Heat treating with a gas forge

[wmhammond]

I need some help. I recently purchased the gas forge pictured below and I have used it successfully to heat treat several smaller knives. A week ago I attempted to heat treat a larger knife. The steel was 1/4" W- and was 13 3/4" long. I ground the blade with a flat grind using 60 grit ceramic blaze by Norton and left an edge thickness (thicker than usual) at 2.5 mm - about the thickness of a Nickle. The blade was coated totally covered with scale prevention coating and then clayed up as pictured below. I use a thermo couple to determine temp inside the forge. I placed the knife as far to the right as I could to keep it out of the gas flame directly and the thermo couple sensor to the right of the blade. I then lit the forge and using the gas flow control I brought the forge slowly up to 1465* which was my target temp. At the time I did this the temperature in my shop was 59* so I assume that my Parks 50 quench Oil was also 59*. When I reached the target temp I soaked the knife for 15 minutes regulating the temp inside the forge between 1470* and 1460*. Pulled it out and immediately and quenched. Allowed it to cool to the point that I could carry it and immediately put it in a 400* oven for 2 hours. Let it cool overnight and then began Grinding it to its final dimensions. Never thought about a crack but low and behold after about three passes on the grinder there it was about 1/2" long with a "J" hook at the top (Pictured below).

OK, so I thought that over and decided that my quench oil was too cold. So, I cut out another knife, same dimensions but this time with 1084 instead of W-2. Same exact procedure on this one as described above except I heated my Parks 50 to 85*. This morning I took it out of the cool oven and cleaned it up with a Brillo pad and there it was, another crack on this new 1084 blank (pictured below). What am I doing wrong? Thanks guys,

Wallace

 

[Frank St. Clair]

I need some help. I recently purchased the gas forge pictured below and I have used it successfully to heat treat several smaller knives. A week ago I attempted to heat treat a larger knife. The steel was 1/4" W- and was 13 3/4" long. I ground the blade with a flat grind using 60 grit ceramic blaze by Norton and left an edge thickness (thicker than usual) at 2.5 mm - about the thickness of a Nickle. The blade was coated totally covered with scale prevention coating and then clayed up as pictured below. I use a thermo couple to determine temp inside the forge. I placed the knife as far to the right as I could to keep it out of the gas flame directly and the thermo couple sensor to the right of the blade. I then lit the forge and using the gas flow control I brought the forge slowly up to 1465* which was my target temp. At the time I did this the temperature in my shop was 59* so I assume that my Parks 50 quench Oil was also 59*. When I reached the target temp I soaked the knife for 15 minutes regulating the temp inside the forge between 1470* and 1460*. Pulled it out and immediately and quenched. Allowed it to cool to the point that I could carry it and immediately put it in a 400* oven for 2 hours. Let it cool overnight and then began Grinding it to its final dimensions. Never thought about a crack but low and behold after about three passes on the grinder there it was about 1/2" long with a "J" hook at the top (Pictured below).

OK, so I thought that over and decided that my quench oil was too cold. So, I cut out another knife, same dimensions but this time with 1084 instead of W-2. Same exact procedure on this one as described above except I heated my Parks 50 to 85*. This morning I took it out of the cool oven and cleaned it up with a Brillo pad and there it was, another crack on this new 1084 blank (pictured below). What am I doing wrong? Thanks guys,

Wallace

View attachment 60285View attachment 60286View attachment 60287View attachment 60288

Sent from my SM-G935V using Tapatalk

 

[Frank St. Clair]

Sorry. Big thumbs on cell phone

Sent from my SM-G935V using Tapatalk

 

[Ty Adams]

I don't have anyway of reading the temp of my forge. But the way you have the thermocouple sticking out the front you are probably getting mixed readings. I have tested heat treating in my forge quit a bit. Several 48" bars. The only crack I have seen is from when I just let the steel sit. Even with the knife on the off side of the forge it will be hotter next to the burner. I now turn the burner as low as it will go and move the steel in and out slowly to keep the heat even on the knife. It takes over 5 minutes to get to non magnetic. Than a minute or so more. When the steel gets brighter not a different color I quench. I don't know if that will help you but it is worth a try.

