Steel Edge Retention testing

I love the tests Vassili,
But I don't see how they can be accurate without having identical knives made of the two steels being tested?
 
steelnut said:
I love the tests Vassili,
But I don't see how they can be accurate without having identical knives made of the two steels being tested?
Click to expand...

It is not knives tested but very edge which has same angle in all cases. So it does not show how good knife itself is, just how edge was affected by this load. And of course it is not steel, but stee heat treatment, and of course may be they will performs differently if angle are different.

For example blade length does not matter - I use only 1" of the blade.
Cross section geometry does not matter - I do not care how much forse I need to cut through, I am looking on what happen on very edge, which affected on the same way - is it flat, hollow or convex grind. Only thickness of the edge may I guess somehow affect this impact on the edge, but in this case when thickness is over 1mm and force applied not like one you have with chopping - I believe it does not matter.

Of course if you looking for full aspects of steel performance you need same knives with different steels, but there are no any!

Thanks, Vassili.
 
nozh2002 said:
Of course if you looking for full aspects of steel performance you need same knives with different steels, but there are no any!
Click to expand...

Spyderco Military, available in S30V, BG-42, and CPMD2 right now.

Benchmade mini-Griptilian, available in 154CM and D2 right now, and you could probably find it in 440C (S30V version has a different edge geometry).

Buck 110, available in 420HC, ATS-34, BG-42, and damascus right now.

Probably others as well.
 
Don M said:
Spyderco Military, available in S30V, BG-42, and CPMD2 right now.

Benchmade mini-Griptilian, available in 154CM and D2 right now, and you could probably find it in 440C (S30V version has a different edge geometry).

Buck 110, available in 420HC, ATS-34, BG-42, and damascus right now.

Probably others as well.
Click to expand...

And many of them I have...

However there is no Military in INFI and 420HC, etc. I do not have Mili CPM D2, but I have 110 in CPM 154 and CPM S30V. I am not aware of 110 in ATS-34.

Anyway - I am focused on edge retention of steel (heat treated by one or other manufacturers) - believe me this one task is more then enough for one person.

Again if you have in mind some testing... I will welcome your effort!

Thanks, Vassili.
 
hardheart said:
the tests have varied. I think the most common is slicing the rope and then push cutting another material to check loss of sharpness. I recall objections to the D2 slicing tests because it was being done by hand, the length of slice and amount of force applied was considered by some to be too variant for such a light media to cut.
Click to expand...

I wish more people would try the slicing sharpness test I did in that comparison to see what kind of results they get, or maybe venture out and try some of their own slicing sharpness tests. I know my results tracked well with the loss of cleaness of the rope cut, shaving, cutting newsprint, ect., but of course anything done by hand and eye like that is subject to human error. In fact, every portion of all of this testing is subject to human error, it is just a matter of how significant is that error. Doing multiple runs and measurements per sharpness point definately help to smooth things out, but there will always be human error involved. I like the push cutting sharpness testing with thread on scales, but I do think there is a lot of merit to testing the slicing sharpness also, as it is true that some edges that won't pushcut very well will slice very well. It just seems harder to come up with a consistent and repeatable test for the slicing sharpness.

Mike
 
gunmike1 said:
I wish more people would try the slicing sharpness test I did in that comparison to see what kind of results they get, or maybe venture out and try some of their own slicing sharpness tests. I know my results tracked well with the loss of cleaness of the rope cut, shaving, cutting newsprint, ect., but of course anything done by hand and eye like that is subject to human error. In fact, every portion of all of this testing is subject to human error, it is just a matter of how significant is that error. Doing multiple runs and measurements per sharpness point definately help to smooth things out, but there will always be human error involved. I like the push cutting sharpness testing with thread on scales, but I do think there is a lot of merit to testing the slicing sharpness also, as it is true that some edges that won't pushcut very well will slice very well. It just seems harder to come up with a consistent and repeatable test for the slicing sharpness.

Mike
Click to expand...

If you accept the idea that it is not only human error involved, but in major part random nature of steel grain structure and carbides dispersion, very degde geometry affected by abrasive and thread inconsistency then you understand that there is no other way to deal with this but by good number of tests and that statistical average.

Human error is just a little fraction of all random component of this picture, so we better deal with this then try to came up with something more precise (which very well may not exist at all).

I did try this while ago:

testing-06.jpg


It dramatically increases test time, but results have same dispersion! So do not worry of human error there are mny other sings out of our control and this is just the nature of this business.

