Edge Sharpness Tester

Just unboxed mine this morning, rec'd yesterday while at work and didn't get home till 8:30PM.

My kids helped me set it up before I took them to work and I was able to run a few tests.
First off I'd like to give a quick impression - unit is light, easy to understand and set up, no frills/none needed. I'd like to make one just like it only scaled up for cutting 3/8" sisal :D

My results based on very limited testing are showing reasonable consistency. I am seeing a bit of creep in that a delay under load seems to reliably produce lower values but I recall seeing a bit about this in the manual - need to reread that section. Will run a few more tests to see if I can pin down the difference - I suspect it is not a large margin. Adding shot until break point seems like I'm adding more than needed, but again probably not a significant enough amount of weight to mention.

My wife's Chicago utility kitchen knife XXC DMT burnished with smooth steel, supposed to be effective yet have an Oops margin of error=
163/158/168, I should maybe dull it up to the mid 200's

Tops CAT 1095 @ 28° inclusive, finished on Washboard with stock honing compound, my current EDU =
42/47/43

Jarvenpaa Puukko 26° same as above =
38/44/45

Serrated portion of my Frankenfolder, Fred Carter MK4 440C ground to fit Benchmade 720 frame =
101/94/104

Will have to continue as I get some more time, that's all for today. So far I like that the numbers coming back for similar prep across different knives are so close. Also numbers within each set seem to be well within reason. If I'm reading the manual correctly, the margin of error is +/- 10 grams.

Am I supposed to be adding the shot value to the existing 55 grams, or is my BESS value only what is added past the weight of the platen and piston (DE Razor drop)? My #s are just what has been added to make a drop.

Edit to add: am guessing folks are using the additional grams, otherwise some would be severing with less than full weight of platen and impossible to take a number...:confused:

Fun stuff!
Martin
 
Last edited:
Komitadjie said:
Was a bit pressed for time last night, but I was able to start a testing series. I ran four of my favourite knives through my usual sharpening process on the Kalamazoo, and tested them within a few minutes of coming off the belts. The progression used was all Trizact for abrasives, A65, A16, A6, A3, followed by 1u Boron Carbide on a SurgiSharp leather belt, and .5u Chromium Oxide on a SurgiSharp belt. The edge is my normal working edge, and will shave armhair cleanly and easily with all four knives, and if I'm careful pass a low-level HHT. I can do somewhat better on the belts if I take more time, but I wanted to try my "normal" edge out and see how it looked.

Northstar EDC (3V steel, same knife tested in the previous series)
#1 = 86.0
#2 = 82.4
#3 = 86.2
AVG = 84.9

Spyderco Paramilitary 2 (same knife tested in the previous series)
#1 = 82.2
#2 = 81.1
#3 = 86.6
AVG = 83.3

BRKT Mini Canadian (A2 steel)
#1 = 108.2
#2 = 108.1
#3 = 109.2
AVG = 108.5

BRKT Saex Prototype (CPM-D2 steel)
#1 = 109.6 - Discarded this number, see more below
#2 = 123.4
#3 = 118.9
#4 = 121.7
AVG = 121.3


While I was testing the Seax, I noticed something interesting. I used up the media I had in the cup I use for "sifting" in the shot to the top cup on the test unit from the Mini Canadian test, and had to go to one of the little storage containers for a bit more. It took me perhaps ten seconds to reach for more mass, and in that time, the tester dropped on its own! Since I was not entirely sure what I had done there, I discarded that data point from my calculations, and completed the rest of the test series as planned. Since this struck me as something rather interesting, once I'd completed the test battery, I came back and made another pass at it. For this test, I chose the Seax (since it had displayed it once before already) and I chose a weight of 111 grams (since 111 grams is approximately 10 grams below the weight needed to cause a prompt media failure, according to the previous tests). My method for this test was to very, very carefully place the weight cup on top of the ram platform, then move my hand back to start a stop watch, and record when the ram head dropped.

The results of four tests are as follows:

7.3 seconds to cut
3.5 seconds to cut
4.2 seconds to cut
1.1 seconds to cut

I will have to play with this a bit more, I was not expecting to see that! I will also double-check my instrument again, with regard to ram mass. I am absolutely sure that my lock screw was fully disengaged for those tests. I also note that there was a reasonably steady downward trend in the time-to-cut, with no change in mass on the ram, or method of testing, that I have no idea of the cause for. I will play with this one more, I think. Would someone with a decent microscope like to give this a try, and see how it looks under magnification?

