This is where the "art" blends with the science for most of us....I think you'll find that everybody's methodology will vary slightly...and that's OK! The key is to work out the methodology that achieves you the "best" end product you can produce (even using the word "best" is open to various interpretations) I'm just going to copy/paste your steps....and insert the differences in mine

1) Forge to shape

2) Rough grind with the grinding wheel (since I'm not that good at forging to shape)

** Grind to shape it fine, but I do not do it

**Thermal cycle: Evenly heat the blade(s) to approx 1300-1350F....and allow to air cool in a rack to 900F or less....X3

3) Normalize--Just one cycle of heating past critical (1600 F) and then air cooling

**I use a single normalizing heat, but strive to be just slightly above the critical temp of the specific steel

4) Anneal

--High: heat to a little past critical (1500 F), cool in warm ash/charcoal dust "hot box"

--Medium: heat to red/orange heat, cool in warm ash/charcoal dust "hot box"

--Low: heat to red/black heat (shadowy), cool in warm ash/charcoal dust "hot box"

** My annealing method is totally different..a single, even heat to slightly above critical temp, and in vermiculite. Annealing is all about SLOW cooling, which means the better insulating value your annealing media hosts, the better...to date I've not found anything that does a better job than vermiculite.

5) File/grind to shape, leaving a bit extra all around for decarb.

6) Harden (one time)--heat to a little past critical (1500 F) and quench in 140 F canola oil (I'll get quenching oil asap)

7) Temper--two 2-hour heats to 425'F in the kitchen oven, cool between tempering cycles.

8) Draw back the spine once (to dark blue) using a torch (or a metal plate on the forge)

**This is an area that I often cringe at....not because it doesn't work, but because most people who ATTEMPT the soft back draw simply do not have the patience or understanding to do it correctly/successfully. If you use "color" then all your likely getting is about a .010" thick "skin" of tempered material, with the remaining "mass" of the treated area remaining harder. It's all about TIME AND TEMP....anything that changes in steel, requires a specific temp, for a specific amount of time. Generally speaking, in order to achieve tempered material all the way through a blades cross section, will require that the heat be applied for approx. 40 mins......now before anyone freaks out and says they have been successful by "color watching"....I will say that IF that is the case, then you did not achieve full hardness on your blade during the quench! Although it not popular in some circles these days, I'm a fan of differential heat treating, and always encourage folks to use that method when creating their JS or MS performance tests blade(s). It allows me to instill attributes to a blade that I simply cannot achieve with full hardening techniques.

9) Polish.

*********************************************************************

As I mentioned, there will likely be a lot of variance on how each of us do things....again....I don't think there is an exact "right" or "wrong" way.....as long as the end result is produces a quality blade.

Steve, you have already made your job fairly easy by choosing 1084, it opens up about the widest windows and margins for error for the beginner. There are a couple of flavors of 1084 out there that could require a slightly different approach but any are simpler than most other steels. Please be aware that when I give advice it is under the assumption that the person I am helping is out to maximize their results along the same lines of knife properties that I shoot for, with “good enough” rarely in my vocabulary, this has led to some tension in the past with folks who were still getting started.

Most of what you have looks very good and some of what I may suggest is merely a matter of preference in the sequences. For example, I will start with your rough grind before the anneal. With your method of annealing and the fact that you are still doing a lot of shaping with the grinder there probably will not be many problems. But I would put most grinding and stock removal shaping after the anneal just to make my life easier and avoid decarb, I do however understand where you are coming from with shaping what elude you at the hammer; with a little more practice that hammer will get you there.

“critical” is a rather elusive term and has really no set definition in discussing heat treatment, most bladesmiths either pin that label on the point of loss of magnetism (1414F)or decalescence, which are two entirely different temperatures. I tend to view it as corresponding with Ac1 on the iron carbon equilibrium diagram, which is around 1333F where the whole process of putting things into solution actually begins.

At any case I really like your choice of 1600F for one good, full normalizing heat. Normalizing does several things, it evenly distributes carbon, it makes uniform grain size, it obliterates any strain effects (unwanted stresses) from forging, but perhaps most importantly it refines carbide distribution. For the carbide part you need heat! Thus I always start out with an actual normalizing heat like they use in industry, similar to your temperature. This will accomplish all the true normalizing goals regardless of grain size. This I follow up with a couple more heats at progressively lower temperatures to refine things like grain size.

