Hardness testing of a sampling of production and custom knives

R

Rapt_up

Gold Member
Joined
May 4, 2012
Messages
962
Lately I am more and more interested in edge holding and cutting performance I have used a digital Rockwell tester that I have access to, to test the hardness of a number of knives that I own. :)

Here are the results. Accuracy for the tester is +/-0.3 Rockwell C

Production:
OLFA LBB blades (Ultra-sharp Box cutter blades) Mfg SPpec:???? Actual: 62.3-63.1 **See post for details this may not be fully accurate**
Kershaw Boa 1580MC (CPM 440V) Mfg Spec: 55-57, Actual: 56.7
Kershaw Leek (CK&T S30V comboedge random tanto) Mfg spec: 57-59, Actual: 57.0
Kershaw Scallion (Walmart 420 HC) Mfg spec:54-56, Actual: 54.5
Kershaw RAM (14C28N) Mfg Spec:58-60, Actual: 58.7
Zero tolerance ZT 0350 (S30V) Mfg Spec: 57-59, Actual 57.7
Leatherman Juice CS4 (420HC) Mfg Spec:???, Actual: 56.8
Benchmade Snody 211 Activator (D2) Mfg Spec:60-62, Actual:60.5
Benchmade 835-07/00 KoTM (M2HS) Mfg Spec:62-64, Actual: 63.2
Benchmade 710-2 limited (M390) Mfg Spec:59-61, Actual: 62.1
Benchmade 710-2 limited (M390) Mfg Spec:59-61, Actual: 61.4
Benchmade 581 Barrage (M390) Mfg Spec:60-62, Actual: 60.1
Benchmade 710-3 (M4) Mfg spec:62-64, Actual: 62.1
Spyderco Mule MT18 (S110V) Mfg Spec:60-61, Actual: 59.7
Benchmade 940-1 (S90V) Mfg Spec: 59-61:, Actual: 59.3
Spyderco Farid K2 (10V) Mfg Spec: , Actual: 64.5
Spyderco Farid K2 (10V) Mfg Spec: , Actual: 65.0

Customs:
JW knives, Jason Woytaz (O1) Target:???, Actual: 58.6
Greg Shahan (AEB-L) Target:58, Actual 57.4
Josiah Gravelle Custom FK-SM (3V) Target:61, Actual 60.9
J Rosa Custom (S90V) Target:61, Actual 61.5
J Garrison Custom (10V) Target: 61, Actual:60.2 (on choil)
Bradford Orwell (M390) Target:61 , Actual: 62.2
Big Chris SAR (3V) Target: 60, Actual: 60.3
Big Chris Bushcrafter S90V Target: 60, Actual: 60.1
OTK DC slicer in K390 Target: 63-64, Actual: 63.3
Darrin Sanders Kiridashi (10V) Target:??? Actual: 60.6
Koster Karda 6 (3V) Target:??? Actual: 58.7


Feel free to add your own tested data.
Here are some other contributions for data points including some other models and brands.
http://www.bladeforums.com/forums/s...ion-and-custom-knives?p=11081071#post11081071
http://www.bladeforums.com/forums/s...ion-and-custom-knives?p=11238015#post11238015


The photo shows my Leatherman Juice blade with two test marks on it, and the tip of a pair of calipers set to .010" to give scale.
I also see that it needs some maintenance/sharpening. ;)

IMG_0587 by rapt_up, on Flickr
 
Last edited:
Something to consider, The thinner the metal gets, the more the heat treat is gonna harden so the spine may not be as hard as the spine.
 
shawnpatterson said:
Something to consider, The thinner the metal gets, the more the heat treat is gonna harden so the spine may not be as hard as the spine.
Click to expand...

I agree, although I am not a metallurgist or material scientist. In addition, some knife makers intentionally differentially heat-treat their knives so that the knife edge is harder for cutting, and the spine is softer for more toughness. So it might be better for the hardness test to be at a point on the knife edge or bevel. But this _can_ be more difficult to test, especially if you cannot _rigidly_ support it due to uneven knife geometry, not to mention the tilt of the knife bevel. Otherwise the knife may flex or bend during the test.

You might ask Ankerson where his knives get tested for hardness. His thread contains many hardness measurements.
http://www.bladeforums.com/forums/s...based-on-Edge-Retention-cutting-5-8-quot-rope

Sincerely,
--Lagrangian

P.S. For those newer to Rockwell C Hardness testing:

"Rockwell Hardness Measurement of Metallic Materials (NIST Recommended Practices Guide)" (2001)
http://www.tensiletest.com/products_services/dwnlds/NIST Booklet.pdf

Nice tutorial introduction to various hardness measurements (Rockwell, Vickers, Brinell, etc):
http://www.gordonengland.co.uk/hardness/
 
I am aware of this, in the case of production knives the blades are hardened in larger quantity batches and will not have differential heat treat, and for the thicknesses we are talking about ther differences will be very small on commercially hardened blades. Also the range of commercial production blades is to account for these variances.

