Bladesmith]

Good point ,Nathan. I was mainly thinking about simple steels, but you are correct that the complex stainless steels and other high alloy steels can have some problems.
What would you recommend as the best treatment to "reset" the grain boundaries.

 

[AcridSaint]

Stacy - the line I was referring to is here:

"The grain will grow at a rate and size proportionate to the amount over A1 and the time it stays there."

I believe that a more accurate representation is to say that temps over A3 will cause grain growth. The way that I read the sentence above is that any temperature above A1 will cause grain growth.

 

[Nathan the Machinist]

Good point ,Nathan. I was mainly thinking about simple steels, but you are correct that the complex stainless steels and other high alloy steels can have some problems.
What would you recommend as the best treatment to "reset" the grain boundaries.

Stacy, I'm not sure. Most of the heat treat time and temp is based upon spheroidized condition, but spheroidizing D2, high alloy stainless and highspeed steels is a long slow process and may not always be necessary. My old copy of Tool Steel Simplified suggests an anneal for HSS before rehardening, which involves a long cool down from 1600 at 20-25 deg per hour. It doesn't mention the danger with D2, but the ASM Heat treater's* guide does and I tend to trust it more. Again, the anneal for D2 is from 1600 at a rate of 20 deg per hour until under 1000. I'm sure there is a optimal process of multiple heats for grain refinement and spheriodizing to reset these steels, but I don't know what it is, nor do I see it in my literature.


*Edit: actually, it may have been "Tool Steels", 5th addition.

 

[Daniel Fairly Knives]

Thanks Stacy and LRB, this thread rocks!

 

[Rick Marchand]

Excellent thread. Knowing the basics gives you something to work from. There are lots of "ideal" HT recipes floating around the knifemaking community. When anchored in metallurgical fact, they tend to be more believable. I have learned so much from many great smiths here on the forums.

Like Stacy mentions... A little out of the box thinking is okay. With regard to forging heat, I learned early on to start off hot when moving big metal and decrease the heat as you get closer to shape. Then I met Christophe Derringer and he taught me to do my final tweeking/straightening well below the suggested temperature range.

 

[KnifeMaker.ca]

Hope I'm not hi-jacking, but any thoughts on limits to multiple temper? I had a customer ask for something I couldn't find specs for - backspring temper on CPM S35VN. I started at 1000F and incresed 25F per try. (This stuff gets a very pronounced curve drop after about 1075). I nailed it (and repeated it) at 1125 giving 47/48 RHC, but in total, there were 7, 2 hour tempers.

I don't see any problems but this customer will be back. He makes some very unique knives that are being well received.

I'm not conviced that 7 tempers leading up to 1125F is going to be the same as 2 tempers at 1125F. Is my reasoning sound? I'm thinking of starting the next batch at 1075F.... and the next batch at 1100F if ....

It's time consuming, but I enjoy working out a problem - especially for a good customer.

By the way, the first time I attempted this, I did cryo. Scott Devana later recommended skipping cryo for spring temper requirements - so the numbers above are without cryo.

Rob!

 

[gnique]

That was an excellent post, Bladesmith. I believe that you made clear some of the tenants of heat treating modern carbon steel that took an awful long time and effort for me to understand and apply. I know exactly enough about metallurgy to know that I know nothing at all. Your outline jives exactly with what I have been told by people whom I believe do know about metallurgy. Well said, succinct and understandable. But you DID leave out one tiny thing that I believe is very important: steel, even though it may be considered by some to be an organic compound, is a cold, dead, lifeless metal that responds only and exactly to heat, time and pressure.

Steel embodies the very definition of the word inanimate. Steel behaves exactly and precisely to the principles of physics and chemistry. Steel does NOT care what you think or what you believe. Steel reacts to heat, steel reacts to time, steel reacts to pressure. Steel does not react to beauty or love or education or fame or beliefs. There is no soul in steel. There is no she in steel. There is no such thing as lady steel. The essence of steel is its predictability.

Steel is the product of engineers and scientists and metallurgists. Steel BELONGS to them because they made it and they know how it works. The owners of steel have forced it to behave in ways that they want it to behave. The owners of steel do not have opinions; they have data. The owners of steel don't spit and whittle and cogitate; they manipulate steel from a known reference point. What knife makers have to work with is what they have permitted us to have. There has never been and there will never be a knife maker with a hammer and a fire who can make any modern steel better than it came from the forges and the ovens of the owners of steel.

Knife makers are at very best merely hewers and polishers of modern steel. It may sound that I espouse modern steel was handed down to mortals from the very gods. My only comment to that is that integrated over the whole, the entire body of knowledge developed by the engineers and scientists and metallurgists who have given us modern steel is almost indistinguishable from divine knowledge. We challenge their knowledge at our peril.

Just that one small point. Other than that it was a good post, Bladesmith.

 

[Nathan the Machinist]

Steel is the product of engineers and scientists and metallurgists. Steel BELONGS to them because they made it and they know how it works. The owners of steel have forced it to behave in ways that they want it to behave. The owners of steel do not have opinions; they have data. The owners of steel don't spit and whittle and cogitate; they manipulate steel from a known reference point. What knife makers have to work with is what they have permitted us to have. There has never been and there will never be a knife maker with a hammer and a fire who can make any modern steel better than it came from the forges and the ovens of the owners of steel.