And it is here that you sort of take my “normalizing” and make it your “annealing”. My concern for this is that the slow cooling of annealing is not as conducive to the goals of normalizing as an air cool is. Once again, your saving grace is that you are working with 1084, but you would probably get some rather unpleasant side effects from your multiple cycles in wood ash with a higher carbon steel. I personally would take the first or second annealing heat and move it to your normalizing sequence and save the last low temperature heat for the wood ash.

On hardening, the number of times you were careful to point out that you harden makes – you da man! It is a pretty intense operation for both you and the steel, nail it dead on the first time, move on and be happy! Your choice of 1084 allows you to work in a range from 1475F to 1500F but avoid going over 1500F by very much. As a VERY loose and VERY general guideline steels with less than .8% carbon can go above 1500F, steels with more than .8% carbon need to stay below 1500F (better at 1475F) and ones at .8% can go to 1500F.

On your quench, don’t rush out to get any fancy oil on somebody else’s account. It has been common across the internet for people to insert that caveat about quenching oil, almost as an apology to avoid being ambushed by hiding guerilla fighters from the quench wars. Canola will work very well with your steel , just be aware that it is not as stable as a formulated quenchant and will change in performance on you if you do not replace it regularly and thus it will not take too long to equal the cost of a quench oil that could give you years of predictable performance. I would adjust your temperature down from 140F to 130F for the canola as that it where that particular oil’s cooling sweet spot seems to be maximized.

With tempering I always start out low and them bump up the temper to “walk” my hardness in with several cycles, but if you don’t have a method of determining that exact hardness it becomes a moot point. For larger knives that will be choppers, your 425F seems pretty good, but for smaller knives that will be cutters and slicers I would drop it back in the range from 375F to 400F particularly if you are going to draw the spine anyhow. Also, folks need to be aware that what will determine how hot you need to temper is often how long you heated it for hardening. A quick heat to nonmagnetic with 1084 before the quench may only require 375F to drop things to 58HRC, but a full soak at 1480F may require a temper of 425F just to get down to 61 HRC.

On the polishing… heck, maybe you have some pointers for me!

At any rate I hope this helps.

Thank you Mr. Caffrey and Mr. Cashen. It was great to check out this forum today and find, after less than a day, two very detailed, thoughtful, helpful replies from two guys who really know what they're doing.

It is clear to me that I need to get a better gauge of temperature than just my eye and a magnet, but for now that will have to do. This summer when I'm back in the States I may purchase a IR thermometer to help things along--have either of you (or anyone else, for that matter) had much luck using one of those? Maybe a couple templestiks would do the trick too.

I saw some corrections from both of you that seemed to match--for example, the lack of multiple annealing steps (I'm happy to do away with those--what a time eater!) and the addition of thermal cycling in conjunction with the normalizing heat, either before (Caffrey) or after (Cashen). I'll try both and see what I have better luck with.

Mr. Cashen: When you say a "full soak at 1480 F" prior to quenching, how long is this soak? I don't have great temperature-for-time control with my forge (and my skill level), but it's great to keep in mind what to shoot for. Also, for the anneal (the single sub-critical anneal), how low should the temperature be? Maybe 1275?

Mr. Caffrey: When you say "just above critical," you are in the 1425-1475 F range, correct? Also, how do you typically differentially harden your blades? Do you get better results from edge quenching or through using clay/refractory cement to coat the back?

Thank you for your help in refining my protocol.

And how about a post forging quench? Why do folks do this (or do they not do it with 1084)? And when, in the process, would one do this quench--I'm assuming one would do this right before annealing...perhaps after thermal cycling?

Soaking can be various times and its success is greatly dependent on the previous thermal treatments to set you up for it. If previous heats (normalizing and annealing) leave the steels constituents widely distributed and rather coarse it is going to take much more soaking to put things fully into solution, and this is the true importance of making things fine and even in normalizing and annealing.

