Towards 0.1 Micron Archive - The Rosetta Stone

304 Stainless mule cutting edge (6 replies)

cKc — Thu, 18 Jun 2015 01:31:59 +0000

This is in no way serious testing or anything. I just couldn't let it go off to Cliff without at least trying one thing. :)

its interesting that the attempt to roll the edge, actually stabilized it. its so soft you can see the strop rounded the bevel.

[www.youtube.com]

Mule - Carpet (55 replies)

CliffStamp — Tue, 10 Nov 2015 03:03:05 +0000

Experiment proposal :

-measure sharpness on bergia/35 grams and/or 1/8" jute/1000 grams
-cut carpet on a slice, measure sharpness periodically, calculate TCE

This didn't work for a few reasons :

-bergia was too fine a measurement, almost immediately the sharpness was so low it would not make a cut with a full draw
-1/8" jute/1000 grams was workable and allowed an extremely low sharpness measurement

More data is needed on the jute to compare to the bergia, but as a coarse estimate it looked to be able to measure below 0.5% of optimal, the apex is clearly visible at this point, the knife is really dull but it can still cut carpet.

However there were a few issues with how the knives behaved. I started with two extreme blades :

-AEB-L/Kyle Bettleyon
-Maxamet-64/McCullen

This was the AEB-L after 150 slices :

The apex didn't blunt, the actual edge itself deformed, clearly visible by eye and felt, it just folded over. However here was the Maxamet :

This was the worst part. That large white glint is an impaction/roll. The Maxamet looks to show far greater strength and general durability and effective edge retention as it loses far more steel.

--

At this point you might be tempted to explain this by some kind of hardness/carbide explanation, however the first thing is that this is only one run. But the difference is so large that I didn't even think random variation could explain it. I did a more careful measurement of the edge angles as they approach the apex and they are :

-AEB-L : 5 to 5.5 dps

-Maxamet/64 : 8 to 8.5 dps

Now this might not look like a big difference, but stiffness is NOT linear. If you look at the percentage difference and cube it then it shows an almost 4:1 advantage in stiffness to the Maxamet. I was obviously not comparing steels with such a large change in cross section I could have been seeing nothing more than the cross sectional difference. However this difference in thickness is actually a very small change measurement wise. Consider :

-0.008"/0.050" -> 10 dps
-0.006"/0.055" -> 6 dps

However those are really tiny change and you can make them easily through parallax error (the angle that you look at the calipers in relation to their position on the knife). The point of this is that it isn't at all trivial to make sure the angles on the knives are the same, not just in the apex but in the bevels going up to it including the main grind. Even if you measure them, small changes of the order of 0.001" can easily produce large changes in the strength/stiffness of the apex/edge.

I reset the AEB-L to 9 to 9.5 dps to normalize out the angle differences. In order to ensure I was getting accurate/precise angles I measured the apex in multiple spots and then checked as well on the Sharpmaker to ensure the apex would catch at the calculated angles. With the similar edge angle the knife easily cut the carpet and just dulled in the apex, no damage to the edge itself :

--

The mules are all zero ground so most of them are ~6 dps in the final edge. For the carpet cutting I am going to increase this a little to 8-9 dps and the use an x-coarse DMT as the apex finish. I am going to look at both edge retention and ease/speed of regrinding with at least two stones. One of the stones will be designed for hard to cut steels, the other one will not.

Anyway, the point of this post is to make a point that it maybe isn't as easy to get a comparison focused on steels as you might think and that it is very easy to be measuring something very different than what you intend even if you are doing things in what appears to be a sensible manner.

Mule - Cardboard (10 replies)

CliffStamp — Fri, 21 Aug 2015 22:12:09 +0000

Details :

-1/8" cardboard cut on a slice through 2" of blade, approximately 1' cuts made at ~1 ft/s
-sharpness measured on bergia spinning thread, 35 grams load
-TCE calculated when the sharpness is < 1.5% of optimal
-mean based statistics used
-all mules have 6-7 dps edge bevels, 15 dps micro-apex bevels (< ~40 microns or ~0.001" wide)
-edge is finished with Naniwa Aotoshi 2000, apex bevel 600 DMT
-cardboard is random sampled
-all cardboard is used stock (is dirty)

Here are the results after just two rounds :

-the large blue boxes are the TCE values (edge retention)
-the red bars are the scatter in those results

A few immediate observations :

-the results show no significant difference at all
-the scatter is really high in some blades

I am going to show the results after 2, 3, 5 and 10 runs. These are very important because :

-2 is the first group that allows an estimation of scatter, crude but it can be done
-3 will make a large reduction in scatter and will likely point to a trend if it exists
-5 is generally enough to note trends magnitude
-10 will generally quantify it

Of course it depends on the data, but if 10 runs isn't enough to tell if there is a difference then it should be obvious for all practical purposes it doesn't exist.

As a few final notes :

-you can not extrapolate from edge retention on cardboard with one edge angle + one apex finish, cut in one method to edge retention in general
-however you can make general inferences from similar cutting done on similar material (slicing rope for example)

my mule (3 replies)

oldsailorsknives — Fri, 08 May 2015 13:12:44 +0000

hi,
finally have this one finished. will ship to Cliff on Monday.

the blade is high carbon steel, i will post more details after Cliff's test. the handle is bookmatched Chakte or red-heart.
scott

MaxaMet - McCullen (12 replies)

CliffStamp — Sun, 31 Aug 2014 00:40:57 +0000

Intro : [oldforum.beyondt01micron.com]

This is the standard mule pattern, see Jeremy's video's for many details. To start I needed to sharpen the blades, I started with one of the ones which is optimized for toughness (the softer one at ~64 HRC). This is how much of the blade was left to zero :

I asked Jeremy to grind it as close to zero as he felt comfortable, as you can see there is very little metal left at the edge. However it took three thousands passes per side, six thousand passes total to zero the blade with the Bester 700 waterstone. The bevel is now full zero at approximately 7.5 dps and again this is the Maxamet mule which is easy to grind.