Thanks, Vassili.
 
I agree human error is part of the process that may not be significant and can be brought to insignificant by doing lots of runs with an average, but my slicing sharpness testing was challenged as having too much human error. I did multiple runs with several measures of sharpness at each point I tested the sharpness at and averaged the results to try to smooth all of the numbers out, but some still thought my test wasn't precise enough. Your test is definately easier to get consistent results with than my slicing test was, but again I would like to see if anyone has come up with a good, consistent way of measuring slicing sharpness to go along with the very consistent thread testing you and others have done for push cutting sharpness. I really applaud your work, and it boggles my mind how many cuts you have already done for your testing. Even with gloves I was starting to get hot spots on my hands when doing many fewer cuts, along with the muscle and mental fatigue associated with doing so much cutting.

Mike
 
I think, what is really the point here is to quantify perceptible loss of sharpness and the thread and scale is capable of quite a bit more. It needs a very skilled hand to detect a difference in sharpness between an edge that measures 30g or 40g or between one that measures 120g or 130g (in fact I would suspect most would not be able to tell them apart) and only when they are sharpened to the same finish. So the scale offers already a precision that already exceeds the practical necessity.

I find it useless to call for greater precision if you can already not tell the results apart in practice anyway. My point is, if you can not tell in practice, without a scale, whether an edge dulled to 120g or 130g, because they both will feel the same to you when cutting something or when "feeling the edge", then why do you need a higher precision measurement? When you look at Nozh's data, you have to realize that those fluctuations in the data are well below what you would notice without thread and scale. Any knifenut will be able to tell a difference between an edge that comes in around 30 or 40ish and one that is 100+, but to get hung up on the 10g fluctuations is really not helping any, me thinks. The thread cutting test is surprisingly reliable and it is able to detect variations that are below what even a skill user would notice in regular use. Just look at Nozh's FF review. The threadcutting is easily capable of verifying the users impression of the edge.
 
If you order from the Buck site, you can geta custom 110 made with ATS-34 if you re interested Nozh.
 
HoB said:
I think, what is really the point here is to quantify perceptible loss of sharpness and the thread and scale is capable of quite a bit more. It needs a very skilled hand to detect a difference in sharpness between an edge that measures 30g or 40g or between one that measures 120g or 130g (in fact I would suspect most would not be able to tell them apart) and only when they are sharpened to the same finish. So the scale offers already a precision that already exceeds the practical necessity.

I find it useless to call for greater precision if you can already not tell the results apart in practice anyway. My point is, if you can not tell in practice, without a scale, whether an edge dulled to 120g or 130g, because they both will feel the same to you when cutting something or when "feeling the edge", then why do you need a higher precision measurement? When you look at Nozh's data, you have to realize that those fluctuations in the data are well below what you would notice without thread and scale. Any knifenut will be able to tell a difference between an edge that comes in around 30 or 40ish and one that is 100+, but to get hung up on the 10g fluctuations is really not helping any, me thinks. The thread cutting test is surprisingly reliable and it is able to detect variations that are below what even a skill user would notice in regular use. Just look at Nozh's FF review. The threadcutting is easily capable of verifying the users impression of the edge.
Click to expand...

I agree 100%. When I did my testing all of my usual tests like shaving and newsprint cutting, and the cleaness of the rope being cut, it showed the D2 and CPM D2 knives were virtually identical in sharpness and edge retention, and my twine slicing backed up that impression. I got criticized (as I should have, when you put out testing it should be looked at critically by others) about my twine slicing test and whether it was accurate enough to even use as a valid test for sharpness. Since the twine testing was so similar to my real world impressions I feel it is valid, but I would like to see some other slicing sharpness testing to back up or refute what I did. The scale testing on the other hand is pretty well accepted as being very accurate, at least in my mind, so my comments about more precision were mainly aimed at a slicing sharpness test, not Nozh's work. Either way, what happened with my knives appears to be exactly what is happening here, the knives are at a point where after a lot of cuts you would be very hard pressed to tell the difference in sharpness without a scale (with the INFI and 420 HC, at least).

Mike
 
I expand tests to 800 cuts for INFI and just finish with 420HC after 700 cuts.

800 is twice from initial test. It is clearly necessary because all beauty of INFI appear only after 250 cuts and to 800 cuts it's superiority for hard work over 420HC became absolutely clear.

420HC shows pretty quick degradation and from around 450 cuts just fall down, while I really expected more flat curve of steel loosing sharpness.