 

[J. Doyle]

Leaving a 60 grit finish on the blades would be my number one suspect, especially in these cases given the location of the cracks. I'd go to at least 220 personally, even 120 grit would probably be better.

Overheating would be my second guess. Are you using some sort of muffle or baffle pipe in the forge to protect the blades from direct flame and atmosphere in your forge? How are you regulating the heat in your forge and how are you monitoring it?

I wouldn't expect that your oil being too cool was the culprit, especially with parks 50 as that would make it slower and thus, likely less violent. As a side note, I'd use parks 50 between 70 and 90 degrees F.

I'd also be careful about soaking W2 especially for 15 minutes in a forge. That seems excessive to me for that steel and you run the risk of overheating, especially if you don't have a good way to monitor and regulate accurate temps and aren't using a muffle.

 

[J. Doyle]

Sorry Wallace, I just looked again at your first pic and see that you are using a thermocouple but I don't see any muffle pipe or baffle and the way those burners blow flame directly down at the blade could cause some hot spots, especially with 15 minutes of soaking and no protection from the atmosphere in the forge.

If you overheat a little bit AND have a coarse 60 grit finish, you definitely could be inviting cracking issues.

 

[EdCaffreyMS]

I see a combination of issues.....
1. LAYING the thermocouple on anything will effect the reading....it's likely going to read quiet a bit lower then the actual temp (which means in this case the temp of the blade was likely much higher then what your readout was showing.)

2. I'm not a fan at all of soaking blades..... it has gotten way out of hand over the past few years.... while there are some steels that require a soak, many have taken the attitude that "more is better" and allow blades to "soak" far too long...... it's a matter of seconds, and not more then a couple of minutes..... for every second that a given steel is "soaked" at or above it's critical temp, grain growth occurs exponentially, meaning that it's likely at that long of a soak, the grain in that blade was like coarse sand.

3. The finish prior to heat treat. In my experience, anything coarser then 120 grit finish creates a very likely case of stress cracks....basically the expansion, then rapid contraction of the steel when quenched will cause a crack at one or more of those 60 grit scratch lines.

I also suspect that warming the Parks50 had an impact too. When I first started using Parks50, I preheated it, just as I had always done with previous quenchants....more muscle memory then anything, and the first couple of blades I quenched (52100) where just littered with cracks. I learned to never preheat Parks50. (my shop stays at 55F)

Finally, as John mentioned, when you put together the combination of 60 grit scratches, and overheating, particularly for that long of a duration, then add in the preheat of the Parks 50, you'd be lucky if the blade(s) didn't crack.

 

[wmhammond]

As always, you guys are the best. Going to have another go at it this weekend and I'll get back to you. Here are my take-a-ways:

1. Get a baffle
2. Grind to at least 120
3. Don't heat the parks 50
4. Don't let the thermocouple sensor touch anything
5. Don't soak - take it to 1465* and quench it.

Thank you so much,

Wallace

 

[Kevin R. Cashen]

As always, you guys are the best. Going to have another go at it this weekend and I'll get back to you. Here are my take-a-ways:

1. Get a baffle
2. Grind to at least 120
3. Don't heat the parks 50
4. Don't let the thermocouple sensor touch anything
5. Don't soak - take it to 1465* and quench it.

Thank you so much,

Wallace

Well, just when you think you have it worked out along comes more input. My first instinct is what you have been advised about the 60X finish. I personally would never heat treat with a finish as coarse as 60X, and if I did I would not be surprised to find fractures induced by the stress risers, especially in Parks #50 which will allow you to actually achieve maximum hardness and those levels of stress/strain.

That being said there could also be some issues with the heating. The cooler quenchant would not be a problem as it would result in a lower overall hardness rather than stress risers. If there was resulting warpage and other distortion along with the fracturing, indicating uneven cooling mechanisms then maybe.

