Recommendation? Heat Treat Scale "pop"

[N.W. Gean]

Good afternoon, everyone. I have a question about heat treating that does not seem to have many discussions. Does every oil quenching steel have scale that "pops off" from a successful quench?

I have quenched a couple knives of 8670 so far, and the scale is solid black with little to no shedding. Even after a 350 degree temper, it is extremely difficult for a file to dig in.

I have one of those cheesy Rockwell C files marked "60" and it can't do anything to either of the knives after grinding the scale off. The edge (~20 degrees with bte of ~.005) on the 1st one seems like it rolls a little after chopping a nail (haven't ground bevels on #2 yet). I broke the tip off and the grain structure seems a little sandy but not egregious, and that was after 3 quenches trying to get it right.

My process right now is:
Heat blade to non magnetic
Turn gas on forge down
Hold at roughly same color shade for 2.5 minutes (checking at intervals for magnetic to make sure my eyes aren't tricking me)
Tempered at 250 (edge chipping on brass rod and nail test), 300 (still a little chippy), then 350 (now rolling a little on both tests)

I arrived at soaking for 2.5 minutes after 1.5 and 2 did not seem to harden the knife. The edge is considerably tougher after the third quench but still rolls on hard use and if I push hard on the brass rod (springs back if I stay at sharpening angle).

But then there's that scale issue. Why isn't it acting like 1075 and 1084, unless it either didn't harden correctly or 8670 is a different animal. I'm leaning towards the former.

Thanks for reading that wall of text.

 

[Joshua Fisher]

I’ve been working with a lot of 8670 recently and would say in a forge you really don’t need more than a minute soaking if that, if it didn’t harden chances are you may not be hot enough, on a oven you want to set it for 1550, ideally it should be about 1525 when it goes in the quench for a forge ht you really have to get good at watching color and knowing what it should look like in your shop. I bring it up to non magnetic and then go up about 2 shades, I’ve been getting blades around 61-62 at a 350 temper and 60-61 at 375. I temper camp knives and hard use blades at 375 and can chop through a 1/4” brass rod with no edge deformation or dulling, I typically hollow grind most of my blades and leave the camp knives with about .03” behind the edge then convex to sharp on the grinder before touching up with stones. In my opinion tbe depends on your grind type, I’d take a flat grind thinner or turn it into a convex grind for hard use, hollow grinds can be left a bit thicker because of their nature, if I took it down to .005 that would be like a super thin chef knife or almost a razor. But as for your original question the scale does seem different, I’ve noticed similar scale with 15n20 and I wonder if it’s just the added nickel. It wipes right off before tempering. Also if your grain isn’t as fine as it could be that will effect abusive tests, try a few sample blades all quenched once and tempered to different degrees then repeat the tests.

 

[N.W. Gean]

I have no doubt the grain structure is mediocre. I'm fairly certain I overheated it on quench 2. The scale on mine doesn't wipe off. I'll play around with another blank before work tomorrow and see what I get. The problem for me is that I see orange before it's even non magnetic so my tendency is to pull too early. You'd think 2.5 minutes would get it hot enough but I know from the thermocouple I had in it that the forge can drop below 1500 quickly with low gas pressure.

 

[Joshua Fisher]

I have no doubt the grain structure is mediocre. I'm fairly certain I overheated it on quench 2. The scale on mine doesn't wipe off. I'll play around with another blank before work tomorrow and see what I get. The problem for me is that I see orange before it's even non magnetic so my tendency is to pull too early. You'd think 2.5 minutes would get it hot enough but I know from the thermocouple I had in it that the forge can drop below 1500 quickly with low gas pressure.

Everyone’s perception of colors is different and color can look different depending on lighting, I always turn off the lights and lower the roll down door to the shop so it blocks natural light where the forge is. Use a magnet to give your self a bench mark if it looks dull orange when it becomes non magnetic then go a couple shades higher to like a bright orange then try quenching there, if it didn’t harden then go up another shade of it did harden snap the piece and check the grain if it looks coarse then go down a shade on the quench once you find that balance of fine grain with a hardened blade remember the color and shoot for that each time. An oven definitely takes some of the guess work out but that doesn’t mean you can’t get really good results in a forge

 

[N.W. Gean]

An oven definitely takes some of the guess work out but that doesn’t mean you can’t get really good results in a forge

That's what I've been thinking. I've seen tons of people here and other places nail their HT in a forge. No reason I can't do it, just need practice.

 

[Blackdirt cowboy]

Don’t feel bad, N.W. My first attempts at heat treating resulted in broken blades after tempering. I’m still struggling with normalization to reduce the grain. I’m just not entirely confident I’m following the right process. I’ve had no issues hardening a blade, just in grain reduction.