It is absolutely clear that dulling curve really do not follow any pattern - edge is to complicated subject to be fall into simple modeling.

Again it will be to simplistic to say that one is better then other, however it is clear that 420HC hold high edge better on small load, while INFI hold good edge (edge which average customer probably puts on his blade using average sharpening tools, edge which many manufacturers initially put on their knives) under heavy load.

Thanks, Vassili.

P.S. I stop with 420HC at 700 cuts because it is too dull - 260g. Range is from 210 to 320 - I had to do twice tests 41 instead of 21 to verify result. It is also harder to do testing and I guess pointless. I think I should stop cutting after I get over 200g for several test in the row.
 
what do you mean by hard work and heavy load, isn't the rope cutting with about the same amount of force since the thread is?
 
hardheart said:
what do you mean by hard work and heavy load, isn't the rope cutting with about the same amount of force since the thread is?
Click to expand...

Sorry, I may be do not understand you correctly, but amount of force to cut 1/2" manila rope is way bigger then amount of force to cut "Classic Crochet Thread" - mercerized cotton thread #1.

I think it is a good load and 800 cuts for sure heavy load. Here rope cutting:
http://www.youtube.com/watch?v=3LgFny0ZQEU

And here thread cutting:
http://www.youtube.com/watch?v=MsN3DghiYcw

You may see the difference.

Manila rope is used for knife testing by many considering it to be good load on the edge. ABS competitions and Phil Wilson tests:

http://www.bladeforums.com/forums/showthread.php?t=421773

And if you see he has 5/8" manila rope (almost 1/2"). In his tests number of cuts way lower probably because he has Alder base which most likely makes biggest part of dulling. Plus his 20 pounds limit not quite clear for me because it may vary a lot from cut to cut.

But for sure he consider manila rope cutting as a close emulation of feild experience.

Thanks, Vassili.
 
I mean that since the thread is still being cut at 130g, that means the knife is still plenty sharp, so the rope should still be cut at lower force. I can see how the 420HC at 260g on the thread would mean a lot of force to cut through the rope, but the INFI only went from 110g to 130g.
 
hardheart said:
I mean that since the thread is still being cut at 130g, that means the knife is still plenty sharp, so the rope should still be cut at lower force. I can see how the 420HC at 260g on the thread would mean a lot of force to cut through the rope, but the INFI only went from 110g to 130g.
Click to expand...

OK I understand now. I meant "load" in terms of how many rope it cuts, not how much pressure I put on cutting. 800 cuts is hard work and heavy load at least to my wrist and elbow... It hurts!

Thanks, Vassili.
 
nozh2002 said:
800 cuts is hard work and heavy load at least to my wrist and elbow... It hurts!

Thanks, Vassili.
Click to expand...

You are going to have carpal tunnel after doing so much cutting! I really applaud the work you are doing, my arms would fall off after that much cutting, and it is also takes a significant amount of time to do your sharpness testing.

I used the same 5/8" Manilla rope that Phil Wilson used in my testing, and I can verify that it does indeed start to dull a knife out quickly. Phil stops at a point where he feels most guys would resharpen their knives, I believe, though he probably has a higher standard of sharpness than most. I stopped my personal testing at 64 cuts over a 2" section of edge because the knives were starting to tear a bit when slicing newsprint, which to me is the point where a knife is really in need of a sharpening (plus I didn't have time to go any further). I only went to a DMT fine finish, not intending to go for optimal number of cuts, but merely to compare the steels in a limited amount of time. I also did a pure slicing motion (heel to tip), and my poly cutting board background was a major factor in dulling (if not THE major factor). However, if I handed the knives at that point to the average guy they probably would keep on cutting without complaint for quite a while considering the blades were D2 and CPM D2, which are known to slice a really long time even when "dull". I guess that is the curse of using sharp knives, you get to where you can't stand to use dull or even average sharpness knives after a while.

Mike
 
I think I finish all three steel. I think it will be pointless to do more cuts - too much work. I need to find different media which will affect edge more then manila rope with less work for tester (me). May be I should not even go after let say 500 with manila rope, but resharpen knife and start again with something else.

But, I will continue with different steels on manila rope updating that table. I found this very interesting and useful.

Thanks, Vassili.
 
I have found that cardboard will fairly quickly dull an edge. I think it would be interesting to find out if different things will give different dulling results. Good idea to finish one test and then try another.
 
You must log in or register to reply here.
Back
Top