Absolutely for custom makers and singlely treated blades the closer to the edge the hardness test is done the better, and I test accordingly. :) With standard Rockwell C testing you can't get too close or too thin a section without getting false results as the indenter is penetrating into the steel and can't be bearing the back side when doing the test if you want valid results.

Lagrangian said:
I agree, although I am not a metallurgist or material scientist. In addition, some knife makers intentionally differentially heat-treat their knives so that the knife edge is harder for cutting, and the spine is softer for more toughness. So it might be better for the hardness test to be at a point on the knife edge or bevel. But this _can_ be more difficult to test, especially if you cannot _rigidly_ support it due to uneven knife geometry, not to mention the tilt of the knife bevel. Otherwise the knife may flex or bend during the test.

You might ask Ankerson where his knives get tested for hardness. His thread contains many hardness measurements.
http://www.bladeforums.com/forums/s...based-on-Edge-Retention-cutting-5-8-quot-rope

Sincerely,
--Lagrangian

P.S. For those newer to Rockwell C Hardness testing:

"Rockwell Hardness Measurement of Metallic Materials (NIST Recommended Practices Guide)" (2001)
http://www.tensiletest.com/products_services/dwnlds/NIST Booklet.pdf

Nice tutorial introduction to various hardness measurements (Rockwell, Vickers, Brinell, etc):
http://www.gordonengland.co.uk/hardness/
Click to expand...
 
Rapt_up said:
I am aware of this, in the case of production knives the blades are hardened in larger quantity batches and will not have differential heat treat, and for the thicknesses we are talking about ther differences will be very small on commercially hardened blades. Also the range of commercial production blades is to account for these variances.

Absolutely for custom makers and singlely treated blades the closer to the edge the hardness test is done the better, and I test accordingly. :) With standard Rockwell C testing you can't get too close or too thin a section without getting false results as the indenter is penetrating into the steel and can't be bearing the back side when doing the test if you want valid results.
Click to expand...

Hi Rapt_up,

Thanks! Good to know. :)

I'm curious which Rockwell Hardness tester you're using, not that it would mean anything in particular to me. Just curious to see a photo of it.

Sincerely,
--Lagrangian
 
Here are some. I think I made this jpg for a thread. I know I've posted it before.

NOTE: In the comments section for the two stockman knives, you will see the word "annealed". In order to bend the blades into position on a knife with more blades than springs, such as a two spring stockman, the tang of each blade has to be annealed. Otherwise the blade would break in bending it. The bending process being called, "crinking". On a stockman, the only place to measure a blade is on the tang because that is the only place the top and bottom of the steel are parallel to one another. So the only measurement you can get on a two spring stockman is that of the annealed tang, which of course, is quite a bit softer than the rest of the blade. You really don't know how hard the edge is on a stockman.

2nd NOTE: IIRC I have since heard that the Gerber EZ-Outs have 420HC blades.

bladehardnesses.jpg
 
I hope a lot of hardness testing happens, even for production knives which were already tested. This will give us some idea of variation in HRC. However, we should be very careful when interpreting the variation (if any), because the variation could easily be due to different testers (human and machine) rather than strictly from variations in hardness.

I think Ankerson is rather cautious about giving HRC figures that he believes have "false precision", which is why he only gives HRC up to about half a point (0.5). Any additional accuracy and/or decimal points may simply be, in effect, a hallucination.
https://en.wikipedia.org/wiki/False_precision
http://www.bladeforums.com/forums/s...based-on-Edge-Retention-cutting-5-8-quot-rope

For people who are not familiar with the technical details of Rockwell C Hardness testing, it is definitely worth understanding the test conceptually. Basically, a small indenter (or "stylus") is pressed into the sample with a standard force. The machine then measures how much the indenter sinks into the surface. Each point of HRC hardness represents 2 microns of sinking in. So for an HRC of 59, the indenter has sunk into the sample 2 microns more than for a sample with an HRC of 60. (btw, I'm not explaining the bit about the pre-load force and so on. Please see the links below for the full details, including the precise shape of the indenter, and best practices, etc.)
http://www.gordonengland.co.uk/hardness/rockwell.htm
http://www.tensiletest.com/products_services/dwnlds/NIST Booklet.pdf

Half a point of HRC represents one micron of indenter penetration.

