Hamons with NO CLAY?

I would like to discuss the process and why it works, when you can get a hamon when not using any clay.

So lets look at a knife out of W2 and trying to produce a hamon with no clay.

From my research there are 3 things that help with the hamon in W2 with no clay.

1) Temp = 1435 to 1445 degrees F.
2) Time = (this is an interesting concept in my mind because you still need to be able to have the soak time, and every time I see something about time when dealing with no clay they talk about quenching time, like 7 seconds in the Parks #50 and then 3 sec out and then back in till cool, but I don't hear about soak time. I am guessing you still have to soak the 4 or 5 minutes and then quench accordingly. )
3) Blade Geometry ( specifically speaking about cross section, say a thicker ricasso and then a really thin point )

Now, the reason I think this happens is because when you bring it to the lower austenitizing temp you still get austenite, but the geometry and the interrupted quench help with the edges getting hard and then a pause allowing for the peralite hamon area to form and then the rest to cool.

What do you guys think and what have you seen and tried for the process? Got any good active hamons?

I can't give you the Tec speak on what & why but years ago when I started I used 1084 steel in some knives that I differently heat treated in ATF fluid by dipping the bottom third of the blade into the fluid in a Painter's tray after heating to non magnetic. Then I etched in Vinegar for a day or two and it brought out a noticeable & pretty Hamon.

I can tell you that there was a line at the quenched steel & softer steel.

This was without any clay packing.

Laurence

www.rhinokinves.com

This can happen with shallow hardening steel with fine grain that limits the depth to which the steel will harden. With something like a knife blade that is less than around 1/4" thick there is not much difference between the rate that the surface and the middle of the blade hardens. As a result, the thinner parts of the blade towards the edge will cool fast enough to form martnesite and the thicker parts can't and will form pearlite. Somewhere between you will have a section that will slack quench and form a mixture of both structures to give a transition line.

Doug

I looked into this a while ago. If you haven't yet, I recommend the section in Verhoevens book for bladesmiths on grain size and its effects on hardenability. He took 1080 down to an ASTM size of 14 or 15 on a wedge shape 1 inch wide and 1/4" thick at the spine, 1/32" at the edge. When quenched, it was hardened on the thin section, and pearlit on the spine and about half way to the edge. The same would likely apply to any 10xx steel or W1 or W2. I think it was either Nick Wheeler or J. Nielson who had some nice hamons without clay, just cross section and grain refinement. Verhoeven didnt water quench, but used agitated fast oil IIRC. I dont remember what the others used.

Weatherman, some of the most beautiful hamons I did in the past were what I call “natural” (no clay, just let it happen) with 1095 and W2. Your terminology is good, you got the hot stuff, the hard stuff and the soft stuff nailed, but there are a couple of points that need clarification, only a couple, once again you seem to have it. Almost any temp over 1350F will get you some degree of austenite, longer times will get you more, and higher temperatures will put more carbon into it, and thus the time and temperature will be determining factors in how hard the quench can make your blade even if it is ideal.

Blade geometry does play a key role but it is edge to spine where it really counts, the gradually increasing “wedge” cross section is what is causing the hamon as the maximum hardenable thickness is reached. When doing an interrupted quench with oil there is no reason to go back into the oil in fact it sort of defeats the purpose of the interrupt. I can give you the full rundown on why, or I can say I have been working with such techniques for 20 years and you can trust me, but I won’t mind explaining either. The interrupt has very little bearing on the hamon because hamons actually form at around 1000F. Hamons are determined by pearlite formation and whatever austenite survives the pearlite formation will become martensite from around 450F to room temp so by the time you reach your interrupt everything has already been determined.

Limiting your austenitizing heat works well with clayed hamons because it lowers hardenability and allows the hardening to stay within the lines you have set, but with a natural hamon it will not have the same affect. Grain size and geometry replace austentizing temp in a natural hamon. But perhaps the most critical factor in a natural hamon is the quenchant. Parks #50 is not a good quenchant for the natural hamon and may force you to go with under-austenitizing to get a noticeable line. #50 is designed to harden these steels fully, so it was actually made specifically to defeat what you are trying to get. A medium speed oil will have a slower curve in the 1000F range and allow more pearlite formation. With that approach you could fully austenitize the edge at a comfortable 1475F and still get wild pearlite formation halfway up the blade.