Can you see a distinct chain of events here? Your forging techniques will determine how extensive your normalizing should be, your normalizing and annealing will determine how long you must soak to harden, and how long you soaked for hardening has a profound effect on how high you have to temper. The numerous cause and effects can be listed all day, but it is safe to say that no heat treatment is an island unto itself and thus to unlock the potential of the steel for any given application requires that you not neglect any part of the process for the sake of another. If it were easy, it really wouldn’t be worth pursuing.

Anyhow, different levels of refinement in pearlite, from fine to coarse will determine your soak time; and the same is true for spheroidized structures. Pearlite is what you form when you heat above critical (say 1500F) and cool in the wood ashes, and the slower you cool, the coarser it is. Spheroidal carbide is what you make when you heat to below critical for annealing. For heavily spheroidized steel I would recommend at least an 8 minute soak. But a very fine pearlite can sometimes work well with much less, so there is no need to fret the absence of pyrometers and controlled ovens too much since you are still working with 1084 and can reduce the amount of insulated cooling or go subcritical and make very fine spheroidal structures. 1275F would be a very nice temperature for sub critical annealing.

The post forging quench, if used, really must not be looked at as a quench but a thermal cycle for the purpose of refinement of the internal structure, sort of a heavily modified normalizing. Don’t get carried away with it, there is a whole lot of wishful thinking and hoopla wrapped up in such treatments. In general, one quench can replace two normalizing cycles in grain refinement. The idea is that the faster you cool, the finer the resulting strucures, but there is a bit more to it than that.

Another thing I was going to mention was another often forgotten heat treatment called “stress relieving” it is not all out normalizing which, in my opinion, is too much for after the anneal. After grinding and shaping and before the quench it may be desirable to have another treatment to relieve stress and prepare for the hardening. Stress relieving consists of heating the steel evenly to around 1200F and holding it there for a little bit before air cooling, it can help with distortion and give a little boost for the new grain formation in the hardening heat.

OK, here is my new and improved protocol. I tried to include elements from Mr. Cashen and Mr. Caffrey that seemed to mesh well, at least in my mind. Please let me know if I'm closer to where I need to be and where the process can still be improved. I've tried to sum up the metallurgy and what is happening at each step--I could have misinterpreted, so please feel free to correct me.

1)Forge to shape with as few high heats as possible, reducing the intensity of the heats as I get closer to shape and especially during edge shaping)

2)Rough grind with the grinding wheel (since I’m not too good with the hammer yet)

3)Normalize (1600 F and air cool)

4)Thermal Cycling 3X (1350 F, 1325, 1300 F)—air cool to 900 F between cycles (I’m thinking that doing the thermal cycling after normalizing is good for grain size reduction—is this accurate?)

5)[Optional: Instead of doing the last two thermal cycles, heat once to 1325 and quench in oil]

6)Annealing (I’ll try both of these and see which works better for me)

A.One Sub-Critical Anneal (1275 F)—cool in hot vermiculite or charcoal dust (to get very fine spheroidal carbide)—this will require a longer soak at hardening

OR

B.One Super-Critical Anneal (1500 F)—cool in hot vermiculite or charcoal dust (to get pearlite—the faster the cooling the finer it is, but slow cooling in vermiculite is the way to go) Question: would this undo all of my grain reduction work from my thermal cycling? Or is getting above critical for a short time not a big deal for grain size?

7)File/Grind to Shape without overheating—lots of dipping in water to cool.

8)[Optional: Stress Relieving Cycle (heat to 1200 and soak a bit, then air cool)—use if I notice I’m getting lots of warped blades at hardening or if I heat the blade too much during step 7.]

9)Hardening

A. Differentially Harden—1480 F (soak if possible, especially if I did a sub-critical anneal) and edge quench in 130 F canola oil. Or butter the back of the blade with refractory cement and give that method a try (haven't done this yet).

OR

B. Fully harden (1480 F) and quench in 130 F canola oil. If I do this I'll need to at least attempt to draw the spine (after oven tempering) with a torch while keeping the edge in oil.

10)Temper—start with one 2-hour cycle at 375 F to start (or at 425 for choppers)…work up to higher temps if necessary.