Now on the subject of soaking… sorry Ed but you know I have to go there, and I think you would expect no less. Ed and I profoundly disagree on some areas of the soaking topic, but I have often used our friendship to point out how people can have difference of opinions on knifemaking and still like each other if we remember knives are knives and people are people. Soaking is very critical for proper heat treating in some alloys and less critical with others. When tight temperature controls (that caveat is important) are used I have never in almost 40 years of heat treating experienced grain growth of any kind from proper soaking, even outrageously long soaks, provided temperatures remains below the grain coarsening temperature at which all pro-eutectoid phases are dissolved. If anybody assembles a lab approaching my facilities and puts the time studying grain conditions with methods matching ASTM E112 type standards for properly measuring grain size and has found the opposite of all my observations I think that could be exciting! But until then I have to go with what all the literature and research I have read that happens to coincide exactly with my own observations from years in front of a forge and a metallograph.

That being said, just because I have not found the same threat to grain size does not mean there are not many detrimental effects from over-soaking, and I would much rather dwell on the points that I agree entirely with Ed about, so that others can benefit from our combined coinciding experiences.

I agree with Ed that there is a lot of oversoaking going on in simpler steels, and if it helps I can say that I rarely exceed a ten minute soak on any steel I work with and have found if it does exceed 10 minutes to obtain proper solution then I probably did something wrong during normalizing or annealing (please note that I do not work with stainless). One thing oversoaking will give you is retained austenite, at which point soaking is robbing your blade of hardness rather than adding to it. Soaking increases decarb and scaling, oversoaking even more so. And, going back to that caveat and the need to do soaking properly, soaking in a forge or any other less controlled heating methods can be a recipe for disaster.

The good news to the soaking confusion is that with W series and 10XX series you can minimize or even eliminate the need for a soak. Avoid going into the hardening operation in a heavily annealed condition, but instead harden from a finer normalized structure and there is no need to keep the steel at those hardening temperatures for any troublesome times. From fine pearlite W2 or 1084 should be in solution almost instantly, and with the W2 the finer primary carbides from normalizing will make a much keener and stable edge- it is a win/win without the hassle of the soak and Ed and I can both say “see, we told you so!”:3:

 

[scott.livesey]

I noticed you put the blade in a cold forge then lit the fires rather than heating to target temperature then adding the blade. could this be part of why the blade cracked?

 

[EdCaffreyMS]

Kevin is right...... there are different paths to same destination, and he and I are a prime example. We agree to disagree on some things, and remain friends. I'd be dumb not to defer to Kevin when it comes to metallurgy..... his knowledge of the subject far surpasses mine. It's as I often mention.....it's not about "the right way", "wrong way", nor "the only way" of doing something, it's about achieving the desired outcome, and however that applies to each individual's skills and shop.

One of the reasons I tend to become more and more "anti-soaking" is because of it being overdone so much these days. Not long ago I conducted a performance test where the blade failed....and the broken edges had grain like coarse sand....the individual stated he had soaked it for "about 15 mins"..... the grain size clearly told me the temp was too high, and I have no doubts that large grain size, likely from the extended soak, at too high a temp, is what caused the failure. As with many things we do in Bladesmithing, applied correctly, it can be a good thing, but when improperly, or overly applied, it becomes a detriment.

 

[wmhammond]

Thanks again guys. heat treated this morning - no visible cracks and it totally skated my file. It's in the oven now at 395* for two hours. I'll get it out and clean it up and give you the final report wit pics and all the details. Thanks

Wallace

 

[wmhammond]

SUCCESS!!!!!!! So here is what I did. I bought a 12" X 3" steel pipe (Baffle). I ground the blade to 120 grit. I found a ceramic "condom" to fit over the end of my thermocouple. I did not heat my Parks. I put the blade inside the baffle pipe with the protected thermocouple and slowly heated the forge slowly to 1470 added a piece of mild steel to act as a heat sink, which it di and dropped the temp to 1460, took out the heat sink and allowed the temp to raise to 1470 and did that twice more so I guess I sort of soaked it at about 1465 for 4-5 minutes. Took it out and quenched it. Tempered it for 2 hours at 395*. Cooled it down in the oven. Cleaned it up with steel wool - NO CRACKS!!!!! I can't tell you guys how much I appreciate your help. What a great site. I'm a nobody and I get the help of three of the best knifemakers in the business and several more just great guys and they solve my problem. Thank you so much. Life is good. I put some pics below. The last one is sort of a Happy New Year card - a view out of the door of my garage (shop) - snow in the mountains.

