BESS Score to Edge Width correlation in the lower BESS scale

[KnifeGrinders]

I start with the conclusion, and then list arguments; you can do your own analysis of the sources I give, or just take my word for it:

50 BESS approximates to 0.1 micron in the razor range of the BESS scale (10-100).

Sources

Chase Anderson's raw data of DE razors sharpness by PT50A, including that of Feather, Gillette, Wilkinson, Kai, and Derby brands.
https://www.refinedshave.com/razor-blade...s-testing/
Thoroughness of his measurements is worthy of respect.

Todd of the scienceofsharp.wordpress.com also known as ToddS on bladeforums.com and fuzzychops on straightrazorplace.com
Todd, a professional with access to SEM microscope, gave edge apex width for Feather, Gillette, Wilkinson, Kai, and Derby DE razors.
http://straightrazorplace.com/honing/968...rop-5.html
https://bladeforums.com/threads/how-low-...402/page-2
https://scienceofsharp.wordpress.com/

Correlation

avg. 32 BESS, Feather (variation 30%) - 0.05 micron 
avg. 46 BESS, Gillette & Wilkinson Sword (variation 15%-47%) - 0.075 micron (I took an average of the 0.05-0.1 given by Todd)
avg. 53 BESS, Kai & Derby (variation 16-33%) - 0.1 micron 

Approximations

As seen from the raw data at http://www.refinedshave.com, where BESS measurements are taken at 4 points along each razor edge, variation in BESS scores is huge, up to as much as 47%.
This is because the razor edge itself is far not that even in width as one would expect a priori.
Test clips used in that measurements vs BESS-certified media have difference in readings less than the dispersion on razor edges, and can therefore be neglected.

SEM measurements depend on the person expertise and subjective judgement.
Even with fully trusted SEM numbers, we know they were not taken on the same razors that were BESS-tested, let alone not on precisely the same points of their edges - we therefore can expect at least the same variation in results as we see in BESS scores along the edge.

CONCLUSION
I know it is oversimplified, yet to a first approximation it can be taken that with the step of 25 BESS, this correlation might be linear.

25 BESS = 0.05 micron edge apex width
50 BESS = 0.1 micron edge apex width
75 BESS = 0.15 micron edge apex width
100 BESS = 0.2 micron edge apex width

This is in line with Mike Brubacher's findings, summarized in the "The Science Behind Edge On Up Instrumentation and BESS Test Media" section of the "Support Document for All Edge Tester Models" available in the EdgeOnUp library.

[Jan]

KnifeGrinders, thanks for your important report concerning razor blades in the lower range of the BESS sharpness scale.

I have read with appreciation your conclusion, that in a first-order approximation the BESS readings may be directly proportional to the edge apex width.

This simplifies further considerations and complements Mike’s measurements showing, that for BESS 50 to 300 gf the relationship between BESS and edge width is essentially linear.

Jan     

[KnifeGrinders]

To pass Jan's vetting means a lot to any concept, it is like switching from red to green at traffic lights. Thank you

[Jan]

Please do not overestimate my judgement skill. Thank you.

[Rupert Lucius]

To pass Jan's vetting means a lot to any concept, it is like switching from red to green at traffic lights. Thank you

Rupert also agrees.

[EOU]

Once again, an excellent exchange of information and ideas but what we we really want to say is this; three guys, three totally different backgrounds and emanating from three different continents having a friendly and informative discussion on the BESS Exchange. Makes us want to "bust our buttons" and we hope that saying translates well in all regions of the world. Thank You all!

[KnifeGrinders]

I've mimicked Chase Anderson's guillotine and compared his scores for test clips used in his measurements to BESS-certified media.
I tested 3 Wilkinson Sword razors on both sides with a PT50A, the same instrument he used.
Wilkinson Sword averaged 46 on test clips in Chase Anderson's tests.
First I tested with a couple of test clips, and got the same score of 46-47 he got, meaning my testing technique is close to his.

Average of the 6 tests with the BESS-certified media is 30, and the scores varied from 23 to 38.



The score of 30 for a BESS Certified media meets our expectation for the approx 0.075 micron edge apex width of this razor by SEM.
We can also see that in the lower BESS scale scores may be lower than those obtained by test clips.

[grepper]

Nice setup. 

When you say, "BESS-certified media", do you mean that the media is in the test media fixture rather than test clips?  If so, the results seem backwards to me.  Test clips should show 20% lower (sharper) readings than test media in the test media fixture, but your results seem the other way around.

[KnifeGrinders]

Yes, that's what the instrument shows.
Test clips aren't BESS-certified.
Under 100 on the BESS scale, test clips score higher than the BESS-certified spooled line, but in the 100s range of the scale Mike says test clips score by 20% less.

Speaking of non-certified test media, Mike said: "Some may resemble the BESS on the bottom end of the scale but miss badly in the middle or top and vice versa."

Yet that difference in scores might be due to my little experience in testing razors, I'll be testing more with the guillotine after I add a counterweight to it, and if the numbers get corrected, will update them.
Razors are interesting, far from knives, have never seen that many popping & kachunks as with them. Had to re-read the KN100 Manual in the part about testing razors to understand them better.

UPDATE:
With an added extension spring as a counterweight, I got better consistency in scores, but still the same number of 30 for the spooled BESS-certified test media, versus 39 for test clips.
Another interesting difference between the two, is that the BESS-certified media gives notably less variation in scores compared to the test clips - on PT50A, I am getting +/- 1 for the BESS-certified line (which is the instrument resolution), while for the test clips +/- 6.


I used a range of screws of 4 to 5mm diameter in the 5mm holes to align the horizontal bar perpendicular to the test media

[Jan]

KnifeGrinders, thanks for sharing the image of your guillotine for sharpness measurements. Really very nice and simple setup.

I have successfully experimented with Proxxon Drill Stand. It has a well designed feed lever, which operates via ball bearing system and offers very sensitive feed. Smooth and reproducible blade feeding reduces the scatter of BESS readings.

   

For me, as a Tormeker, it is sometimes important the measure the sharpness while the blade is mounted in the knife jig. Alternatively the Proxxon Machine Vise can be bolted to the headstock with jaws facing the table. 

       

Jan