11)Polish

Another question: Would it be a good idea to keep a bucket of saturated borax solution next to the forge to dip the blade in before each normalizing, thermal cycling, annealing, stress relieving, and heat treatment step? All told, I’m heating the blade 5-7 times just to HT, which causes a lot of decarb. I’ve tried using a reducing flame (which makes temperature control a problem because the forge burns so hot when the blast is so high) and I’ve tried just turning the blast off completely (which gives great temperature control but excessive decarb). Advice?

Thanks for all the advice--I hope I've summed it up correctly.

I should say that my advice is to help you fine tune your methods in your shop. Due to unique variables to each shop, each maker and his tools, there really is no one set recipe to follow for optimal success for all. For instance I hope to advise you on how to optimize what I understand to be your methods. You should not be surprised to find that I do things differently in my shop based upon my separate circumstances. Likewise I would encourage you to play around with this formula a bit since only you can fully fine tune it with first hand access to your shop and blades.

I would opt for the first thermal cycle to be around 1475F-1450F and then proceed to the lower temps.

You can get the necessary soak time down with the spheroidal by merely cycling a few times from heat to air cooling instead of the slow cool. You will also get finer and more evenly distributed carbide by quenching and then low temp cycling. Something like a high heat normalizing to evenly distribute everything and then a 1475F-1500F to lower the grain size in preparation for a quench from 1450F-1475F. This would then be followed by several cycles from 1275F to black heat which would both finish the refinement and spheroidize. Then you could shape and go for the final treatments of optional stress relieving and hardening

I personally would only pursue the edge quench if I were going for a test blade, otherwise I would full quench or use the clay if the customer wanted the hamon for appearances. Maximum strength is obtained by full hardening and tempering. The ideal may be to soak for a while at 1475F and then quench, but if you have limited ability to hold for long times at temp then the 1480F to 1500F would get better solution.

Starting at 375F for tempering is the safest bet, and I would only go for 425F if you find the hardness to be where you want it, even for a chopper. If the soak was insufficient you may find 425F making things a bit softer than you may want.

I handle the decarb issues by doing all the heavy heating with all the forge scale still on and grinding it, along with all the decarb, away afterwards. Be careful not to confuse decarb with oxidation (scale), which can actually be used as a counter to decarb- within reason. The scale that forms on the outside of the steel is iron oxide, in which case you are losing iron and not necessarily carbon. The carbon leaves in the form of a gas and is not as evident. So long as the scales if nice fine flakes and not large ugly sheets, it could probably help you balance out the decarb a bit.

|quoted:

I should say that my advice is to help you fine tune your methods in your shop. Due to unique variables to each shop, each maker and his tools, there really is no one set recipe to follow for optimal success for all. For instance I hope to advise you on how to optimize what I understand to be your methods. You should not be surprised to find that I do things differently in my shop based upon my separate circumstances. Likewise I would encourage you to play around with this formula a bit since only you can fully fine tune it with first hand access to your shop and blades.

I would opt for the first thermal cycle to be around 1475F-1450F and then proceed to the lower temps.

You can get the necessary soak time down with the spheroidal by merely cycling a few times from heat to air cooling instead of the slow cool. You will also get finer and more evenly distributed carbide by quenching and then low temp cycling. Something like a high heat normalizing to evenly distribute everything and then a 1475F-1500F to lower the grain size in preparation for a quench from 1450F-1475F. This would then be followed by several cycles from 1275F to black heat which would both finish the refinement and spheroidize. Then you could shape and go for the final treatments of optional stress relieving and hardening

I personally would only pursue the edge quench if I were going for a test blade, otherwise I would full quench or use the clay if the customer wanted the hamon for appearances. Maximum strength is obtained by full hardening and tempering. The ideal may be to soak for a while at 1475F and then quench, but if you have limited ability to hold for long times at temp then the 1480F to 1500F would get better solution.

Starting at 375F for tempering is the safest bet, and I would only go for 425F if you find the hardness to be where you want it, even for a chopper. If the soak was insufficient you may find 425F making things a bit softer than you may want.

I handle the decarb issues by doing all the heavy heating with all the forge scale still on and grinding it, along with all the decarb, away afterwards. Be careful not to confuse decarb with oxidation (scale), which can actually be used as a counter to decarb- within reason. The scale that forms on the outside of the steel is iron oxide, in which case you are losing iron and not necessarily carbon. The carbon leaves in the form of a gas and is not as evident. So long as the scales if nice fine flakes and not large ugly sheets, it could probably help you balance out the decarb a bit.