Common hand-held micrometers, are accurate up to about 0.0001" to 0.00016" (for imperial micrometers), which is about 2.5 to 4.1 microns. This means a hand-held micrometer is not accurate enough to measure how deep the indenter penetrates.
http://www.amazon.com/DIGITAL-ELECT...-attachment/dp/B004CZ4QB8/ref=pd_sbs_indust_5
http://www.amazon.com/Mitutoyo-103-...342538947&sr=8-4&keywords=mitutoyo+micrometer

Metric micrometers are often slightly more accurate. But even if the _resolution_ of a micrometer goes down to 1 micron, that does not mean it has an accuracy of 1 micron. For example, these Mitutoyo digital micrometer has a resolution of 1 micron, but only an accuracy of +/- 2 microns.
http://www.amazon.com/Mitutoyo-293-...342538947&sr=8-8&keywords=mitutoyo+micrometer

These digital micrometers are a bit better, but their accuracy is still only +/- 1.27 microns.
http://www.amazon.com/Mitutoyo-293-...342538947&sr=8-3&keywords=mitutoyo+micrometer

If we used a micrometer to measure the indenter's penetration depth, none of the micrometers mentioned above would be accurate to half a point of HRC. Of course, an Rockwell Hardness Testing machine can have very special and precise measurements which are more accurate than a hand-held micrometer. The I'm trying to make is simply that: It can be difficult to measure things to an accuracy of less than a micron. To do so would require either an exceptionally accurate micrometer, or else a more sophisticated measurement technique (there are many, which I won't go into).

Measuring things to micron or sub-micron accuracy can be messed up easily by not sufficiently cleaning the surface (oil, dust), thermal expansion from not letting the machine or sample equilibrate to room temperature, human error during calibration or measurement, and just general finickyness. Here's another way to think about it: Modern razor blades are sharp to about 0.4 microns, and the wavelength of visible light is 0.4 to 0.7 microns. So a micron or half a micron is very precise!

So I would like to see a test of repeatability for HRC measurements: Take a given knife (probably sacrificial), and measure its hardness once a day for ten days. Then show a table and/or histogram of the resulting measurements. The results would be interesting, but we might have to be careful about interpreting them. If there is very little variation, then I think that's probably a good sign. If there is significant variation, it's more difficult to interpret: Is the variation because the sample does not have uniform hardness, or is it due to uncertainty in the measurements?
https://en.wikipedia.org/wiki/Accuracy_&_precision

One tester said to try the following deliberate error in making HRC measurements (just for fun and education):
(1) Take a HRC measurement normally.
(2) Put a thin layer of anything (paper, plastic wrap) below the sample (or maybe below the base plate of the sample holder), and repeat the measurement.
He claimed the difference between (1) and (2) would be "enlightening".

btw, please take what I'm saying with some salt. I'm a researcher in computer science; So I'm not an engineer, nor a machinist, nor a material scientist. Even so, in any experimental field, concern for accuracy and precision is universal.

Sincerely,
--Lagrangian

P.S. Just for fun: It is an interesting oddity of Rockwell C Hardness that an _infinitely_ hard material would be measured as having 100 HRC. At 100 HRC, the indenter does not penetrate into the sample at all (ignoring the pre-load, etc.). So if anyone tells you they have a material that is 100 HRC or harder, then they are either just wrong, or they have some crazy exotic meta-material in a research lab.
 
Last edited:
Absolutely... I would personally consider 56.7 the same as 57.3 and typically lump them all as Rc 57... getting much beyond that is as you say false accuracy.

And yes its more about curiousity and seeing where production knives typcally fall. The other issue that arises is taking tests in close proximity to each other. They affect the subsequent results.
 
Rapt_up said:
Absolutely... I would personally consider 56.7 the same as 57.3 and typically lump them all as Rc 57... getting much beyond that is as you say false accuracy.

And yes its more about curiousity and seeing where production knives typcally fall. The other issue that arises is taking tests in close proximity to each other. They affect the subsequent results.
Click to expand...


Hi Rapt_up,

You mean close spatial proximity right? Because of the indentation deforming a region of metal around it, etc. The NIST document has rules about how close indentations can be to each other, and how close to the edge of samples, sample thickness, etc. I thought it was really interesting, but I don't remember the zillions of details (the NIST guidelines are about 100 pages).
http://www.tensiletest.com/products_services/dwnlds/NIST Booklet.pdf

Or do you mean close temporal proximity? If so, I wouldn't have any idea about that.

Sincerely,
--Lagrangian
 
LOL I meant spatial... :) Temporal issues are usually addressed by the machine having a controlled rate of indenting. Yes there are lots of details in the documents, and they are very useful reading for anyone doing hardness testing.