The best oil I ever had for natural hamon in 1095 was worn out Amo-quench, a medium speed oil made by Amoco that I am not even sure is available anymore, I do remember that it smelled like ripe road kill when you quenched in it.

The pinch of vanadium in W2 does something with particle drag at Ar1 that creates the wildest natural hamon I have seen on any steel, everything from cool patchworks to flame like shapes. 1095 will form very nice rolling thunderheads along the transition that will have some wondering if you clayed it for a lazy choji style. I don’t believe that it is coincidental that both of these steels are hypereutectoid in nature and that the extra carbide plays a role, once again I can go into details and hijack your thread with metallography work, or you can trust me:3:

Kevin R. Cashen said:

Weatherman, some of the most beautiful hamons I did in the past were what I call “natural” (no clay, just let it happen) with 1095 and W2. Your terminology is good, you got the hot stuff, the hard stuff and the soft stuff nailed, but there are a couple of points that need clarification, only a couple, once again you seem to have it. Almost any temp over 1350F will get you some degree of austenite, longer times will get you more, and higher temperatures will put more carbon into it, and thus the time and temperature will be determining factors in how hard the quench can make your blade even if it is ideal.

Blade geometry does play a key role but it is edge to spine where it really counts, the gradually increasing “wedge” cross section is what is causing the hamon as the maximum hardenable thickness is reached. When doing an interrupted quench with oil there is no reason to go back into the oil in fact it sort of defeats the purpose of the interrupt. I can give you the full rundown on why, or I can say I have been working with such techniques for 20 years and you can trust me, but I won’t mind explaining either. The interrupt has very little bearing on the hamon because hamons actually form at around 1000F. Hamons are determined by pearlite formation and whatever austenite survives the pearlite formation will become martensite from around 450F to room temp so by the time you reach your interrupt everything has already been determined.

Limiting your austenitizing heat works well with clayed hamons because it lowers hardenability and allows the hardening to stay within the lines you have set, but with a natural hamon it will not have the same affect. Grain size and geometry replace austentizing temp in a natural hamon. But perhaps the most critical factor in a natural hamon is the quenchant. Parks #50 is not a good quenchant for the natural hamon and may force you to go with under-austenitizing to get a noticeable line. #50 is designed to harden these steels fully, so it was actually made specifically to defeat what you are trying to get. A medium speed oil will have a slower curve in the 1000F range and allow more pearlite formation. With that approach you could fully austenitize the edge at a comfortable 1475F and still get wild pearlite formation halfway up the blade.

The best oil I ever had for natural hamon in 1095 was worn out Amo-quench, a medium speed oil made by Amoco that I am not even sure is available anymore, I do remember that it smelled like ripe road kill when you quenched in it.

The pinch of vanadium in W2 does something with particle drag at Ar1 that creates the wildest natural hamon I have seen on any steel, everything from cool patchworks to flame like shapes. 1095 will form very nice rolling thunderheads along the transition that will have some wondering if you clayed it for a lazy choji style. I don’t believe that it is coincidental that both of these steels are hypereutectoid in nature and that the extra carbide plays a role, once again I can go into details and hijack your thread with metallography work, or you can trust me:3:

Click to expand...

Kevin, If you don't mind I would like to hear about the explanation about the interrupted quench.

Another thing that I was looking at but forgot to mention when I started the thread was the grain refinement. I have heard of doing SEVERAL declining thermal cycles to refine the grain structure and lower the harden ability.

Now after saying that, I think that is something that when reading I am missing something. How do you know that several meaning like 4 - 7 is the right number compared to 2 or 3? And if I was to do several thermal cycles what would be the temps that I should do? And doing more research I only see some of the references talking about doing it only once, and doing this when forging. Is there any benefit when doing stock removal?

And after thinking about it some more I think it would work like this:

1) cut out profiles
2) Thermal Cycling 1700, 1600, 1500, 1400x3
3) Pre Grind
4) Stress Relief 1250
5) HT to like 1435-1475, soak for 5 minutes
6) Quench in oil

Only thing that I think I am missing is thermal cycling times ( Which should be determined by the cross section of the material, so if I am using a .25" piece of steel I am guessing somewhere around 15-20 minutes each time).... I know you just remove the knife from the heat and let air cool to room temp and then continue.

