Sharpness Chart

[maximus83]

Maximus, these super sharp edges do survive on soft wood cutting boards even if made not of a super steel.
I sharpened my wife's Global Santoku to 0.1 micron, the same as a Gillette safety razor, and after 3 days of constant use it still shows under 0.2 micron edge.
This honestly surprised me, as I used to think that level of sharpness be more volatile.

That is good to know! Don't want to derail your thread here off into a different subject, but would be interested sometime to in a different thread discuss what combinations of edge types, and edge angles, that you find hold up the longest in your commercial sharpening business.

 

[maximus83]

Similarly to you, with my blades I've seen all variations between a perfectly clean cross push-cut, and rough.
The PT50 Edge Sharpness Tester shows that:

- roughens (still cuts, not tears) - 35 BESS or 0.07 micron edge apex width;
- almost clean cut with rough spots - 30 BESS or 0.06 micron edge;
- clean cut - 20-25 BESS or at least 0.05 micron edge and under.

Ok, so this makes sense. In simplistic terms, a "fail" is a knife that rips the paper. For knives that "pass", how clean the cut is correlates to edge width and thus sharpness. Taking the above pic I posted as an example, under magnification I can see really easily that although the Buck cut the Rizla paper, it is more 'jaggy' especially near the starting point of the cut. The CS cut is so much cleaner. And from all this, I draw this rather stunning and unexpected conclusion: Josh @razor-edge-knives is better at sharpening than I am, and should have no fear that I am any competitive threat to his sharpening business.

Q for wootzblade : How far from the point at which you are push-cutting the cigarette paper are you holding the paper? For the above pictured cut, I was holding about 3/4 inch from the cut point.

 

[Natlek]

For me, the point is just an easy low-tech way I can use at home to indicate I've gotten a knife up to some acceptable minimum standard of sharpness, without having to go test-cut on a bunch of tomatoes, boxes, or other tests that are expensive or a hassle. In practical terms: if I'm sharpening a kitchen knife, I'd rather find out while I'm still at the stone, working on it, that it's not as sharp as I'd prefer yet. Rather than waiting until all my stones are cleaned and put away, and I'm slicing a tomato in the kitchen later, and find out it's still too dull. Or have to go cut a fresh tomato every time I sharpen the knife to find out it's sharp enough. That gets expensive. Most of my knives I can test up to the more basic level I posted above and don't need the cigarette paper testing or extreme level of sharpness. It's just helpful to know which of these informal tests correlates to what level of real-world sharpness I can expect.

As for the extreme sharpness and edge testing, the cigarette paper and all that: for me at least, this is not a necessity, it's more a matter of interest, and a challenge, to see how far you can go in sharpening. I'd have no expectation that edges that sharp are going to survive very long. I won't even use the cigarette paper test probably 90% of my real world knives.

Don t get me wrong , but I KNOW how sharp the knife will be before I start to sharpen it ? I do not need to do any test after sharpening to see how sharp is it..... ? I decide how I will sharpen /how I want that knife to cut/ and I just do it ....... It's not rocket science .
But as you say , I like challenge The knife in my picture in practice is probably unusable ,it is good only for demonstration . On brass rod test it flex deep in blade .........that thin is it behind the edge now .But it slice like laser ...........

Okay, I am sorry for of topic

 

[wootzblade]

...
Q for wootzblade : How far from the point at which you are push-cutting the cigarette paper are you holding the paper? For the above pictured cut, I was holding about 3/4 inch from the cut point.

For the rolling paper I follow the Hanging Hair Test guidelines - cut at about 2 cm (3/4 inch) from the holding fingers.
HHT was standardized by Todd S (scienceofsharp.wordpress.com) and his friend, full description is here:
http://shavelibrary.com/index.php/Hanging_Hair_Test,_from_trick_to_probing_method

Smilarly to HHT, edges that carry microscopic sawteeth cut rolling paper easier vs polished.