 

[JTknives]

A big problem with using a forge is your eyes. If your trying to soak for any length of time you most likely will over heat it. Your eyes adjust more and more the longer you stare at the forge. You could be 200° over temp and you would not even know. This is why setting up a pid controlled forge is a must if your going to try and heat treat in one. Not saying you can’t do it by eye but then it’s always a guess and you can not make any guarantees about your heat treat. If your just playing around then no problem but if your making knives to sell then that’s where your doing your customer a disservice. Even if your pid is not controlling anything it will give you a color reference point. So adjust your forge till the pid says the temp you want then soak the blade and keep it the same color as the thermocouple and that will get you close. 25°-50° Off can make a huge difference in the outcome of a heat treat.

 

[N.W. Gean]

The problem I've had with the thermocouple is it'll read 1500 and still stick to the magnet. That's why I took it out.

JT, I've thought at great length about sending my knives to you for HT, but there's a part of me that thinks I can still figure it out myself you might still get a chopper from me though. No way I could do a 16" blade in that forge.

 

[N.W. Gean]

Well this one definitely hardened. Ran the forge a little hotter, and went about a minute past non magnetic. I'm concerned about the pattern though. It looks like blistering from overheating
Regardless, I'll put an edge on after temper and see what happens.


Update: Fail. Tempered at 325F, edge at .005 to .007, rolling on the brass rod.
Even if that edge is too optimistically thin, it should be chipping badly at the hardness a 325 temper would give, correct? I'm thinking of building an HT oven for...unrelated reasons

 

[Joshua Fisher]

Don’t worry about time with a forge ht it will give you poor results, putting any bias on whats best aside and just what will give you results with what you have, I’ve not had good luck with thermocouples in a forge, you have to place them in a very specific spot and it has to be rated for over 2600f otherwise you’ll burn it up and it won’t give good read outs. Don’t look directly into the forge as that will damage your eyes and cause them to adjust to the color of the forge, you should ht in low to no light then see what color the steel becomes non magnetic remember that color then go 2 shades brighter and test there, you can go up or down a shade until you find what color is right. From non magnetic to the right temp to quench could be 10-15 seconds depending on your forge. 8670 really doesn’t need a soak so sticking to it is a good choice for a forge ht. It also has a wide temp range that will give you good results. It just takes a couple test blades to get a good idea of color and what’s right. You may not be able to say an exact temp but if you test your blades and know what they are capable of then you know how your ht was. A oven just takes some guess work out but you should still destruction test and do abusive tests on blades from time to time to ensure that you are getting a optimal heat treat.

 

[Bigfattyt]

A big problem with using a forge is your eyes. If your trying to soak for any length of time you most likely will over heat it. Your eyes adjust more and more the longer you stare at the forge. You could be 200° over temp and you would not even know. This is why setting up a pid controlled forge is a must if your going to try and heat treat in one. Not saying you can’t do it by eye but then it’s always a guess and you can not make any guarantees about your heat treat. If your just playing around then no problem but if your making knives to sell then that’s where your doing your customer a disservice. Even if your pid is not controlling anything it will give you a color reference point. So adjust your forge till the pid says the temp you want then soak the blade and keep it the same color as the thermocouple and that will get you close. 25°-50° Off can make a huge difference in the outcome of a heat treat.

I have a new Graham Atlas blown forge. I'll be damnes if I have figured out how to get it lower to heat treat temps!!

I have a pid that maxes our at 2000° and my forge seems to be running above that!

Not sure how to get the temp down and still not blow out/flash out.

I've tried turning the blower and propane down...

I HT'd a puko by moving it in and out, with constsnt magnet checks. May try forging a piece of 1080 this weekend.

 

[Joshua Fisher]

I have a new Graham Atlas blown forge. I'll be damnes if I have figured out how to get it lower to heat treat temps!!

I have a pid that maxes our at 2000° and my forge seems to be running above that!

Not sure how to get the temp down and still not blow out/flash out.

I've tried turning the blower and propane down...

I HT'd a puko by moving it in and out, with constsnt magnet checks. May try forging a piece of 1080 this weekend.

Put a muffle pipe like a 2x2 square tube into the forge it can help distribute heat and avoids direct flame on your knife, if you’ve ever seen a pid controlled forge they shut off and on quite a bit to get low temps, some steels don’t need to soak at temp and some do, you can ht 1080,1084,8670,5160 with minimal to no soak time.

 

[Joshua Fisher]

Also a neat trick to heat treat by color is put some salt on test coupon and watch for it to melt then pull the piece out and see what the color of the coupon is. Salt melts at about 1475F so that gives you some idea of what that color looks like in your shop. Red in my shop could be orange in yours but salt always melts at 1475.