If you are talking about tempering cycles in an oven 15-20 minutes will not give the steel time enough to come up to heat. Also, even though temperature is the main factor, time is relevant. I would give the knife blank at least one hour per cycle and two would be better. I would do at least two cycles and a third wouldn't hurt anything even if all it does is get the steel hot. I would temper at 425° if the W2 has a carbon content of around 0.9-1%. If the actual assay is closer to 0.8% I might try 400°. That's the problem with the W series of tool steel; the carbon can be anywhere from about 0.6-1.4%. You really need to know the actual assay.

On step 2. Your stock should come to you annealed or spherodized so there really is no reason to try to soften the steel at this point or to relieve the stress. I also would not go all the way up to 1700°, especially if the steel has a higher carbon content or you could produce grain growth. I would recommend 1600°, 1500°, and a single 1400° cycle. Be aware that this is more like wishful thinking when using a forge to heat treat with if you are going by color. Color is very dependent on ambient lighting and can be misleading. You could try to get some temperature crayons at those temperatures, I think Tempel Stix is one brand name, to increase your accuracy. I would move the normalization to step 4 and do away with the stress relief.

Your austinizing temperature prior to quenching looks good as does the soak time for W2 if it has a higher carbon content. Just make sure that you warm the oil slightly. It's counter intuitive but warm oil cools faster than cold oil.

Doug

Doug Lester said:

If you are talking about tempering cycles in an oven 15-20 minutes will not give the steel time enough to come up to heat. Also, even though temperature is the main factor, time is relevant. I would give the knife blank at least one hour per cycle and two would be better. I would do at least two cycles and a third wouldn't hurt anything even if all it does is get the steel hot. I would temper at 425° if the W2 has a carbon content of around 0.9-1%. If the actual assay is closer to 0.8% I might try 400°. That's the problem with the W series of tool steel; the carbon can be anywhere from about 0.6-1.4%. You really need to know the actual assay.

On step 2. Your stock should come to you annealed or spherodized so there really is no reason to try to soften the steel at this point or to relieve the stress. I also would not go all the way up to 1700°, especially if the steel has a higher carbon content or you could produce grain growth. I would recommend 1600°, 1500°, and a single 1400° cycle. Be aware that this is more like wishful thinking when using a forge to heat treat with if you are going by color. Color is very dependent on ambient lighting and can be misleading. You could try to get some temperature crayons at those temperatures, I think Tempel Stix is one brand name, to increase your accuracy. I would move the normalization to step 4 and do away with the stress relief.

Your austinizing temperature prior to quenching looks good as does the soak time for W2 if it has a higher carbon content. Just make sure that you warm the oil slightly. It's counter intuitive but warm oil cools faster than cold oil.

Doug

Click to expand...

Doug, sorry I was all over the place and looking into 2 or 3 books trying to get good info to answer my own questions and I think I edited that post at least 5 times.

I was specifically talking about normalizing the steel to refine the grain structure. In the ASM Heat Treater's Guide it says you can normalize W2 for 1.10 to 1.5 Carbon to 1600 to 1695 degrees F so I just bumped it to 1700 degrees F. Now, I would end up buy some W2 from Aldo so it says it has .916 C so I would just do what you said and hit 1600 degrees F and then go to 1500, 1400x3. I guess the important question is: Normalizing to refine grain structure is important, but does doing 6 or 7 thermal cycles really refine the grain structure more? And if there is a benefit does going 1600, 1550, 1500, 1450, 1400x3 help and why?

Hopefully Kevin Cashen has some data on that.

Then I would pre grind and then HT and quench. ( I would do this all in a HTing Oven )

But you did bring up something that I have not seen before Doug, the thought about tempering at different temps based on carbon content........

Three cycles is considered the point where you get diminishing returns. Also, keep in mind that W2 is already a water hardening steel, and the finer the grain, the lower the hardenability. If it gets too low, you might not be able to harden it at all. It is concievable for excessive grain refinement to cause trouble with knife shaped sections of even oil hardening steel, though the degree of refinement would have to be extreme.

theWeatherman said:

But you did bring up something that I have not seen before Doug, the thought about tempering at different temps based on carbon content........

Click to expand...

I believe this is also true but not mentioned much when austenizing at the higher and lower temps. If one were to austenize at the low end and the soak stayed the same as what one does on the high end there will be less carbon in solution and therefore when queched the blade will not be able to get as hard as it could if it were austenized at the higher end. Because you are starting with a blade that did not reach max hardness you will need to adjust the tempering heats and times. start on the low end of the tempering chart and use the brass rod test to see if its too hard. If so go up 25 degrees and try it again till you find the balance you want.