Mr. Cashen,

Thank you for all the advice and the options--this is exactly what I was hoping for when I posted my protocol. I understand that using salt baths or an oven can give more consistent results and more options for heat treating. I appreciate your help in working out a protocol that fits my shop and my skill set. One thing is clear: I need to get a better handle on temperature.

You made interesting points about decarb and oxidation. I assumed that for the carbon to come out of the steel it needed to combine with oxygen in the atmosphere--is this not the case? Can carbon leach out even in a reducing fire? (I kind of assumed the carbon might even get pushed in from all of the unburned carbon coming out of the charcoal.) I see how scale can protect from decarb by sealing in the carbon. Does coating a rough polished blade with borax only protect from oxidation and not from decarb? If that is the case, then coating the blade with borax before hardening would actually increase decarb by inhibiting oxidation/scale formation. Hmmmmmm...

If I understand correctly, the following is a potentially good protocol for me, since I will have trouble soaking during the hardening step:

Normalize (1600 F and air cool)

Thermal Cycles:

a.once to 1500-1475 F and air cool to black.

b.once to 1450-1475 F with quench in oil.

c.Several cycles (3?) from 1275 F--air cooling to black.

The several cycles from 1275 F to black are substituting for the sub-critical anneal, correct? And at that point if I found the steel too hard to file to shape I could go ahead and do a sub-critical anneal...and then at the hardening step increase the temp a bit (or increase soak time, if possible.

Again, thank you for all your advice. Next I'll be asking about W2 and 5160! <img src=' http://www.americanbladesmith.com/ipboard/public/style_emoticons//smile.gi f' class='bbc_emoticon' alt=':)' />

Steve, I think we are getting things sorted out. The decarb vs. oxidation is a fascinating study that is a bit more chemically complex than many believe, but it all comes down to what the atmosphere has a greater affinity for- iron or carbon. Working with smelting my own steel I have learned much more about true "reducing" atmospheres which will remove extra oxygen from the equation. The best resource for information on the decarb/oxidation reaction would be the book "Tool Steels Simplified" by Palmer and Luerssen, it has an entire section on furnace atmospheres and includes testing done by Carpenter Steel that showed the effects of heavier oxidation from atmosphere on carbon concentration. More critical perhaps than O2 levels is the presence of moisture which will shift the game to decarburizing even in a heavily carburizing atmosphere. Either way the oxidation just from the room you are in can be more than enough to make the old idea of carburizing in a forge a pipe dream unless great efforts are taken to overcome it, i.e. much more work than just buying good steel <img src=' http://www.americanbladesmith.com/ipboard/public/style_emoticons//wink.gi f' class='bbc_emoticon' alt=';)' />. I would think the whole borax thing (which will result in the equivalent of a glass coating on your blades) would be more hassle than it is worth, certainly more so that simple anti-scale compounds.

|quoted:

This is where the "art" blends with the science for most of us....I think you'll find that everybody's methodology will vary slightly...and that's OK! The key is to work out the methodology that achieves you the "best" end product you can produce (even using the word "best" is open to various interpretations) I'm just going to copy/paste your steps....and insert the differences in mine

1) Forge to shape

2) Rough grind with the grinding wheel (since I'm not that good at forging to shape)

** Grind to shape it fine, but I do not do it

**Thermal cycle: Evenly heat the blade(s) to approx 1300-1350F....and allow to air cool in a rack to 900F or less....X3

3) Normalize--Just one cycle of heating past critical (1600 F) and then air cooling

**I use a single normalizing heat, but strive to be just slightly above the critical temp of the specific steel

4) Anneal

--High: heat to a little past critical (1500 F), cool in warm ash/charcoal dust "hot box"

--Medium: heat to red/orange heat, cool in warm ash/charcoal dust "hot box"

--Low: heat to red/black heat (shadowy), cool in warm ash/charcoal dust "hot box"

** My annealing method is totally different..a single, even heat to slightly above critical temp, and in vermiculite. Annealing is all about SLOW cooling, which means the better insulating value your annealing media hosts, the better...to date I've not found anything that does a better job than vermiculite.