Finally upgraded to one of my good notebooks!
IMG_8009_zps4spkkijd.jpg
Click to expand...

Yes this is what Jason pointed out... something needs to be standardized about this and how you specifically add media to the cup as there are different ways to do it and immediate vs. delayed severing of the BESS media
 
Martin, you are correct in how you are interpreting it. The "zero" is the 50-gram weight of the piston assembly. The scale begins with "1" at the first gram you add to the mass holder. :)

Josh, I think you might be right there. I'm going to play with this a bit, clearly there is going to be a zero point at which the media will NOT sever with time... I'd like to find and confirm that for a known edge, then play from there. I know my results are at least consistent to each other, since I apply the weight the same way each time, but there might need to be a standard added for how you apply the mass, and how long you wait before adding more.

EDIT: Mike just emailed me, and helpfully pointed out that the exact situation I am seeing regarding a slightly delayed severing on the test media is covered on page 33 of the user's manual, regarding proper application of mass to the ram assembly. It is a known (and specified) piece of data that when you are within 10 grams of the severing weight, a delay may cause the piston to drop with less mass. That correlates well with the large deviation in time-to-drop that I was seeing in my quick test on that subject, since I was right on the fuzzy edge of ten grams less than prompt-cut mass. The manual specifies that mass should be added smoothly until the media severs, and if the mass is indeed added without pause, the reading will be consistent. That also correlates well, given how I did not observe that phenomenon at all until I had to pause very near the cut point to add additional mass to my cut.

I will test this myself this evening, but it looks like what I've run into has already been tested and verified by the crew over at EdgeOnUp.
 
Last edited:
Did some experimenting last night and today using one knife, tested it, touched it up, tested again, brought up a burr on the edge and removed that etc then tested it again.

All at 400 grit, no stropping, M390, .006" behind the edge, 10 DPS.

1st test, sharpened a year ago.

52, 43, 53 = 49.3 ave

2nd Test Touched up

45, 46, 53 = 48 ave

3rd test - Brought up a bur sharpened again:

48, 48, 43 = 46 ave

So I have a spread of 3.3 grams in the 3 runs.

Ave of all 3 runs = 47.7 grams


I would say that is pretty accurate. :thumbup:
 
Those readings are pretty consistent. My question is whether two different testers -- both getting good consistency in their own readings -- would also come up with the same BESS scores on the same knife edge.
 
Just ran quick test with an old Imperial carving knife. Pretty low RC stainless, about 30° inclusive, maybe a touch larger.

320 grit wet/dry
196/205/213

Smooth steeling for 8 passes
126/134/121

Washboard with paper and stock compound
67/60/64

These numbers are falling very close which is a good sign. I'm also coming to the conclusion that respectably sharp edges start around 200 or so. Have been running some fine newsprint cut tests in parallel, some edges that show marked improvement cutting paper do not register much change on the BESS in some cases - no conclusion, just observation.

So far so good, am going to have to cross reference the BESS numbers with some other standardized cut test(s) once I get more familiar with it.
 
Regarding two different testers it could be as follows and might be done that way already:
Take a standard blade test it and set the value you get to 100 (or any other number)
The difference the actual measurement of that standard has has to be subtracted (or added) to all the other tests that day.
This should reduce variance in numbers caused by slightly different ways of measurement.
 
Ankerson said:
Did some experimenting last night and today using one knife, tested it, touched it up, tested again, brought up a burr on the edge and removed that etc then tested it again.

All at 400 grit, no stropping, M390, .006" behind the edge, 10 DPS.

1st test, sharpened a year ago.

52, 43, 53 = 49.3 ave

2nd Test Touched up

45, 46, 53 = 48 ave

3rd test - Brought up a bur sharpened again:

48, 48, 43 = 46 ave

So I have a spread of 3.3 grams in the 3 runs.

Ave of all 3 runs = 47.7 grams


I would say that is pretty accurate. :thumbup:
Click to expand...

Great results Jim! What do you do for burr minimization?
 
razor-edge-knives said:
Great results Jim! What do you do for burr minimization?
Click to expand...

I knock the burr off cutting into plastic lightly, straight into it like once or twice.

Then make some very light passes across the edge on the Edge Pro flipping the blade every 1 or two passes, checking for a burr as I go.

That's what works for me anyway. :)
 
Last edited:
Twindog said:
Those readings are pretty consistent. My question is whether two different testers -- both getting good consistency in their own readings -- would also come up with the same BESS scores on the same knife edge.
Click to expand...

I can't see why not if they are doing it correctly.