 

[bucketstove]

Hi
nice work

Finally I've got a full assortment of Rizla cigarette rolling papers... I've been told the Tally Ho paper is thinner than many other rolling papers, and 10 times thinner than a human hair.
http://knifegrinders.com.au/Manuals/Sharpness_Chart.pdf
Cross push-cut:
Longitudinal push-cut:

Labels easily get misused as a description, I'd prefer if the labels were absolutely honest
cigarette paper Cross slice-start-push-cut:
cigarette paper Longitudinal slice-start-push-cut:
because slice to start then push cut the rest of the way is much easier than a push cut with no slice from the start
That assumes the knife is 90 degree to the paper (perpendicular) in all 3 axes

Speaking of papers, it would be useful to include paper thicknesses,
~0.0040 inch is ~100 microns average a4 printer/copy paper
~0.0020 inch is ~50 microns for phonebook or yellow pages paper
~0.0010 inch is ~25 microns for Tops cigarette rolling paper
~0.0005 inch is ~12 microns for OCB Hemp cigarette rolling paper
ref Re: Practical Ways to Evaluate Sharpeness

HHT was standardized by Todd S (scienceofsharp.wordpress.com) and his friend, full description is here: http://shavelibrary.com/index.php/Hanging_Hair_Test,_from_trick_to_probing_method

What does that mean exactly, I can't say I've read ToddS write anything about HHT.... shavelibrary/coticule/bart HHT description is dated 2010 and before

Before out study, the only available definition of the demarcation between sharp and dull was Steve Bottorff's "Any edge thickness under a few thousands of an inch could be considered sharp."

Cute If only it were true

razor patents and john d verhoeven talk about edge thickness and thats 2004 and before,
ref Measures of cutlery sharpness (Review) : Cliff Stamp
ref what-is-possible.1008190/#post-11478832


My people would call this sharp not dull 2-3 thousands of an inch , 0.002 - 0.003 inch or 50.8-76.2microns , after all food is soft and we cut forcefully

 

[maximus83]

What does that mean exactly, I can't say I've read ToddS write anything about HHT.... shavelibrary/coticule/bart HHT description is dated 2010 and before

I've only recently become familiar with Todd and read his blog. He mentioned the HHT test in his site a few times and refers to his friend's article.
https://scienceofsharp.wordpress.com/2015/03/24/sharp-and-keen-part-2/
https://scienceofsharp.wordpress.com/2016/04/14/simple-straight-razor-honing/

 

[miso2]

50 nm apex radius is a norm for straight razors - but astonishing for knives.
Freehand sharpening has always been a magic to me.
The Maximus's M390 folder edge apex radius is 35 nm. And he is still not happy.
To get my knife to fillet a receipt docket like he does, I had to use CBN wheels, jigs and computer software.
How they can achieve that sharpness on knives of that hard steel freehand without bartering away their soul, I will never comprehend.

I think this is an over (or under?) estimation.

The measurements by Dr. Verhoeven in 2004 using SEM showed on some commercial razor blades showed that the edge width of around 200~300 nm (average minimum) is typical. This is consistent with what ScienceofSharp reported on the web site (~200 nm wide, ~100 nm radius on average) using FIB-SIM.

Also, ScienceofSharp describes that 50 nm edge width was the thinnest (not a norm) among them and is near the limit of what can be achieved by honing and stropping.

I am sure that even a 30 nm wide edge is possible (Standard for Flatness, I believe, was maintained by the hand of meister at the atomic level in the past). But it would not be an easy task.....


I am just questioning the relationship between BESS and edge thickness. That between the traditional sharpness tests and BESS would still stand.


Edit: Correct for consistency of edge radius and width.

 

[wootzblade]

Actually, Prof. Verhoeven estimated Gillette DE razor edge width of 0.35-0.45 micron (page 6), i.e. 350-450 nm.
Doesn't that surprise you as much as surprises me?
Human hair is 0.3-0.5 micron thick, how on Earth can it be shaved with the Verhoeven's razor?
It's like chopping down a tree with an ax having edge width of the tree diameter.