The only purpose of interrupting an oil quench is to approximate marquenching techniques. As I mentioned in my other post, the hamon is already established at least 500F higher in the cooling curve. The interrupt should occur as close to Ms (martensite start) as possible. For 1095 this *should* be around 420F, but using lower asutenitizing temperatrures or different soak times will skew things a bit. The idea behind borrowing benefits of marquenching is to allow martensite to form more evenly and without undue stress, but the most notable benefit is what is known as the auto-tempering effect. When one allows the blade to cool from 400F to ambient in the air as much as 30-40% martensite can form at temperatures where the blades own thermal mass can get e head start on tempering, this is particularly beneficial in hypereutectoid steels like 1095 and W2 which can form plate martensites. The auto-tempering effect will buffer some rather stressful things which can occur in these steels and produce a blade with some gains in impact toughness and a higher tolerance to the stressful period before the tempering. It takes several minutes for the auto-tempering range to complete and in this time you can gently keep the blade straight, should any warps begin to show, effortlessly with just you gloved hands. Not only would returning the blade to the oil not offer any benefits after the martensite transformation began, it would negate any of the benefits of the marquench.

Much ado about nothing has been made over the thermal cycling thing. One can refine grain fine just by repeating a heat to the exact same temperature. Just what happens as the grain does start to nucleate at a higher rate, and as a result increases the number and thus the size of the grains, the grain coarsening temperature will get lower. Austenitizing, grain formation and eventual growth are diffusion based processes, so they are profoundly affected by points of nucleation such as the grain interfaces known as the boundaries. The more of these boundaries you have, the greater will be the rates of these processes. So finer grain equals more grain boundaries which equals easier initiation of grain growth. So you get a point of diminishing returns unless you drop subsequent temperatures to stay ahead of that issue, but even then you will encounter a point of diminishing returns when you approach the threshold of proper autenitizing temperatures.

On the inverse, the exact same phenomenon is what is responsible for the lower hardenability. Pearlite needs points of nucleation to initiate its formation, and the grain boundaries are very good for this. The more grains, the more grain boundaries, and the quicker pearlite can form. This also applies to carbide structures in these two steels.

We also really need to separate traditional normalizing from “thermal cycling”. Normalizing is done at higher temperatures, typically in excess of 1600F, and homogenizes grain size, but also is quite affective at conditioning carbide distribution. Many of the temperatures used by bladesmiths for thermal cycling do not affect carbide distribution much at all and just merely cycles grain. Many steel, like 1095, will benefit from a good normalizing heat even if stock removed, in fact due to the lack of the forging cycles, especially with stock removed blades.

Weatherman, it's basically that if you don't have as much carbon trapped in the untempered martensite you don't have as much to release to achieve the target hardness. Basically there is less carbon to trap in the untempered martensite with 5160 than in 52100 so it doesn't take as high a temperature to let it out.

Doug

Kevin R. Cashen said:

The only purpose of interrupting an oil quench is to approximate marquenching techniques. As I mentioned in my other post, the hamon is already established at least 500F higher in the cooling curve. The interrupt should occur as close to Ms (martensite start) as possible. For 1095 this *should* be around 420F, but using lower asutenitizing temperatrures or different soak times will skew things a bit. The idea behind borrowing benefits of marquenching is to allow martensite to form more evenly and without undue stress, but the most notable benefit is what is known as the auto-tempering effect. When one allows the blade to cool from 400F to ambient in the air as much as 30-40% martensite can form at temperatures where the blades own thermal mass can get e head start on tempering, this is particularly beneficial in hypereutectoid steels like 1095 and W2 which can form plate martensites. The auto-tempering effect will buffer some rather stressful things which can occur in these steels and produce a blade with some gains in impact toughness and a higher tolerance to the stressful period before the tempering. It takes several minutes for the auto-tempering range to complete and in this time you can gently keep the blade straight, should any warps begin to show, effortlessly with just you gloved hands. Not only would returning the blade to the oil not offer any benefits after the martensite transformation began, it would negate any of the benefits of the marquench.