Or extremely close anyway within a few grams and that's basically nothing.
 
HeavyHanded said:
Just ran quick test with an old Imperial carving knife. Pretty low RC stainless, about 30° inclusive, maybe a touch larger.

320 grit wet/dry
196/205/213

Smooth steeling for 8 passes
126/134/121

Washboard with paper and stock compound
67/60/64

These numbers are falling very close which is a good sign. I'm also coming to the conclusion that respectably sharp edges start around 200 or so. Have been running some fine newsprint cut tests in parallel, some edges that show marked improvement cutting paper do not register much change on the BESS in some cases - no conclusion, just observation.

So far so good, am going to have to cross reference the BESS numbers with some other standardized cut test(s) once I get more familiar with it.
Click to expand...


My own opinions on the sharpness levels based on the BESS numbers from what I have seen, only my personal opinion though.

  • 50 and under is crazy sharp
  • 51 to 100 is nuts
  • 101 - 200 very sharp
  • 201 - 300 Reasonable
  • 301 - 400 Working edge
  • 401 and up needs to be sharpened
 
Last edited:
Had enough time to prep and run one knife -

Bluntcut Metalworks s110v @ 63 RC 24° inclusive, finished on Washboard with stock SiC compound over paper.

Don't have the thickness behind the edge as the shop Verniers walked off again - primary is very thin into a near full curvilinear edge.

16/13/15
 
HeavyHanded said:
Had enough time to prep and run one knife -

Bluntcut Metalworks s110v @ 63 RC 24° inclusive, finished on Washboard with stock SiC compound over paper.

Don't have the thickness behind the edge as the shop Verniers walked off again - primary is very thin into a near full curvilinear edge.

16/13/15
Click to expand...
That's pretty much insane. So... convex edge?
 
Omega Leather Works said:
That's pretty much insane. So... convex edge?
Click to expand...

Hey old buddy, haven't had a chance to say howdy (I realize you gotta promote the enterprise, but I liked your old avatar better)!:)

Thanks to another thread, I realized that sort of edge isn't really convex, because its thinner than a V bevel with the same apex. Now I refer to it as "curvilinear"...

Yeah, it is crazy sharp and some seriously tough steel with great HT. I've used it some but haven't really EDUd it yet.

And I lied about only testing one knife. Prepped my Bark River Necker 2 in 12C27 - another curvilinear edge at about 24 inclusive.

Sharpened on 600 grit wet/dry over Washboard

137/166/141 stropped/burnished on plain paper over WB and re-tested

88/63/83

Real catchy.

I do not believe there is any way to test draw-cutting qualities with this particular media, though the equipment is a good platform for any small edge testing if a different media could be found/agreed upon. The current material is very abrasion resistant - dragging a hacksaw blade across it several times and it didn't fuzz up or show any real breakdown - dragging an edge across would either snag and cut, or dance across. It would be easy to rig up a roller bearing guided clamp.
 
WB + SiC compound over paper is the key for producing these insanely low BESS #s. My hand-waving theory... Those fine(fractured) sharp SiC probably shaped & decorate the apex with large VC teeth and sub-micron NbC teeth in the trough. That combination of uneven high pressure points: VC teeth partially cut & wedge, NbC teeth cut from below. To improve, I think a hard rubber coated twisted fiberous string could expand the BESS test range.

When you've a chance, it would be informative to see BESS # for BCMW D2 at same 12dps & your WB magic. And impatiently waiting for(upgrading to) your next generation WB...

HeavyHanded said:
Had enough time to prep and run one knife -

Bluntcut Metalworks s110v @ 63 RC 24° inclusive, finished on Washboard with stock SiC compound over paper.

Don't have the thickness behind the edge as the shop Verniers walked off again - primary is very thin into a near full curvilinear edge.

16/13/15
Click to expand...

Thanks, Chris!

Chris "Anagarika";14798865 said:
BCMW rocks & it's insane indeed!
Curvilinear @ 24 inclusive is exactly sharp :D
Click to expand...
 
Maybe a dumb question?

I'm curious how the tension of the test thread is set, and how stable that is. So if you set the thread many times, what is the distribution of thread tensions? Or just the standard deviation of the tension. I'm also curious about how sensitive the test is to different thread tensions. So if we ran tests at a slightly higher (or lower) tension, how much does that affect the result?

Originally, I was guessing that the tension would be set by hanging a standard weight on the thread while the thread goes over a bend or pulley, and then is clamped by pincers (that way there is no pulling or pushing of the thread while tightening down a screw or nut). I think this would be a reasonable way to consistently set the tension and take another variable out of the test.