See to yourself that Verhoeven's numbers is an unfortunate mistake by checking the Gillette patent, and posts by
Todd Simpson of the scienceofsharp.wordpress.com also known as ToddS on bladeforums.com and fuzzychops on straightrazorplace.com

https://www.google.com/patents/WO2010132645A1?dq=EP2429777+A1&ei=lhTYVITCD4rLsASK9IEY&cl=en
http://straightrazorplace.com/honing/96867-why-do-we-still-use-clean-leather-strop-5.html
https://bladeforums.com/threads/how...to-go-in-microns-for-stropping.1259402/page-2

This, of course, in no way diminishes the importance of prof. Verhoeven's study on grinding with Tormek.
We turned Verhoeven's trigonometry into a computer algorithm, to set grinding & honing angle with 0.1 degree accuracy on all slow wet rotary grinders, and paper wheels. By now hundreds of sharpeners on every continent use our software http://knifegrinders.com.au/05Equipment_scripts.htm

Miso2, when you say "I am sure that 30 nm edge is possible..." i see that you confuse edge radius with width.
When we say that Maximus' edge apex radius is 35nm it means apex width of double that, 70nm or 0.07 micron.

By SEM studies DE razors average apex radius is 50 nm +/- 12.5 nm, i.e. width at about 100 nm or 0.1 micron.
Most of these "average" razors will fail the Tally-Ho/Rizla Green cross push-cut test; only the best like Feather which is twice sharper than average will pass it cleanly.

 

[Natlek]

I don t think that my knife has THAT /sharpness / thin edge radius ...........BUT have VERY ....... VERY thin geometry .......................??

 

[HeavyHanded]

Actually, Prof. Verhoeven estimated Gillette DE razor edge width of 0.35-0.45 micron (page 6), i.e. 350-450 nm.
Doesn't that surprise you as much as surprises me?
Human hair is 0.3-0.5 micron thick, how on Earth can it be shaved with the Verhoeven's razor?
It's like chopping down a tree with an ax having edge width of the tree diameter.

Human hair is approx 50 micron + 10 or so. Microtome blade manufacturers claim sample thickness as small as 2 micron.

I'd also point out that SEM has limitations when it comes to making precise measurements at that scale directly from the image. While it is somewhat common practice, there are image distortion issues that can effect accuracy.

 

[miso2]

By SEM studies DE razors average apex radius is 50 nm +/- 12.5 nm, i.e. width at about 100 nm or 0.1 micron.

Thanks for the reply. I edited my post to be consistent in the use of radius and width.
Anyway, I would like to know the source of the above claim that DE razors have 100 nm edge on average.

 

[wootzblade]

Thanks for the reply. I edited my post to be consistent in the use of radius and width.
Anyway, I would like to know the source of the above claim that DE razors have 100 nm edge on average.

Historically first was Sackman's estimate of 0.1 micron for a Wilkinson razor - his work is among references in the "Experiments on Knife Sharpening" by Verhoeven.
Most exhaustive data known to me are in the Gillette patent.
In line with them are Edge-On-Up SEM studies - e.g. see SUPORT DOCUMENT For All Edge Tester Models at http://www.edgeonup.com/Library.html

Finally, you can come to the same average around 100nm when you match DE razor edge widths given by Todd at http://straightrazorplace.com/honing/96867-why-do-we-still-use-clean-leather-strop-5.html
with BESS scores at https://www.refinedshave.com/razor-blade-sharpness-testing/
and extrapolate to the rest of the DE razors measured by BESS tester, that don't have direct SEM estimates.

Personally I think they should average lower than 100nm, you think - higher, but the general consensus is somewhere around 100nm.

BTW, matching of the above sources with BESS scores should relieve your doubts when you say "I am just questioning the relationship between BESS and edge thickness."
Establishing this correlation didn't come easy, there's lots of work behind it.
It is true for polished edge of finer-grained steels and CPM;
It is not true for ragged edges, edges off a coarse grit with pronounced toothiness, and unpolished edges with huge carbides protruding here and there.

This BESS_score - to - edge_width correlation is almost linear in the scale range from 10 to 400, and becomes significantly non-linear above 750.
Since we are talking of knives, we don't care much about the scale range over 500, which is a dull knife - that range is for certain industrial blades, chiefly in plastic manufacturing.

 

[wootzblade]

Human hair is approx 50 micron + 10 or so. Microtome blade manufacturers claim sample thickness as small as 2 micron.

I'd also point out that SEM has limitations when it comes to making precise measurements at that scale directly from the image. While it is somewhat common practice, there are image distortion issues that can effect accuracy.

Thanks for the correction.
I must have had a momentary mental blackout, I was thinking of hair cuticles when typing that, and that a razor as thick as prof. Verhoeven says wouldn't be able to get between them and split the hair, but my fingers typed nonsense about the tree and an axe.

You are right about SEM - size estimates are subjective, and for this very reason even professors may mistake.

 

[BluntCut MetalWorks]

A cusp has radii. E.g. if top of an apex is flat 1nm wide (a line - i.e. no curvature), radius would be infinite/undefined. I believe edge 'radius' describes - the line between arc end points, where actual geometric radii values are much larger for these 2 end points. So width = cross-section/truncated-radius.

 

[wootzblade]

True.
For that miniscule world, variances around the mean are huge.
Reading the Gillette razors patent, I was surprised by how inconsistent their razors' edge width is along the edge, differing up to as much as 40-50% - in the world of Angstroms it's like beads on a string.

We have no choice but to resort to averages and approximations.
This is implied whenever we talk of them, as no one will have patience to stipulate everything he writes or says by "To a first approximation,..."

Humans simplify reality to make it comprehensible.
As soon as we forget that what we say is a simplified representation of the reality, everything sounds like lies.
That's what the poet meant when he said "A thought once uttered is untrue."

 

[Natlek]

A cusp has radii. E.g. if top of an apex is flat 1nm wide (a line - i.e. no curvature), radius would be infinite/undefined. I believe edge 'radius' describes - the line between arc end points, where actual geometric radii values are much larger for these 2 end points. So width = cross-section/truncated-radius.

BluntCut MetalWorks , let see if I understand what is about in this topic....we basically speak HOW thick is bevel behind edge ???

 

[maximus83]

Couple of interesting checkpoints, just playing around.

• A Feather blade cleanly and effortlessly push cuts cross-grain on Rizla green, with no ragged edge visible anywhere, and no need to start a cut for it. Extremely impressive. I happen to have feather blades on hand since that's what I shave with. Confirms why I like them so much.
• Josh @razor-edge-knives sharpened the CS UH that I mentioned earlier, this will ALSO will push cut cross-grain without the need to start a slice. It is SLIGHTLY--under 30x magnification--more ragged at the start point than the feather. But it's definitely slicing not tearing even at the start point. I have a few other knives that can do the push-cut without tearing if I start the slice point for them per the suggested test. But this knife is the only one in my collection I've found up to this point that will cleanly start this slice without tearing it and without any help.

 

[BluntCut MetalWorks]

No, just the very cutting edge, maybe 0.01-2 micron tall, where topology reached (whatever subjective) uniformity to consider that is the apex/cutting-wedge. Bevel shoulder is 20 to thousands of microns (micro to none) away/up from apex, where thickness is ~ commonly 0.1mm and up

BluntCut MetalWorks , let see if I understand what is about in this topic....we basically speak HOW thick is bevel behind edge ???

 

[wootzblade]

Couple of interesting checkpoints, just playing around.

• A Feather blade cleanly and effortlessly push cuts cross-grain on Rizla green, with no ragged edge visible anywhere, and no need to start a cut for it. Extremely impressive. I happen to have feather blades on hand since that's what I shave with. Confirms why I like them so much.
• Josh @razor-edge-knives sharpened the CS UH that I mentioned earlier, this will ALSO will push cut cross-grain without the need to start a slice. It is SLIGHTLY--under 30x magnification--more ragged at the start point than the feather. But it's definitely slicing not tearing even at the start point.
  ....

Thank you for sharing this, valuable findings.
Todd S. estimates Feather edge apex width to be around 50 nanometers, i.e. 0.05 micron - thanks to your tests, we know now that this is the minimum requirement to cross push-cut through the Rizla Green paper edge.
Before you posted this, I used to think it should be around 0.01-0.02 micron...

That's what I call the power of collective intelligence
With your permission, I update the chart with this new data.

 

[maximus83]

Sure go for it. Note that I tried this with a single blade on a single paper, so should be tested a few more times to confirm. But I'm telling ya', it's a night and day difference between the Feather and even my best knives. It just glides right into it, almost zero friction, and no training wheels starter cut.

Here's the specific Feather blades that I get, if that helps. ETA: I wonder if the 'platinum coating' on this specific blade impacts the results.