Much ado about nothing has been made over the thermal cycling thing. One can refine grain fine just by repeating a heat to the exact same temperature. Just what happens as the grain does start to nucleate at a higher rate, and as a result increases the number and thus the size of the grains, the grain coarsening temperature will get lower. Austenitizing, grain formation and eventual growth are diffusion based processes, so they are profoundly affected by points of nucleation such as the grain interfaces known as the boundaries. The more of these boundaries you have, the greater will be the rates of these processes. So finer grain equals more grain boundaries which equals easier initiation of grain growth. So you get a point of diminishing returns unless you drop subsequent temperatures to stay ahead of that issue, but even then you will encounter a point of diminishing returns when you approach the threshold of proper autenitizing temperatures.

On the inverse, the exact same phenomenon is what is responsible for the lower hardenability. Pearlite needs points of nucleation to initiate its formation, and the grain boundaries are very good for this. The more grains, the more grain boundaries, and the quicker pearlite can form. This also applies to carbide structures in these two steels.

We also really need to separate traditional normalizing from “thermal cycling”. Normalizing is done at higher temperatures, typically in excess of 1600F, and homogenizes grain size, but also is quite affective at conditioning carbide distribution. Many of the temperatures used by bladesmiths for thermal cycling do not affect carbide distribution much at all and just merely cycles grain. Many steel, like 1095, will benefit from a good normalizing heat even if stock removed, in fact due to the lack of the forging cycles, especially with stock removed blades.

Click to expand...

Ok, lets make sure that we separate the terms of normalizing and "Thermal Cycling". What I want to do is thermal cycling. If I understand correctly, then by reducing the the temps and cycling you allow for finer grains, equaling more grain boundaries, and the reason you reduce the temps is to allow for "greater" refinement and prevent the diminishing returns. Another benefit to more grains equaling more boundaries, is allowing perlite to form quicker, which is what we want when we are trying to get a hamon!

However, you are saying that 1600F is to high to start a thermal cycle because the grains will grow? Temps above critical will give you coarse grain growth, so a thermal cycle of 1450, 1300, 1200 might be better? Or I have even heard of 1350F to start at? Or is my original thoughts correct?


I need to sit and think more about what you said Kevin, on the interrupted quench. Thanks for the info.

Doug, thank you also!

There are two major things affected by heat cycles like this, grain size, and carbide size. When it comes to edge quality and stability, believe it or not, carbide size is the more critical of the two. In a way carbides are the opposite of grains in that high heat without a slow cool makes them finer, and low heats with a slow cool ten to makes things courser. With 1084 or steels with less carbon, this is a non -issue but with W2 and 1095 you have carbide. A full normalizing heat will refine carbide and equalize grain size (perhaps coarser but at least evenly sized- which is almost as important). Lower heats will bring the grain size down.

I regularly abrade portions of grains away during metallography sample preparation, it is what makes proper grain size estimation possible, but carbides are a different story. You can get an edge finer than the average grain size, it may not be as strong if the grain size is coarse but you can do it. It is very difficult to get an edge finer than the average carbide size because the carbides will want to resist abrasion to such an extent that they pull out of the matrix and leave a big crater where you edge should be; they will also do this during use. This is why some steels with lunky carbides have a reputation for taking a terrible edge and holding it forever.

“critical temperature” doesn’t really have a meaning, despite its common use in knifemaking. It is simply whatever you target temperature is, if you are tempering critical temperature could be 400F. Most often it is attached to the Currie point, which is not the temperature at which grains will grow. The grain coarsening temperature is the temperature above Ac1, Ac3 and Accm (depending on the carbon content) where the grain boundaries finally have enough energy, with no inhibiting factors, to begin moment. With a pure iron/carbon system this is above 1335F for carbon content of .8%. But NONE of our steels are perfect iron carbon systems. Not only do 1095 and W2 have Mn, and V for the W2, but they also have plenty of trace elements and a healthy dose of aluminum nitrides from the deoxidizing process. These extra goodies create drag and lend a stabilizing effect to the grain boundaries.



So by the very nature of austenite recrystallization, before the grains can even consider growing, they need to first dissolve all the primary carbide, cementite for the 1095 and cementite as well as vanadium carbide in the W2. This would be designated by Accm on the chart above (with a healthy bump up due to the other elements) and this temperature is well above the 1475F used for hardening heats. Below Accm you have a mix of austenite and carbide, above Accm you have all austenite. If you then continue heating you have the possibility of overcoming the other factors and growing grain, but not in proper heating for hardening. This is why normalizing if often 1600F, 1700F or even higher, it takes those kinds of heats to dissolve heavy carbide. Vanadium is particularly tough (the third toughest to dissolve) and that is why it makes W2 “fine grained”. Add enough vanadium and you can expect to have to go over 1900F to break those carbides.