This steel/hardening really is too much for the Bester which gets slick and wears down and doesn't release abrasive well, but of course it isn't made to grind this steel, in fact not many stones are given the extremely high carbide volume and ability to reach extreme hardness levels (~70 HRC).

cKc AEBL 60RC Mule (1 reply)

cKc — Fri, 23 May 2014 13:47:52 +0000

This is going to be interesting, and I wonder if Cliff is probably going to need to thicken the edge a little.

I have noticed that one side is a fraction thicker than the other above the apex creating a slight assymetry that will doubtfully effect much.

I measured depth and cross section near the centre of the blade and readings from 0.01", 0.02" and 0.03" give DPS measurements of 4dps approx at the first 2 and moving to 3.5dps at 30 thou. Thats a pretty thin knife. 0.03" is 0.252" from the apex.

I just did an 0.004" depth reading at 1 point. have to use the microsope to measure that, and it works out at around 8dps so I'd say thats where the thickening starts, I"m sure there is a micro under that also.

Calton : 1095 (34 replies)

CliffStamp — Thu, 25 Dec 2014 17:38:43 +0000

Reference thread : [oldforum.beyondt01micron.com]

I have a set of the mules from 58 to 66 HRC in the same pattern, they look like this :

They are all used, so the edges have been damaged to some extent. Here is the #221 one (the numbers don't mean anything) :

+

+

I will be doing a series of :

-hemp cuts
-cardboard cuts

and recording sharpening time/observations. I will update this each time I get a run with each knife.

I may also rig up a small impact tester to subject the edges to a specific load into a hard object like a drill bit and see if there is any difference in how the edge behaves in regards to deformation/fracture.

Spyderco thread : [www.spyderco.com])

some new mules (20 replies)

Joe Calton — Sun, 02 Mar 2014 00:06:04 +0000

got to spend a few hours in the shop today. got a good start on the 1095 mule project! and I have a pic!

the mules are the bottom row in the middle. Cliff's smaller pattern of those, with a hidden tang so I could get more out of a single piece of steel to reduce that variable, I was able to get 11 mules out of one piece of steel, and a couple more out of some scraps of 1095 from the same supplier that will be finished out and ill either keep them or sell them if I like how they turn out. the others are the next 2 weeks worth of normal sales, and I will heat treat them here in the shop.

this will be a multi part series of mules with my heat treat and a series heat treated by peter's heat treat. I hear that they do good work, and are reliable and such, but have never sent a blade out to get heat treated. but I spoke with someone at peters and here is how the experiment will be run:

of the 11 that I got from the single piece of steel, they are all marked with a "P" at the end of the tang so they don't get mixed up with the others, 10 will be sent to peters to be treated to varying hardnesses:

2 at 58rc
2 at 60
2 at 62
2 at 64
2 at 66

the 11th one I will heat treat here in the shop along with one other one from another bar of steel

they are all at 3/32. and are pretty close to the pattern. I did not grind bevels in any of them, just cut them out and cleaned up the profiles to 400 grit, and drilled the pin holes.

When they get back, I will grind them all with a flat grind to the same thicknesses {as close as my micrometer and I can get anyways} id guess maybe .010" at the shoulder of the edge, on my water cooled grinders. and make 2 handles that can be installed on any of the knives for testing. I will probably soften the ends of the tangs and mark 1-6 on each of them, and write the hardness number in my notebook as to which number is what hardness.

so when I am finished I will have 2 sets of mules, and the fella at peters told me that he would include the heat treat that he used in the box when he ships them back to me, and my heat treat to compare! I will ask peters if they could get a Rockwell on one of mine at the same time also.

once I have the sets done, I will test one set, and would like to send the other set to Cliff for his testing and thoughts. and then have a friend of mine test my set again. with just numbers on the tangs, and no knowledge of what hardness each is at, it should be about as "blind" as I can make it for cliff and my friend, but I will have to just try and forget which number is which, or have someone install the blade in the handle for me so I don't know for my testing.

does anyone have any suggestions for improving the experiment? or thoughts about it?

Mule Project (60 replies)

CliffStamp — Tue, 12 May 2015 02:24:49 +0000

The Toward 0.1 micron mule project will consist of specific blade patterns offered in multiple steels. The end goal of this program is to have a collection of mules in a number of limited specific designs to evaluate steel materials.

Imagine a knifemaker who wants to check out some new material and he just requests a mule fro the "library" in O1 for a reference, or imagine a user who is getting a custom made and wants to see how A2 performs in a large chopping blade.

Now yes there are logistic issues, but these are the end goals, I will concern myself with the practical issues later. For now we need a design for a small blade.

Design folder :

[drive.google.com]

The small blade should be capable of :

-utility
-food
-hunting

It should not be extremely narrow in scope and should also not be extreme in application, i.e, 1/4" or 1/16" are not good choices for these reasons.

[www.youtube.com]