New testing session.

nozh2002

nozh2002

BANNED
Joined
Jun 9, 2003
Messages
5,736
After 6 months from last testing:

http://www.bladeforums.com/forums/showthread.php?t=509097

I kind of rethink it a little bit and finally started new session two days ago. From last time I change - set up clear procedure for how to cut thread to avoid any influence from myself. Now I am cutting thread applying force step by step from one point on scale to another waiting few seconds (counting to 5) and move to another point right away.

Another change - new postal scale which is more precise (I hope) in oz.

And I reduce number of measurements and decided to stop at 200 cuts - it will continue to cut rope even more but there is no too much point on doing this to me. It take me a day to do full test circle with one knife. So this is much better then a week it takes before.

http://playground.sun.com/~vasya/Manila-Rope-Results.html

Thanks, Vassili.
 
Last edited:
Interesting results. I'm surprised the S30V and ZDP189 blades gave such close measurements.
 
Vivi said:
Interesting results. I'm surprised the S30V and ZDP189 blades gave such close measurements.
Click to expand...

If you consider all numbers - ZDP189 is quite ahead until last 200 cuts.
What surprise me - SG2, I think I will test 3G next.

Thanks, Vassili.
 
Do you think the tension on the string on the piece of would should be consistent? Is it? Perhaps I'm missing something. I just wonder if there needs to be a way to ensure the same tension level on that string.
 
Reeek said:
Do you think the tension on the string on the piece of would should be consistent? Is it? Perhaps I'm missing something. I just wonder if there needs to be a way to ensure the same tension level on that string.
Click to expand...

Tension play significant role in results. You may try this by holding thread in your hand having knive placed on the scales. In my set up tension is same, because the way it fixed on the stand - it adjust itself if it is too tense.

Thanks, Vassili.
 
Nice work, Vassili. I know from personal experience how much work is involved in doing testing like this, and thank you for your efforts and sharing this. Some of the results are surprising ... but then I've seen strange things, too, like the way some blades will hit a certain level of sharpness and just stay there, cut after cut, for a long time.
 
Sorry for the stupid question, but how must we read these stats? I mean what must we take out from the tests...

Is the number of cuts that occurred before the knife went blunt?

Thanks
 
Buffalohump said:
Sorry for the stupid question, but how must we read these stats? I mean what must we take out from the tests...

Is the number of cuts that occurred before the knife went blunt?

Thanks
Click to expand...

What I am doing

1. Sharpen knife to hair whittling sharpness. Mark place on the edge and make sure it whittle hair there. Measure sharpness level using my statistical thread sharpness test (see below) - this will be '000' row.
2. Cut 1/2" Manila rope by the market part of the edge. Measure sharpness - it will be row '001'.
3. Cut two times, measure sharpness - it will be row '003'
...

So more weight you need to cut thread - less sharp edge. So ZDP189 after 200 cuts has 080 (8 oz) and so more sharp then Sandvic 13C26 which has 105 (10.5 oz).

But this test shows dynamic of edge changes which is more interesting. Like Friction Forged D2 at 20 cuts less sharp then D2 - 80 vs 65. But it hold it up to the 200 where difference is 85 vs 120!


Sharpness test. I cut cotton thread places on the scale (spring postal scales from Office Max) and measure weight when cut happen. It has scale in 1/2 oz. So 10 in my table is 1 oz and 15 is 1.5 oz. There are many factors which make this measurement vary, so to have some solid number I make 21 measurement and then take median - as an average result.

I am working on it probably for 4 years now - if you follow the links, you may see when I first came here with this sharpness test. Now I am pretty confident in what I am doing.

Thanks, Vassili.
 
I'm sure you have it down pat, I was just asking how to read the table and understand the information it contains.

But I must admit I have owned knives from some of these manufacturers and I was less than impressed with the edge they held.

However, I have custom made knives in the same steel and it performs a lot better.
 
nozh2002 - Nice to see someone else's test confirm my own experience with
S30V. It received alot of bashing awhile back, from one person in particular, based on what I thought was pretty limited testing in that great conclusions were drawn from work with one knife.

S30V in its current composition (it was originally 3% vanadium, and is now 4%), and with folks figuring out how to grind and heat treat it, it makes a dang good knife blade. I recently used a Siegle custom in S30V while out river fishing. When confronted with a log jam, I had only this little 3 3/4 inch bladed knife to clear a path to jump the jam with our jet boat. I used a baton to clear branches, some three inches in diameter, and the blade chewed right through knots and joints. After repeated pounding and scraping, the blade was no worse for the wear, and still very sharp. A couple passes with a Ezelap restored it to hair popping sharp in no time. This is good stuff.
 
Hmmm, D2 is the best and the worst, that seems a bit odd and if it is to be believed than perhaps the composition of the underlying steel is of much less importance than the heat treat.
 
Very cool test Vassili.

Very interesting how much a difference there is in the d2.
I have a Kershaw outcast in d2 that preforms like 1095; easy to sharpen, chip resistant, and springy.
I also have a small d2 skinner that is hard to sharpen but holds an edge.
 
theonew said:
Hmmm, D2 is the best and the worst, that seems a bit odd and if it is to be believed than perhaps the composition of the underlying steel is of much less importance than the heat treat.
Click to expand...

I agree 100%. Fortunately he was wise enough to list not only the steels being tested but also the specific knives and manufacturers. Because I believe the HT is very important; if only the steel had been listed the results would have little value IMO. It would be interesting to put Buck's 110 in 420HC with the Pual Bos HT to the test.

Thanks for posting these results! It seems like a well designed test to me.
 
Looking at the graph generates a question in my mind. How come performance improved with additional cuts? While the tendency overall was for increasing thread cut weight as more cuts were made, in many instances the thread cut weight actually went down with more cuts before going back up. This would seem to indicate imprecision in the measurement process from one measurement point to another. Enough imprecision in fact to make me question the overall results, especially where the results were fairly close.

For example, S30V went down between 60 and 80 cuts, and then back up. Again between 150 and 200. This means it could go down again between 200 and 250 and actually rate better than say R3V or 10V. Any ideas about this anyone?
 
Steelhed said:
How come performance improved with additional cuts? While the tendency overall was for increasing thread cut weight as more cuts were made, in many instances the thread cut weight actually went down with more cuts before going back up.
Click to expand...

Though others have claimed that this makes sense, I too find it very suspicious. While I think Vassili tries very hard to keep his tests accurate, I don't have tremendous confidence in the results. Things like CRKT's AUS8 being ahead of BM's D2 strike me as indicative of errors in the process. Rather than continuing to test every steel and knife created, I would much rather see him repeat the test on the exact same knife a few times to get a sense of just how divergent the results can be.
 
Yeah, it doesn't make sense to me either. If it isn't measurement imprecision, it could be variability in the cutting medium, or some other variable including the thread itself.
 
Well, I saw this changes many times, actually few years ago I first notice this and discussed here, and I did several retests in such cases, but it shows same results.

We do not really know what is happening on the edge when it cutting manila rope and it is easy to cam up with one or another explanations - like some bumps causing weaker results got straighten, or some carbides first emerge but then fall out - who really knows?

But in opposite process - sharpening and polishing which is less brutal to edge then dulling it using manila rope, during this more gentle and more controlled process measurements always show less and less weight as expexcted. So here I do not have such surprises with measurements.

So I consider the fact that sometimes edge got little sharper after use as a natural evolution of the edge itself, rather then measurement in-perfection.

I think too that it is better to have not one but many tests for same blade to be more accurate, but this is a lot of work to do - anybody like to join this effort? It will be much better to have alternative independent testing then me do this again and again.

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