Wallace

 

[Kevin R. Cashen]

SUCCESS!!!!!!! So here is what I did. I bought a 12" X 3" steel pipe (Baffle). I ground the blade to 120 grit. I found a ceramic "condom" to fit over the end of my thermocouple. I did not heat my Parks. I put the blade inside the baffle pipe with the protected thermocouple and slowly heated the forge slowly to 1470 added a piece of mild steel to act as a heat sink, which it di and dropped the temp to 1460, took out the heat sink and allowed the temp to raise to 1470 and did that twice more so I guess I sort of soaked it at about 1465 for 4-5 minutes. Took it out and quenched it. Tempered it for 2 hours at 395*. Cooled it down in the oven. Cleaned it up with steel wool - NO CRACKS!!!!! I can't tell you guys how much I appreciate your help. What a great site. I'm a nobody and I get the help of three of the best knifemakers in the business and several more just great guys and they solve my problem. Thank you so much. Life is good. I put some pics below. The last one is sort of a Happy New Year card - a view out of the door of my garage (shop) - snow in the mountains.

Wallace

View attachment 60314View attachment 60315View attachment 60316View attachment 60317

I am glad it worked out. I know how frustrating it can be to just run into a wall like this and not know how to overcome the problem. In such cases a community where somebody else may have overcame the same issue can be invaluable. Success is great but when we share our failures, and the methods of overcoming them, we can help countless others in not having them, this allows the craft as a whole to advance onward without having to reinvent the wheel.

 

[OzarkEdge]

Leaving a 60 grit finish on the blades would be my number one suspect, especially in these cases given the location of the cracks. I'd go to at least 220 personally, even 120 grit would probably be better.

Overheating would be my second guess. Are you using some sort of muffle or baffle pipe in the forge to protect the blades from direct flame and atmosphere in your forge? How are you regulating the heat in your forge and how are you monitoring it?

I wouldn't expect that your oil being too cool was the culprit, especially with parks 50 as that would make it slower and thus, likely less violent. As a side note, I'd use parks 50 between 70 and 90 degrees F.

I'd also be careful about soaking W2 especially for 15 minutes in a forge. That seems excessive to me for that steel and you run the risk of overheating, especially if you don't have a good way to monitor and regulate accurate temps and aren't using a muffle.

Exactly my thoughts. I keep it simple. Move my knives back and forth until I reach critical and then quench on carbon steels. I also get my knife close to finish before quench, knock off the scale and then begin cleaning it up for etching or whatever the final finish. I haven't had a crack since I've been using this method

Sent from my XT1585 using Tapatalk

 

[tapforge]

and the first couple of blades I quenched (52100) where just littered with cracks. I learned to never preheat Parks50. (my shop stays at 55F)

WOW WOW WOW, 52100 in Parks50?!? i thought that was a medium quench steel and needed 10 second oil, iirc i was told never to quench 52100 5610 O1 in parks because it will will be too fast and over stress the steel resulting in a ton of cracks. please correct me if im wrong. because i dont want to have two quench tanks if i can get away with one.

 

[Kevin R. Cashen]

WOW WOW WOW, 52100 in Parks50?!? i thought that was a medium quench steel and needed 10 second oil, iirc i was told never to quench 52100 5610 O1 in parks because it will will be too fast and over stress the steel resulting in a ton of cracks. please correct me if im wrong. because i dont want to have two quench tanks if i can get away with one.

