Hardness testing and decarb layer questions

Cardinal Knife

Cardinal Knife

Joined
Nov 25, 2007
Messages
340
Hardness testing and decarb layer

Hey guys, seeking advice from those of you fairly knowledgeable with hardness testers. I did these two blades over the weekend, and my RC numbers are all over the board. I ground the decarb off ( at least I thought I did) one blade with a 120 belt (handle area only), and had really erratic numbers, so I took it back to the grinder and hit it with 50 grit, 80, 120, 280 gator in hopes of removing more of the decarb layer, and evening out my numbers. My numbers still had a 12 point spread from 50-62, but were a bit higher on avg than the first round when I used just the 120 belt. With this knife I like to keep the “out of quench” look which is why I didn’t use an anti-scale coating.

HT- 1480 for 10 min soak (Evenheat kiln)
Steel- 1075/1080 from Admiral
Quench- Mcmaster 11 sec heated to 130 deg
Time to quench approx 1 sec
Temper- 400 for 70 min (in wife’s oven)
Testing method- Rockford Rockwell tester verified with a 63 RC test block. (tests within 1 point each time)
Other tests- I sharpened the one blade to approx 20deg (40 inclusive) and just did some manual testing shaving walnut, hammered through corner of walnut, push cutting leather, cardboard and such and had no problems with the edge.

My question is how deep should I be sanding to get past the decarb layer? Am I doing something else wrong; soak time, quench ( should I switch to something faster like canola) , quench heat?

Other variables- the anvil on my tester has play, which always bothers me, but it tests consistently on the test block. Could the knife being tested off center on the handle be causing false readings?

Any help would be appreciated.
Thanks Chris
knife 003.jpg
 

Attachments

  • knife 003.jpg
    knife 003.jpg
    95.8 KB · Views: 24
If those rc readings near the tang edge (spine or perimeter) are higher than those further inward, then probably the tang is mix-phase hardened. In other words, the interior of tang has too much mass for given cooling rate, hence resulted with pearlitic transformation. Since there are 62 readings, I think the blade grind bevel should also be 62rc or higher because of less mass due to thinner cross section. I would go ahead grind decarb off the blade grind and rc test the bevel. Just keep in mind the bevel inclusive angle usually indicate lower hardness reading than actually, could be 1 to 3 points depend on tilt.

At any rate, to assure a full harden 1075/1080 (~0.5%Mn or less) I recommend quench in warm canola or parks50.
 
The anvil can have some play, but there shouldn't be any grit or dings or rocking etc.

You should keep your hand off the work during the hardness test. If it's off balance you can set a weight on the light side to even it out during the test, but if you're having to hold the work you'll never get a good reading.

I don't know how deep your decarb will be.

If you don't soak it long enough, the hardness can be all over the place. I am coming to the opinion that all steel should be given a couple thermal cycles first just to un-spheroidize it first.
 
bluntcut said:
If those rc readings near the tang edge (spine or perimeter) are higher than those further inward, then probably the tang is mix-phase hardened. In other words, the interior of tang has too much mass for given cooling rate, hence resulted with pearlitic transformation. Since there are 62 readings, I think the blade grind bevel should also be 62rc or higher because of less mass due to thinner cross section. I would go ahead grind decarb off the blade grind and rc test the bevel. Just keep in mind the bevel inclusive angle usually indicate lower hardness reading than actually, could be 1 to 3 points depend on tilt.

At any rate, to assure a full harden 1075/1080 (~0.5%Mn or less) I recommend quench in warm canola or parks50.
Click to expand...

Hey thanks for the info bluntcut. I'm gonna have to look up mix-phase hardened, sounds self explanatory, but still, never heard that used before.
 
Nathan the Machinist said:
The anvil can have some play, but there shouldn't be any grit or dings or rocking etc.

You should keep your hand off the work during the hardness test. If it's off balance you can set a weight on the light side to even it out during the test, but if you're having to hold the work you'll never get a good reading.

I don't know how deep your decarb will be.

If you don't soak it long enough, the hardness can be all over the place. I am coming to the opinion that all steel should be given a couple thermal cycles first just to un-spheroidize it first.
Click to expand...

Nathan, the anvil does rock a bit. I try to hold the knife as flat as possible before running the anvil up, but like I said it definitely is a variable. Maybe I will try counterbalancing with a weight instead of trying to hold flat with my hand. Do you mind if I ask your method of performing the thermal cycles?
 
I usually use O1. This was my first time using 1075 what is the recommended soak time for 1/8" stock? I read prior to ht it did not require a long soak, but I was unsure how short was too short.
 
If it is spheroidized a 10 min soak is probably necessary. If it isn't, then it isn't.

I don't think 10 min at those temperatures is going to create decarb more than a couple thou.

If you're touching the workpiece with your hand during a hardness test, your numbers will be all over the place. One Rockwell point is only about .0001". You could skew it ten points. Use a weight to balance it close enough that you don't need to touch it during the test.

My thermocycle will be different than yours because my steels are different, but your steel will respond to a standard descending heat technique for simple steels. If you don't know what your grain condition is or where your carbon is, you should set that before starting your heat treat. I'm using very expensive steels and the grain and carbon condition is better defined and I have found it still needs a thermocycle to avoid occasional problems.
 
Cardinal Knife said:
I usually use O1. This was my first time using 1075 what is the recommended soak time for 1/8" stock? I read prior to ht it did not require a long soak, but I was unsure how short was too short.
Click to expand...
Just for a laugh, bring a coupon up to your quench temperature and just quench it, and then rockwell that and maybe break it to get a look at your grain size. You can differential 1075 with a torch, if you've got a steady hand.
I wouldn't do it with 1095 or O-1 but with 1075 I don't think you'll find a big difference between soaked and non-soaked, assuming proper treatment beforehand.
Steels above 1084 are a whole 'nother ball game.
 
J W Bensinger said:
I wouldn't do it with 1095 or O-1 but with 1075 I don't think you'll find a big difference between soaked and non-soaked, assuming proper treatment beforehand.
Steels above 1084 are a whole 'nother ball game.
Click to expand...
Can you please elaborate what you mean by soaked or non soaked. Wouldn't a one minute soak still be a soak? Man, "soak" turn into a weird word when you type it that much.
 
Cardinal Knife said:
Can you please elaborate what you mean by soaked or non soaked. Wouldn't a one minute soak still be a soak? Man, "soak" turn into a weird word when you type it that much.
Click to expand...
That's true. In the example above I actually meant bring it to quench temp long enough for it to be an even temp all the way through-which with 1/8" isn't going to be long at all (which I guess is technically a soak).
1075 is hypoeutectoid, so if your thermal cycles haven't been wacky it should be fine to bring to temp and quench.
I used to use a lot of 1075 and the process was forge, normalize, grind to a 3/32" (ish) edge, normalize, and differential harden with oxyacetylene, and I never had one come back for any reason.
Back to the original question, Nathan has you on the right track with the rockwell tester. I don't think you'll actually hurt the stuff soaking it.
 
Finished those two blades, thought I'd throw you guys a pic
7582e5782dee68d8c430b5bbb5333952.jpg
 
You must log in or register to reply here.
Back
Top