I bring this all up because certain myths that have taken on a life of their own have gotten in the way of many bladesmiths advancing in their heat treatment and the bogey man of grain growth is one of the biggest because it is so misunderstood. It takes effort to eliminate the stabilizing effects of carbides and other particles in the grain boundaries and that equates to massive energy. Time is very weak in comparison to thermal energy in these processes. A steel with heavy carbide formers can be held around Accm for a VERY long time without the grain boundary movement. On other forums I have shared my data of soaks that involved many hours a bit above normal hardening temperatures with no grain growth, the retained austenite was another story. After seeing some of this data my friend and fellow master smith Ray Kirk shared his similar experience with 52100 soaked for 24 hours with no grain growth. Overheating is the number one prime cause of grain growth. Due to these factors that I have mentioned the grain boundaries will resist movement quite well until there is enough energy to achieve the grain coarsening temperature. The reason we have come to fear it so much is probably because once that temperature is reached grain growth is incredibly rapid. Overheat for just a second and your grains can be huge.

It all comes together to show that there are many, many facets to consider in the whole package of heat treating and no single one can give you the ultimate perfect outcome.

Kevin R. Cashen said:

There are two major things affected by heat cycles like this, grain size, and carbide size. When it comes to edge quality and stability, believe it or not, carbide size is the more critical of the two. In a way carbides are the opposite of grains in that high heat without a slow cool makes them finer, and low heats with a slow cool ten to makes things courser. With 1084 or steels with less carbon, this is a non -issue but with W2 and 1095 you have carbide. A full normalizing heat will refine carbide and equalize grain size (perhaps coarser but at least evenly sized- which is almost as important). Lower heats will bring the grain size down.

I regularly abrade portions of grains away during metallography sample preparation, it is what makes proper grain size estimation possible, but carbides are a different story. You can get an edge finer than the average grain size, it may not be as strong if the grain size is coarse but you can do it. It is very difficult to get an edge finer than the average carbide size because the carbides will want to resist abrasion to such an extent that they pull out of the matrix and leave a big crater where you edge should be; they will also do this during use. This is why some steels with lunky carbides have a reputation for taking a terrible edge and holding it forever.

“critical temperature” doesn’t really have a meaning, despite its common use in knifemaking. It is simply whatever you target temperature is, if you are tempering critical temperature could be 400F. Most often it is attached to the Currie point, which is not the temperature at which grains will grow. The grain coarsening temperature is the temperature above Ac1, Ac3 and Accm (depending on the carbon content) where the grain boundaries finally have enough energy, with no inhibiting factors, to begin moment. With a pure iron/carbon system this is above 1335F for carbon content of .8%. But NONE of our steels are perfect iron carbon systems. Not only do 1095 and W2 have Mn, and V for the W2, but they also have plenty of trace elements and a healthy dose of aluminum nitrides from the deoxidizing process. These extra goodies create drag and lend a stabilizing effect to the grain boundaries.

So by the very nature of austenite recrystallization, before the grains can even consider growing, they need to first dissolve all the primary carbide, cementite for the 1095 and cementite as well as vanadium carbide in the W2. This would be designated by Accm on the chart above (with a healthy bump up due to the other elements) and this temperature is well above the 1475F used for hardening heats. Below Accm you have a mix of austenite and carbide, above Accm you have all austenite. If you then continue heating you have the possibility of overcoming the other factors and growing grain, but not in proper heating for hardening. This is why normalizing if often 1600F, 1700F or even higher, it takes those kinds of heats to dissolve heavy carbide. Vanadium is particularly tough (the third toughest to dissolve) and that is why it makes W2 “fine grained”. Add enough vanadium and you can expect to have to go over 1900F to break those carbides.

I bring this all up because certain myths that have taken on a life of their own have gotten in the way of many bladesmiths advancing in their heat treatment and the bogey man of grain growth is one of the biggest because it is so misunderstood. It takes effort to eliminate the stabilizing effects of carbides and other particles in the grain boundaries and that equates to massive energy. Time is very weak in comparison to thermal energy in these processes. A steel with heavy carbide formers can be held around Accm for a VERY long time without the grain boundary movement. On other forums I have shared my data of soaks that involved many hours a bit above normal hardening temperatures with no grain growth, the retained austenite was another story. After seeing some of this data my friend and fellow master smith Ray Kirk shared his similar experience with 52100 soaked for 24 hours with no grain growth. Overheating is the number one prime cause of grain growth. Due to these factors that I have mentioned the grain boundaries will resist movement quite well until there is enough energy to achieve the grain coarsening temperature. The reason we have come to fear it so much is probably because once that temperature is reached grain growth is incredibly rapid. Overheat for just a second and your grains can be huge.