Knife makers are at very best merely hewers and polishers of modern steel. It may sound that I espouse modern steel was handed down to mortals from the very gods. My only comment to that is that integrated over the whole, the entire body of knowledge developed by the engineers and scientists and metallurgists who have given us modern steel is almost indistinguishable from divine knowledge. We challenge their knowledge at our peril.

.

There are frequently a lot of short cuts in steel manufacturing. There are also a lot of broad generalizations and trying to make one thing work for many applications rather than optimized specifically for one. And when industry sets out to optimize a process for a particular application, the HT they develop frequently looks a lot different than the process spelled out by the manufacturers. I expect that most of this information remains proprietary, though we have seen enough of it over time to convince me that steel manufacturers "recipes" are simply starting points that cover the majority of applications.

Something I find interesting is when you look at the literature that is available today, they sometimes contradict each other about recommendations about specific steels, and knowledge about some subjects have changed considerably over time and between editions such as RA values for given steels and heats. So, given all this, I believe it is a mistake to conclude that the scientists and metallurgists that create a steel always know everything about it and give you the optimal HT for your application. I would argue that would be the exception, not the rule.

A good example of what I'm talking about would be "prequenching" D2 to develop ultra fine grain in the neighborhood of intercept grain size of 17 rather than 13, which is a huge difference, created by tweaking a process in a way the deviates significantly from "standard practice". Source: Teledyne VASCO.

Just food for thought...

 

[Stacy E. Apelt - Bladesmith]

Nathan,
I took a look, and didn't find much reference material either.

With CPM steels, I think the somewhat permanent boundaries setup by the particle metallurgy process may help limit the problem a bit. But, I concede that there may be limiting factors in high alloy stainless steels and multiple HT. The vanadium now being added to many blade steels will help a bit,too.

I am going to file this away as a lab project for when the new shop is done. I will take a piece of CPM 154, grind a bevel on it, and cut it into four coupons. Then I'll HT the batch at 1900F. After the cyro and temper, I'll test one coupon for hardness and then check the grain size. I will redo the three remaining coupons and retest one, repeating until the results are graphed for 1-2-3-4 HT cycles. This should provide real use info for bladesmiths.

Rob,
While tempers are a process of accumulated effect, the final temperature is more a factor than the total time. I would think that a double temper at your final 1125F would get you about the same hardness results as your seven tempers ending at 1125F. Maybe a bit higher, but not likely to be lower. You are wise to try it a bit lower the first time and see what the result is. When you do, let me know the results. What people can take from your process is -
" You can always lower the hardness in temper....but you have to redo the HT to raise it".

 

[Kevin R. Cashen]

That was an excellent post, Bladesmith. I believe that you made clear some of the tenants of heat treating modern carbon steel that took an awful long time and effort for me to understand and apply. I know exactly enough about metallurgy to know that I know nothing at all. Your outline jives exactly with what I have been told by people whom I believe do know about metallurgy. Well said, succinct and understandable. But you DID leave out one tiny thing that I believe is very important: steel, even though it may be considered by some to be an organic compound, is a cold, dead, lifeless metal that responds only and exactly to heat, time and pressure.

Steel embodies the very definition of the word inanimate. Steel behaves exactly and precisely to the principles of physics and chemistry. Steel does NOT care what you think or what you believe. Steel reacts to heat, steel reacts to time, steel reacts to pressure. Steel does not react to beauty or love or education or fame or beliefs. There is no soul in steel. There is no she in steel. There is no such thing as lady steel. The essence of steel is its predictability.

Steel is the product of engineers and scientists and metallurgists. Steel BELONGS to them because they made it and they know how it works. The owners of steel have forced it to behave in ways that they want it to behave. The owners of steel do not have opinions; they have data. The owners of steel don't spit and whittle and cogitate; they manipulate steel from a known reference point. What knife makers have to work with is what they have permitted us to have. There has never been and there will never be a knife maker with a hammer and a fire who can make any modern steel better than it came from the forges and the ovens of the owners of steel.

Knife makers are at very best merely hewers and polishers of modern steel. It may sound that I espouse modern steel was handed down to mortals from the very gods. My only comment to that is that integrated over the whole, the entire body of knowledge developed by the engineers and scientists and metallurgists who have given us modern steel is almost indistinguishable from divine knowledge. We challenge their knowledge at our peril.

Just that one small point. Other than that it was a good post, Bladesmith.

Dang this is a good post!:thumbup: I like it.

I have been thinking a lot about "traditional bladesmithing" lately and have come to realize that real "old school" traditional bladesmithing is what I am all about. I mean the guy, on whose supreme knowledge of steel rested the fate of empires. The guy who the whole world turned to when they needed a better steel, because nobody knew more about it or was more cutting edge technology for his time than the bladesmith before gun based metallurgy replaced him. What we tend to think of as "traditional bladesmithing" mostly involves the approach of 19th century blacksmiths attempting to emulate the old bladesmiths.

There was a time that bladesmiths were the "owners of the steel" we didn't turn to industry for their information or their scraps, we were that industry. I sometimes get consulted by serious industrial operations for advice on heat treating, and when I hang up the phone or walk away from the factory, I think of them old sword makers and realize that I am a Bladesmith.

 

[Rick Marchand]

Gnique and Kevin.... cut it out... my pants are getting tight!


Rick:thumbup: