Testing for sharpness?

D

dawgvet

Joined
Jan 25, 2007
Messages
45
I am now progressing in my sharpening skills and would like to develop a consistent, reliable way to test the final sharpness of each blade. In other words, some objective way to tell when I'm "done" with a blade. Most blades would pop hair when finished with Ultra Fine Lansky stone, then started stropping on green chromium oxide and that definitely helped. Just don't know where to go from here. Sharpening has turned into a bit of a hobby in itself.
Thanks,
Jed
 
Well since you are polishing your edges, pushcutting sharpness seems the most sensible measurement for you edges. Far and away the most reliable method for that is IMHO the thread cutting test. A simple setup I have described here:
http://www.bladeforums.com/forums/showthread.php?t=399057

I will add that this is not the only way of measuring sharpness. The newprint paper cutting is also a method which I am not too fond of, since I think that you can cheat too easily by putting tension on the paper with the holding hand or by pulling the knife just a tad, or tilting the knife. For pull-cutting ability, thread cutting is not a sensible measurement. Far better is rope cutting on a scale: Place a cutting board on a scale (a larger one than you would use for thread cutting) and see at what pressure the rope is not cut on a slice. Or mark a distance on the edge and see how far you have to put the rope to cut through it, using a constant weight.
 
My sharpness test is I'll cut off a strip of paper towel about 2 inches from the end hanging off the roll. If it cleanly cuts that off I tightly roll it and then slice that like a small piece of rope. If it cleanlly does both that is the edge I wanted.
 
I have found a couple of things one is push cutting 550 para cord the other is a Styrofoam cup with both you want crisp clean cuts.The Styrofoam seems the be the more stringent test it take a VERY sharp blade to push cut cleanly but it also depends to a certain degree on blade geometry so it works best with smaller thinner blades.
 
dawgvet said:
I am now progressing in my sharpening skills and would like to develop a consistent, reliable way to test the final sharpness of each blade. In other words, some objective way to tell when I'm "done" with a blade. Most blades would pop hair when finished with Ultra Fine Lansky stone, then started stropping on green chromium oxide and that definitely helped. Just don't know where to go from here. Sharpening has turned into a bit of a hobby in itself.
Thanks,
Jed
Click to expand...

This is simpliest way to test final sharpness:

http://www.youtube.com/profile?user=nozh2002

I also develop some thread cutting tests on the weight - it is too not precise so I have to use statistical measurement on 21 cuts I take median.

Sharpness for hair splitting - 40g.
Best out of the box sharpness - 60g by Microtech and Spyderco
Usual sharpness out of the box - around 100g.
 
HoB said:
Well since you are polishing your edges, pushcutting sharpness seems the most sensible measurement for you edges. Far and away the most reliable method for that is IMHO the thread cutting test. A simple setup I have described here:
http://www.bladeforums.com/forums/showthread.php?t=399057

I will add that this is not the only way of measuring sharpness. The newprint paper cutting is also a method which I am not too fond of, since I think that you can cheat too easily by putting tension on the paper with the holding hand or by pulling the knife just a tad, or tilting the knife. For pull-cutting ability, thread cutting is not a sensible measurement. Far better is rope cutting on a scale: Place a cutting board on a scale (a larger one than you would use for thread cutting) and see at what pressure the rope is not cut on a slice. Or mark a distance on the edge and see how far you have to put the rope to cut through it, using a constant weight.
Click to expand...

Somehow I missed your thread HoB.

From my first attempt I develop new gadget to hold thread. And this is how to fix it using scotch tape:

thread-01.jpg


thread-02.jpg


thread-03.jpg


thread-04.jpg


thread-05.jpg


thread-06.jpg


And also I do 21 cuts as minimum to get reliable result which is median. I am very confident in it and now test each new knife to see out of the box.

Right last week I finally started steel testing (INFI) because I now can sharpen knife to stable sharpness and precisely measure sharpness after cutting manila rope...

Thanks, Vassili.
 
Hi, Vassili, nice piece of work there, as always! I like the wingnuts though. It is very easy to clamp the thread.

I am almost get the impression you are superstitious :D, why 21 cuts and not 20 cuts, is 21 the magic number? :D

I agree by the way, that you have to make several cuts and also test along the edge to see whether you missed honing near the choil or at the tip. I never really count my cuts and do a statistical average. I usually take around 10-15 cuts per section of the edge and see what ballpark the numbers come in and then take the middle of the ballpark. That is precise enough for my purposes and I can do it in my head. On the Rayon thread I am looking for at least 5 g of difference. Whether a blade comes in at 20 or 23 g doesn't really matter to me, both is bloody damn sharp. Even at 30-40 g you are in the top level of sharpness achievable with typical folding knife type edges, 40-50 is still quite nice, 50-60 g is servicable as long as you don't do high precision work and 60+ means it is time for a re-sharpening. So all I really care is what "grade" an edge makes: 20-30 (A), 30-40 (B), 40-50 (C), 50-60 (D), 60+ (F). This is of course only my classification. If I wouldn't be on a knife forum, most people would probably still be satisfied with a 100+ g edge.
 
I picked up some thread the other day from Wally world. Coats & Clark, cotton hand quilting, glace(?) finish. I also got an inexpensive 35 lb digital scale, figure I can use it for rope slicing as well. I was trying to figure out a good way to hang a 105g piece of wood I had lying around for tension, but if you guys don't use a weight, maybe I'll skip it. Think I'll pick up some disposable Bics too, maybe we can get some baseline going, lots of difference in measurements going from thread to thread.
 
Just wave it over a ripe tomato. If the tomato screams and divides itself into even slices, it's sharp enough.
 
Hey Jed,

I have several quick tests to determine when I'm satisfied.

One is to do what Chiller2 does: use a length of 550 over a block of wood (maple, mahog., whatever). At various points along the edge, push straight down. After awhile, you'll begin to know when it's sharp by the sound, quality of cut, and pressure needed to push-cut through the paracord.

Two is to take a rectangular piece of 7-8oz leather - about 6" long by, oh, 1/2" wide. Stand it on it's edge (not end, but edge) and, again at various points along the edge, push cut straight down, slicing off 1/16" to 1/8" slices of the leather. This quickly becomes wasteful, I assure you, but it truly will show the difference between sharp, very sharp, and sharp!. My Bravo-1 and Northstar will slice like a knife going through butter. The mini-Canadian will slice like a hot knife going through butter!!! If there is any portion of the edge that needs work, you'll know it right away.

Then there's the sharpness by feel method. I'll hold the knife - edge up. With my thumb, I'll touch the edge going slightly downward and away....using only the slightest pressure. A very sharp edge immediately finds and grabs a fingerprint ridge. It takes practice to find just the right angle and pressure, but shortly you can very quickly ( in three or four seconds) test both sides of the edge all along the blade - finding any "weak" spots.

Last, but not least, there are the arm shaving and the hair-popping methods.

All of these in combination are what I use.
 
Pushcutting newsprint is a fast and consistent test for sharpness, most suited to fine polished edges. I prefer it because it's easy to use, and I can test an edge much more quickly than I can using a medium like thread that has to be tensioned, and applied cutting force weighed or otherwise measured.

As HoB points out, results will vary if your technique isn't consistent, but then you'd only be cheating yourself. IMO the best way to do the test is to compare two or more knives side-by-side, which helps you maintain consistent technique, and of course if you use a knife of known sharpness for comparison, it will be very easy to tell when you've done a good job sharpening. The knife that's able to consistently start a pushcut furthest away from the other end of the paper where it's being held is the sharpest; simple as that.

Bear in mind with any test of initial sharpness, actual measurements taken are only meaningful for the quality of edge you're trying to achieve. A coarser edge, better suited to slicing in many instances, isn't going to cut newsprint, or thread for that matter, as readily as a finer, polished edge. However if you're comparing two blades sharpened to the same level of finish, then the results are meaningful: even with a coarse edge, better push cutting performance means a better edge was attained.

BTW, although it's not a quantified way of testing, I've noticed the same thing that Stretch discusses, above, when cutting leather, which is very sensitive to edge sharpness. Another test is pushcutting styrofoam plastic packing peanuts, an edge has to be quite sharp and have decent polish and consistency to really cut them clean and with no squishing.
 
HoB said:
Hi, Vassili, nice piece of work there, as always! I like the wingnuts though. It is very easy to clamp the thread.

I am almost get the impression you are superstitious :D, why 21 cuts and not 20 cuts, is 21 the magic number? :D
Click to expand...

This is just because median will be eleventh from the start in sorted list of results:

90 +
100 ++++
120 ++++
130 +X++++
140 ++++
150 +
160 +

X - median.

I just cut thread around same point 21 times marking in the table results, then just from the top count 11th. You may see that dispersion even around single point on the blade is quite big, so good statistic is nessesary. Otherwise it may vary from 90 to 160! On this example it looks quite Gaussian - nice hill like chart. If it is not like this I just do more cuts and fill table more, like 41 (to count to 21st).

This is Busse GW after 400 cuts of 1/2" manila rope (I almost ingure myself doing this!). Started from

30 ++++
40 ++++++X++
50 ++++++++

I'll post complete results later. I finished I think with Busse and go for CS ODA with 420 steel, as, possible, an other side of the range.

I use different weight then you are and different thread - cotton. So my results should be different that what you have.
On my scale -
40 - hair splitting sarpness (my hair)
130 - bad out of the box sharpness

Thanks, Vassili.
 
hmm, just sharpened a knife and gave it a shot. Blade wasn't tree topping, but I put on a pretty typical edge for 10 minutes effort.

anywho, this is what I got, in order, it's really all over the place

40 70 60 45 100 75 80 100 90 70 70 105 105 90 95 80 100

by this point, I was wondering why the numbers were creeping up, so I did 4-5 passes/side on the leg of my jeans (no abrasives, just denim), here's the last 5 cuts

45 55 65 70 65

it's like the edge noticeably dulled/damaged and was brought back. A couple of the triple digit measures were cuts near the tip, where I wasn't completely happy with the edge just from some arm hair shaving before stropping & the thread cutting. A few times when I cut through the thread, I also lightly tapped the wooden mounting block, which must have affected the edge more than I thought it would.
 
hardheart said:
hmm, just sharpened a knife and gave it a shot. Blade wasn't tree topping, but I put on a pretty typical edge for 10 minutes effort.

anywho, this is what I got, in order, it's really all over the place

40 70 60 45 100 75 80 100 90 70 70 105 105 90 95 80 100

by this point, I was wondering why the numbers were creeping up, so I did 4-5 passes/side on the leg of my jeans (no abrasives, just denim), here's the last 5 cuts

45 55 65 70 65

it's like the edge noticeably dulled/damaged and was brought back. A couple of the triple digit measures were cuts near the tip, where I wasn't completely happy with the edge just from some arm hair shaving before stropping & the thread cutting. A few times when I cut through the thread, I also lightly tapped the wooden mounting block, which must have affected the edge more than I thought it would.
Click to expand...
Sounds like you had a bit of a burr which stropping on the jeans cleaned up.
 
hardheart said:
hmm, just sharpened a knife and gave it a shot. Blade wasn't tree topping, but I put on a pretty typical edge for 10 minutes effort.

anywho, this is what I got, in order, it's really all over the place

40 70 60 45 100 75 80 100 90 70 70 105 105 90 95 80 100

by this point, I was wondering why the numbers were creeping up, so I did 4-5 passes/side on the leg of my jeans (no abrasives, just denim), here's the last 5 cuts

45 55 65 70 65

it's like the edge noticeably dulled/damaged and was brought back. A couple of the triple digit measures were cuts near the tip, where I wasn't completely happy with the edge just from some arm hair shaving before stropping & the thread cutting. A few times when I cut through the thread, I also lightly tapped the wooden mounting block, which must have affected the edge more than I thought it would.
Click to expand...

Well, on the first run, I would simply discount the 40ies and call the mean 85 (this is btw only by looking at it without any statistical analysis). My numbers are a lot lower and my results are usually a lot closer together. I think you are using thread that is too thick. If you read this thread http://www.bladeforums.com/forums/sh...d.php?t=399057 it is quite important to use very thin thread, otherwise the numbers will become very dependent on how you hold the blade: If you tilt it slightly, you increase the chances of denting the edge, especially if your edge is thin. Also, I would not group measurements taken at different parts of the edge into one. Instead I would measure the tip, the middle and the heel. Otherwise you get a totally screw idea of your edge if the tip or the heel is not overly sharp. Also you are missing bad technique, if you are, for example always neglecting the tip for example.

So I got curious and calculated the average with out the 40ies and it is 86, so my guess of 85 wasn't bad at all :).
 
I came to conclusion that average does not really represent sharpness, because of random nature of results. I use median instead and will take care of picks in one or other direction. Median - is result taken from the middle of the list of results sorted by values - in my case it is eleventh result marked by X on the table.

If you are measuring results of sharpening - tests range will be narrow because during sharpening effect on the edge will be even all over it length.

Busse GW hair splitting

30 ++++
40 ++++++X++
50 ++++++++

In ideal case results should show Gauss chart - look like hill.

Bob Dozier KS7 out of the box:
50 +
60 ++++++
70 +++X+++++
80 ++++
90 +
Pretty good sharpness

Measurement always looks this way - law of nature, and we are looking for maximum value - top of the hill. Actually, if you results looks this way - this mean that you are doing right thing.

Busse AD out of the box

70 +
80 +++
90 +++
100 +++X
110 ++++++
120 +++
130 +

Average sharpness

Busse AD after my sharpening - hair splitting

30 ++++
40 ++++++X++
50 +++++++
60
70 +

And it really looks like this when quality of the edge is even all over it - right after sharpening. If you edge not ideal than pattern of results will be different. If you have chips on the blade then it looks like two independent Gauss hills one for good edge and one in a different range with much bigger results.

Similar thing happen if you measure sharpnes after blade cut something. I found that even when I do measurements in around same point of the edge I may have two headed hill in result, which mean that edge is not even there, one part sharper then another. Likeon this example two maximums 100 and 130:

Busse GW after 230 cuts of 1/2" manila rope

80 +
90 +
100 +++++
110 +++X
120 +++
130 ++++
140 ++
150 +

Thanks, Vassili.
 
I stropped 50 passes/side after 8000 King, so I hope there was no burr. There's a thumb stud in the way at the heel, and I didn't pay as much attention to the tip as I usually do (in a hurry to try this)

is this what you mean, Vassili?

40 +++
50+
60+++
70+++X+
80++
90+++
100+++++

I guess the tip is the duller peak. Gonna get some thinner thread at a quilting shop nearby, and test the edge in segments. Think I'll also get some disposable Bics as ghost squire suggested.
 
hardheart said:
I stropped 50 passes/side after 8000 King, so I hope there was no burr. There's a thumb stud in the way at the heel, and I didn't pay as much attention to the tip as I usually do (in a hurry to try this)

is this what you mean, Vassili?

40 +++
50 +
60 +++
70 +++X+
80 ++
90 +++
100 +++++

I guess the tip is the duller peak. Gonna get some thinner thread at a quilting shop nearby, and test the edge in segments. Think I'll also get some disposable Bics as ghost squire suggested.
Click to expand...

Is it around one point of the edge? I usually mark some point with marker and try to test around it. If it is from different parts of the edge it may be anything. In general it will be some random sum of different Gauss. Like in your case it may be three of them one for tip + one for middle + one near the handle.

Excluding 40 and 100 it is pretty good Gauss curve. But for some reason you have two maximums at 40 and 100. In such case I increase number of cuts - only way to be more certain - increase number of tests.

Thanks, Vassili.
 
You must log in or register to reply here.
Back
Top