It all comes together to show that there are many, many facets to consider in the whole package of heat treating and no single one can give you the ultimate perfect outcome.

Click to expand...

Do you mean Acm? And the above graph is not showing the "bump" in temperatures due to the other elements correct?

So because of the other elements starting a thermal cycle at 1600F for W2 and 1095 is not creating grain growth but allowing it to come to solution because of the other trace elements?

And from there you can lower the the thermal cycle temperatures to stay ahead of the diminishing returns?

The chart is generic, so it only shows full equilibrium temperatures, as designated by the simple “A”. When the temperatures refer to heating they are designated with an ”Ac” (Arret Chauffant), when they are cooling they are designated “Ar” (Arret Refroidissant ). Ac transformation temperatures are notably higher, and Ar temperatures are notably lower and due to hysteresis are made more or less so by rate of heating/cooling, so to simplify the concept the charts will often pick an equilibrium spot on the middle and call it “A”.

The chart is for iron and carbon and does not account for other elements. The closer you get to get to this simple system the more reality with match this chart, but carbide formers will push all of the temperatures up. This is why you can harden 1084 at 1450F but need to go to 1800F or better for heavily alloyed and stainless steels.

With simple steels you could get grain growth with normal normalizing temperatures, but that is entirely in keeping with the true nature of “normalizing” which is to provide homogenous conditions within the steel. Perhaps larger grains, but they will be the same size. The subsequent cycles are for bringing the size down. Very slow cooling from temperatures above Acm are what you want to avoid with steels like 1095 and W2 as the carbide will pool up in coarse structures and will seek out the grain boundaries. Thus air cooling is what is used. Traditional annealing via slower that air cooling is not used by industry for these steels because it will make machining miserable and result in embrittlement issues. Spheroidizing was essentially invented to meet the needs of these steels.

One of the reason 1095 has gotten a bad rap among bladesmiths is they have tried to heat treat it like it was 1084, wood ash, vermiculite, or overnight in a cooling forge is not good for this steel. Air cool and then spheroidizing creates finer carbides and a softer, more machinable, result.

I think I understand Kevin. I am going to send you a PM.

Because we are talking about 1095 and W2, lets see if I understand this correctly:

We have carbides. The carbides are important because of edge quality and stability. A full normalizing heat (1700, 1600) will refine carbide and equalize grain size (perhaps coarser but at least evenly sized. Subsequent lower heats 1500, 1400 allow for grain size reduction.

Hey this is an open forum! Don't leave me hanging with a secret PM! This is really good reading and I for one welcome whatever crumbs fall from the table. I think many of us want to grow in knowledge and craft a better product but dont yet know what we dont yet dont know.

Shane Wink said:

Hey this is an open forum! Don't leave me hanging with a secret PM! This is really good reading and I for one welcome whatever crumbs fall from the table. I think many of us want to grow in knowledge and craft a better product but dont yet know what we dont yet dont know.

Click to expand...

Ah! There wasn't anything special about it. I just asked Kevin if he could give me a definitive answer on my process. I can see where he is going. He wants the correct info out there, and not just say yes it could work or no it wont. There are just so many variables to deal with that there really isn't any correct answer because you can change different parts and get a hamon.

I then asked him the same questions as above:

"So since we have 1095 and W2; we have carbides. The carbides are important because of edge quality and stability. A full normalizing heat (1700, 1600) will refine carbide and equalize grain size (perhaps coarser but at least evenly sized. Subsequent lower heats 1500, 1400 allow for grain size reduction. Is that the right idea?"

If I am correct then my thermal cycle of 1700, 1600, 1500, 1400x3 would work without creating oversize grain, by equalizing the carbide grain size and then reducing the cycle temps to reduce the grain size.

I feel like I understand what is being said and then second guess myself, so a yes or no would help me....

I sent Kevin a pm concerning 1095 pretty much asking for a definitive how to as well as why to do it. I have read several post concerning 1095 normalizing and interrupted quenching written buy Kevin that span from 2005 to prestent and there are changes in each which I have no doubt that time and testing have improved the entire sequence.