Tapforge, Parks #50 was indeed designed for shallow hardening steels (1095, 1084, 1075, W-2, W1 etc...) that were originally quenched in water. Parks #50 is 7 to 10 second oil and is described as having speeds approaching water, and I can attest to the fact that it is indeed capable of such speeds. I personally would strongly advise against quenching any steel in Parks #50 that you would not want to quench in water. Alloyed steels with deeper hardening qualities, which were made for oil quenching, are better suited for 10 to 14 second oils, and thus Parks also developed their AAA oil for such steels, but there are many more medium speed oils on the market.


I have cracked many O-1, 52100 and 5160 metallurgical samples in Parks #50, but I did it intentionally in order to insure full conversion without having to mess around with water. Chrome is an incredibly powerful alloy element for increasing hardenability and it doesn’t take much to push a steel into the medium speed oil category. I personally would stick with your two quench containers. I have both.

 

[tapforge]

Tapforge, Parks #50 was indeed designed for shallow hardening steels (1095, 1084, 1075, W-2, W1 etc...) that were originally quenched in water. Parks #50 is 7 to 10 second oil and is described as having speeds approaching water, and I can attest to the fact that it is indeed capable of such speeds. I personally would strongly advise against quenching any steel in Parks #50 that you would not want to quench in water. Alloyed steels with deeper hardening qualities, which were made for oil quenching, are better suited for 10 to 14 second oils, and thus Parks also developed their AAA oil for such steels, but there are many more medium speed oils on the market.


I have cracked many O-1, 52100 and 5160 metallurgical samples in Parks #50, but I did it intentionally in order to insure full conversion without having to mess around with water. Chrome is an incredibly powerful alloy element for increasing hardenability and it doesn’t take much to push a steel into the medium speed oil category. I personally would stick without your two quench containers. I have both.

Thanks thats what i thought. ill keep two quench tanks for now...

 

[EdCaffreyMS]

Guess I don't know any better I quench 52100 and 5160 in Parks 50.... I don't use any O1, so can't speak to that...... I just learned NOT to preheat it, leave my edges a bit thicker prior to quenching, and get blades in the tempering oven ASAP after the quench. I've not had any problems...with the exception of that first time when I preheated it, and that was more muscle memory taking charge... then my brain kicking in.

There's a lot of "hearsay" out there about many things, and as I always say, don't take anybody's word for any of it (including mine) until you prove or disprove it for yourself. I think there's a lot of headway that can be (and has been) made in Bladesmithing when people "think outside the box". Sometimes it works, and sometimes it doesn't, but nothing ventured...nothing gained.

 

[Kevin R. Cashen]

I wanted to mention that Parks #50 was designed specifically for shallow hardening non-alloy steels earlier in this thread, but I bit my tongue because Ed is experienced and needs to do whatever works for him in his own shop, it’s not up to me to meddle in that. But Tapforge specifically asked about what steels are actually suited for #50 and I felt it wouldn’t be right to keep silent and let him come away without knowing what Heatbath/Park Metallurgical recommends. We all do indeed have different experiences in our own shops; that is undeniable. I have probably been working with Parks #50 longer than any other knifemaker that I know, with the exception of Dan Maragni who introduced me to it around 20 years ago. I only mention this because Ed is right that there are a lot of green armchair experts whose experience seems to reside in Google searches or other reading, rather than doing, and what I have experienced in my shop happens to correspond very well with what Heatbath/Park Metallurgical lays out in their literature.

I will share this pdf. from my quenchant library for anybody wanting Heatbath/Park Metallurgical literature on the #50:

http://www.cashenblades.com/forums/50.pdf

In it Park describes this oil as- “ 50 QUENCH OIL produces high hardnesses in many types of steel ordinarily considered “water quenching”. In the initial stages of the quench, 50 QUENCH OIL cools like water. As the cooling nears the martensite formation range, this remarkable oil cools the steel slowly and uniformly to preclude cracking or distortion.

The recommended operating temperature range is ambient to 120F.”


I didn’t want to contradict anybody, I actually never want to do that, but so often were are forced to choose between politeness and what we know to be the facts, I am just wired in a way that will not allow me to compromise factual information. If this offends I do apologize, but I have to stick with my principles or I wouldn’t be me, or anybody else for that matter.