Although in general for a given application, a knifemaker for instance, could read it once, and build a set of "rules" that would cover their application so they didn't have to remember everything. At least thats my take on it.
 
Here's the deal. The machines are accurate. It is the calibration standards that vary. I know. I am an aerospace materials engineer and one day I had a rather thorough conversation on the subject with the metallurgist who runs the metallurgical test lab at work.

If you test two blades on machines calibrated to the same calibration standard, the results are accurate to a decimal place or two and you can compare those knives to one another with that level of precision. I have made multiple measurements on a knife blade. The results were all within about 0.02 units of each other.

If you test blades on machines calibrated to different calibration standards, the results are accurate to ±1. Two different standards are only guaranteed to be that close to one another. Since it would be unusual to have Rockwell testers in two facilities both calibrated to the same standard, you have to compare hardness values of knives tested by different people on that basis: good to ±1 unit.

Now in point of fact, I have on occasion compared my measurements to those of someone else using the same knife. The results we got were closer than 1 to one another. But ±1 is all you can guarantee.
 
Frank,
Thanks for both your data and the additional information re: precision etc... I know that the machine i=I'm using (which is older) is not as accurate as the newer ones, (Its still pretty accurate) and yes the calibration always matters. :)

The data is still useful, just wanting to make sure that other readers, especially ones without the technical background don't read TOO much into it. I suspected you were aero-something engineer from your username and avatar.

James (Mechanical, Electrical, Structural - Professional Engineer)
 
knarfeng said:
Here's the deal. The machines are accurate. It is the calibration standards that vary. I know. I am an aerospace materials engineer and one day I had a rather thorough conversation on the subject with the metallurgist who runs the metallurgical test lab at work.

If you test two blades on machines calibrated to the same calibration standard, the results are accurate to a decimal place or two and you can compare those knives to one another with that level of precision. I have made multiple measurements on a knife blade. The results were all within about 0.02 units of each other.

If you test blades on machines calibrated to different calibration standards, the results are accurate to ±1. Two different standards are only guaranteed to be that close to one another. Since it would be unusual to have Rockwell testers in two facilities both calibrated to the same standard, you have to compare hardness values of knives tested by different people on that basis: good to ±1 unit.

Now in point of fact, I have on occasion compared my measurements to those of someone else using the same knife. The results we got were closer than 1 to one another. But ±1 is all you can guarantee.
Click to expand...

Hi knarfeng,

Many thanks! :)
Very interesting, and important to know, I think. Especially, as you mention, for comparing measurements from different people. As someone without any experience in hardness testing, it is a surprise to me that the hardness calibration is the limiting factor in accuracy. But it's nice to hear that the precision/repeatability is very good.

Let's see... 0.02 points of HRC represents 0.04 microns of indenter penetration. That is 1/10th of an optical wavelength (visible light is 0.4 to 0.7 microns). In other words, that is a measurement which is as accurate as a precision mirror in an astronomical telescope (typically accurate from 1/8th to 1/20th wavelength).

Wow, that is very accurate! I'm curious to know how they measure the indenter's position... is it optical (ie: interferrometry, linear glass scale, or deflection of a laser (like in atomic force microscopy)), or mechanical (something like a micrometer or dial test indicator), or electrical (ie: capacitive linear encoder) ?

I'm also curious about the accuracy/repeatability of other types of hardness measurements, such as Vickers Hardness. If anyone knows or has experience, please chime in.

Sincerely,
--Lagrangian
 
Last edited:
knarfeng said:
I am an aerospace materials engineer and one day I had a rather thorough conversation on the subject with the metallurgist who runs the metallurgical test lab at work.
Click to expand...

Hi knarfeng,

If you have any additional notes or comments about hardness testing (Rockwell C or other), then we would love to hear about it. I found your earlier post outstandingly interesting and useful.

Sincerely,
--Lagrangian
 
Rapt_up said:
I can also test knives if people are interested.
Click to expand...

Hi Rapt_up, knarfeng,

I might be interested in sending a knife in for hardness testing, but probably much later from now.

For now, would you (Rapt_up and/or knarfeng) be willing to test some common blades out of curiosity? I'm thinking you could test the HRC of xacto, scalpel, and/or razor blades. And if it is not too difficult, how about a cobalt-steel or carbide drill bit? (Some drill bits have a "tri" or "hexagonal" shank with three flat regions that might be usable.) Although, tungsten carbide might be too hard for HRC (I mean at the upper end of hardness that can be accurately measured by HRC).

Sincerely,
--Lagrangian
 
You must log in or register to reply here.
Back
Top