5) File/grind to shape, leaving a bit extra all around for decarb.

6) Harden (one time)--heat to a little past critical (1500 F) and quench in 140 F canola oil (I'll get quenching oil asap)

7) Temper--two 2-hour heats to 425'F in the kitchen oven, cool between tempering cycles.

8) Draw back the spine once (to dark blue) using a torch (or a metal plate on the forge)

**This is an area that I often cringe at....not because it doesn't work, but because most people who ATTEMPT the soft back draw simply do not have the patience or understanding to do it correctly/successfully. If you use "color" then all your likely getting is about a .010" thick "skin" of tempered material, with the remaining "mass" of the treated area remaining harder. It's all about TIME AND TEMP....anything that changes in steel, requires a specific temp, for a specific amount of time. Generally speaking, in order to achieve tempered material all the way through a blades cross section, will require that the heat be applied for approx. 40 mins......now before anyone freaks out and says they have been successful by "color watching"....I will say that IF that is the case, then you did not achieve full hardness on your blade during the quench! Although it not popular in some circles these days, I'm a fan of differential heat treating, and always encourage folks to use that method when creating their JS or MS performance tests blade(s). It allows me to instill attributes to a blade that I simply cannot achieve with full hardening techniques.

9) Polish.

*********************************************************************

As I mentioned, there will likely be a lot of variance on how each of us do things....again....I don't think there is an exact "right" or "wrong" way.....as long as the end result is produces a quality blade.

what is differential heat treating

Hello Phillip. Differential heat treating is heat treating for a hard edge and a less hard spine and a tang that is even less hard. There are three basic methods for doing this, there may be more.

First is an edge quench, where the blade is brought up to quench temp and then only the edge is quenched in the quenchant. This is done by using a limiter plate in your quench tank, so the blade can only go so deep. The blade is rocked so that the point area is also quenched. The spine area is not quenched.

The second method is a full quench, where the whole blade is brought up to the correct temp for whatever steel you are using and the whole blade is submerged in the quenchant. After tempering the blade, you can then use a water bath and put the edge in the bath up to a certain height and use a torch to draw the spine, usually to a blue color. This color will stop approximately 1/4 inch above the water. The tang and ricasso area are also drawn with the torch using the water to stop heat transfer.

The third method would be the japanese method, where the blade is partially coated with a clay or refractory, "satanite". The blade is then brought up to heat and quenched in the quenchant. The whole blade is submerged. The area where there is no clay will get hard, the spine will be softer. This is the method used to get hamons. An o/a torch can also be used to just heat the edge area to the correct temp and then the blade full quenched. Jim Crowell is the master at this method.

These methods are all used to produce a hardened edge and a softer back, ricasso, and tang area.

Brion

|quoted:

Hello Phillip. Differential heat treating is heat treating for a hard edge and a less hard spine and a tang that is even less hard. There are three basic methods for doing this, there may be more.

First is an edge quench, where the blade is brought up to quench temp and then only the edge is quenched in the quenchant. This is done by using a limiter plate in your quench tank, so the blade can only go so deep. The blade is rocked so that the point area is also quenched. The spine area is not quenched.

The second method is a full quench, where the whole blade is brought up to the correct temp for whatever steel you are using and the whole blade is submerged in the quenchant. After tempering the blade, you can then use a water bath and put the edge in the bath up to a certain height and use a torch to draw the spine, usually to a blue color. This color will stop approximately 1/4 inch above the water. The tang and ricasso area are also drawn with the torch using the water to stop heat transfer.

The third method would be the japanese method, where the blade is partially coated with a clay or refractory, "satanite". The blade is then brought up to heat and quenched in the quenchant. The whole blade is submerged. The area where there is no clay will get hard, the spine will be softer. This is the method used to get hamons. An o/a torch can also be used to just heat the edge area to the correct temp and then the blade full quenched. Jim Crowell is the master at this method.

These methods are all used to produce a hardened edge and a softer back, ricasso, and tang area.

Brion

Thanks Brion,

I appreciate all of your comments on all topics.

Bruce