As a physics major, I would guess (without having worked out all of the details), that if the test media were very inelastic and brittle, then tension could matter a lot. This is simply due to the geometry of how a thread stretches when being clamped at both ends and then loaded so that it changes from a straight line to an angle (or v shape). If you are an engineer, or physics major, you can see this is a very interesting problem. (You can analyze it several ways, such as the method of virtual work.) What you would study is how the stiffness of the thread affects the tension of the thread for a given weight. It would be interesting to get some formulas and plots for:

Input:
Thread Tension when unloaded (no load weight)
Thread Stiffness
Load weight

Output:
Thread Tension when the load weight is applied.

Pretty busy, but might be fun to work this out? If you're a technical person, feel free to do so and we can check each other's results, etc.

A couple of caveats: In the above analysis I'm assuming the thread acts like a very stiff spring (linearly elastic). If the thread is specifically designed to be non-linear, then all bets are off (much harder to analyze). But this could be on purpose, as it might improve the tests repeatability (precision).
 
Lagrangian said:
Maybe a dumb question?

I'm curious how the tension of the test thread is set, and how stable that is. So if you set the thread many times, what is the distribution of thread tensions? Or just the standard deviation of the tension. I'm also curious about how sensitive the test is to different thread tensions. So if we ran tests at a slightly higher (or lower) tension, how much does that affect the result?

Originally, I was guessing that the tension would be set by hanging a standard weight on the thread while the thread goes over a bend or pulley, and then is clamped by pincers (that way there is no pulling or pushing of the thread while tightening down a screw or nut). I think this would be a reasonable way to consistently set the tension and take another variable out of the test.

As a physics major, I would guess (without having worked out all of the details), that if the test media were very inelastic and brittle, then tension could matter a lot. This is simply due to the geometry of how a thread stretches when being clamped at both ends and then loaded so that it changes from a straight line to an angle (or v shape). If you are an engineer, or physics major, you can see this is a very interesting problem. (You can analyze it several ways, such as the method of virtual work.) What you would study is how the stiffness of the thread affects the tension of the thread for a given weight. It would be interesting to get some formulas and plots for:

Input:
Thread Tension when unloaded (no load weight)
Thread Stiffness
Load weight

Output:
Thread Tension when the load weight is applied.

Pretty busy, but might be fun to work this out? If you're a technical person, feel free to do so and we can check each other's results, etc.

A couple of caveats: In the above analysis I'm assuming the thread acts like a very stiff spring (linearly elastic). If the thread is specifically designed to be non-linear, then all bets are off (much harder to analyze). But this could be on purpose, as it might improve the tests repeatability (precision).
Click to expand...


The manual describes using the weight of the spool, basically just take the slack out and give a little tension. Anecdotally the amount of deflection with the weight of the bare platen/piston appears pretty uniform. The testing head could have lines inscribed that show where it should be, but that might not be practical. I've played around with it a bit, and seems to have very little stretch if any. Is slow to take a hard set as well. Am not sure, but having it too tight almost seems to increase the load it can tolerate before severing.

Supposed to be +/- 10 grams. I lock one side down, gently pull the line on the other side and lock it down. I haven't had but a handful of readings that fell outside the margin of error and might even be the blade itself, as for safety reasons I remove it from the stage every time I spool more test media and cannot be 100% sure it lines back up in the exact same spot.

My next question is to see if on lower RC steels the edge is being visibly effected by the test media prior to a drop in a way that would effect the outcome. It might just be me, but the higher RC/ higher carbide steels I've tested seem to reliably score a little higher at what I believe to be comparable finish values and results on newsprint.

Won't be easy to tell even with a good microscope, and suppose I do determine this, to what degree will also be near impossible. Would be enough to say that there is a trend (if there is!), and I hope the other eval participants might keep this in mind as they go to see if they have similar experience. I haven't done enough testing to say and at 400x I'm not seeing anything obvious, just a very slight gut feeling with no data to back it up.
 
Lagrangian,

Well put, I expressed concern for the how the test media bows over the edge in earlier posts and what you have said in a way backs up my though on the subject. This picture illustrates the real size difference between the edge and test media, and how with varying test media tension there could be varying degrees of wedging and binding.

EE208840-1834-4A95-BFAF-1069748B5AA1.jpg
 
Jason, what are you using as a relative gauge for tension? I had considered using a small spring scale we have at work - like a fish scale but fairly precise.
 
You must log in or register to reply here.
Back
Top