Ranking knife brands by CATRA testing

wootzblade

wootzblade

Joined
Feb 24, 2014
Messages
240
Australian Knife Magazine
is embarking on our most ambitious article to date! A community-based project led by Knife Grinders and Australian Knife Magazine aiming to rank the cutting performance of World's leading knife brands by means of independent CATRA testing.
 
madcap_magician said:
I would think that this test would be more heavily influenced by blade geometry of the test knives than it would by the brand of the knife.
Click to expand...

I fully agree. Maybe they could just try to find a similar pattern blade in several manufacturers at the time. But will need to insure the same edge angle, sharpening grit, behind the edge thicknes, etc... and how about the steel?

Question is... what are you trying to compare? Same steel/knife geometry, etc... so you can compare the heat treating? Same knife geometry but different steels? Then you would be comparing steels, not manufacturers.

What is this all about?

Mikel
 
I agree with the above comments. This will go off the rails quickly if they attempt to compare different thickness and geometries in the same steel. I can see some lower hardness steel with better geometry outperforming harder and even more wear resistant steels. We have seen it before. What would be the point of this article? It wouldn't prove the superiority of anything but geometry and sharpening technique and quality.

Why do you think Larrin and some others use test coupons? To eliminate the very issues in this article.
 
This is somewhat silly. How could you rank knife brands just based on a few models performance on a carta test?

If BMW produced a bicycle, can you test how fast that bicycle go and rank BMW cars as slow? Even cars, I'm sure different make will perform very differently, and you can't just 'rank' a car company based on selective models.
 
  • Like
Reactions: DMG
I think this is great. I've been experimenting with some of the unused knives in my collection using a goniometer, BTE measurement, and weighted thread cutting and from my preliminary results it's obvious that some brands come sharper than others and have geometry more conducive to edge retention.

It would be interesting to see how stock edges stack up. If one brand comes sharper and stays sharper longer, isn't that the sharper knife - irrespective of geometry?

We might want to equalize everything in the lab to get robust results, but the real world doesn't work that way. If you give a guy knife x and knife y, he's not going to regrind them to the same angle and profile to make sure his opinion is formed by repeatable statistical method. He's going to use them and notice how long it takes each to get dull, and form opinions over time. Which was the process for 99% of knife reviews up until about 5-6 years ago.

Here's a direct link to the first report, took an excessive amount of clicks to get there from the original post: http://knifegrinders.com.au/CATRA-report.htm

I've worked in professional kitchens for 20 years and it is zero surprise to see Global sitting at the top. I would have expected Henckels and Wusthof to be in opposite places but they are similar enough I suppose.

Regardless, I will be checking the results as the project continues. It is already proving interesting and I'm hopeful it will generate some good discussion.
 
Last edited:
spoonrobot said:
It would be interesting to see how stock edges stack up. If one brand comes sharper and stays sharper longer, isn't that the sharper knife - irrespective of geometry?
Click to expand...

If all they did was to compare specific model of different knife company, and rank those stock knives sharpness based on the result, surely I think that's a fair test.

But what I think is borderline silliness is we don't even know what knife they choose from each brand, and ranking companies(not knives) on those results alone? That's not exactly a fair test, and I would think a blade magazine would know better?

The geometry of the knife definitely will play a part in sharpness and edge retention. Think about how a Chinese cleaver vs a chisel grinded Santoku, that is not exactly what I would call apple-to-apple test.
 
I also don't see where multiple offerings were tested to account for one-off issues.

I have/had several Spyderco Delicas. I got one with a really bad grind but all the others were fine. If that one bad one was used in testing as representative of the entire line, let alone the brand, it would be irresponsible.
 
  • Like
Reactions: DMG
It goes without saying that we compare knives of the same function.
Project #1 is for Kitchen & butcher knives.

For kitchen knives, 2 staple Chef's knives of a given brand are tested, and the numbers are averaged. Where a maker produces several lines of knives, each line is tested.
Example:
GIESSER produces 3 staple lines of knives: PremiumCut, BestCut and PrimeLine.
For the Giesser brand we test 2 Chef's knives of each line, 6 knives total.

As we get results for Chef's knives for 10-20 brands, by comparing the test results we get an objective ranking of knife brands.
We will honestly name those who use the best steel for Chef's knives, heat treat it the right way, design the blade and cutting edge geometry for the optimal performance, and sharpen properly.

The first provisional report on our website lists Chef's knives.

For butcher knives we plan to test boning knives of the 4 brands used in meat industry. They all are of the similar geometry.

We could do CATRA testing ourselves, but these results will draw adverse criticism. Testing must be done by an independent party - that is why we contract CATRA. Everyone can dispute results that we obtain, but no one can dispute results obtained by CATRA themeslves.
 
There's a couple of logical problems here. Unless you can eliminate the variability between identical specimens of the same brand, you'd have to test a very large number of the same model of knife. That will then give you a baseline nominal performance curve. The number of specimens required to achieve a normal performance distribution isn't known until you see the variability among the population of the model tested. It's been a while since I had to pass statistics, but this is all pretty basic.

In the case of a model with high variability from specimen to specimen, you may need to test 50 or more specimens to achieve an accurate representation of typical cutting performance.

What about different lot runs of the same model knife? What about multiple manufacturing locations? What about the human effect on hand finished/assembled knives?

MAYBE then you could compare the nominal performance curves of one model of knife to another model of knife. At that point you could rank models, not manufacturers. I don't see how you can ever rank manufacturers using a method like this. You may be able to say that Brand A's Boning knife cuts better than Brand B's similar Boning Knife, but that leads nowhere into an overall ranking of the brand.

You might want to consult with a researcher about steps you can take on the frontside to make this effort have more meaning. The data, as you're proposing to gather it, can only lead to some invalid assumptions.
 
Hackenslash said:
There's a couple of logical problems here. Unless you can eliminate the variability between identical specimens of the same brand, you'd have to test a very large number of the same model of knife. That will then give you a baseline nominal performance curve. The number of specimens required to achieve a normal performance distribution isn't known until you see the variability among the population of the model tested. It's been a while since I had to pass statistics, but this is all pretty basic.

In the case of a model with high variability from specimen to specimen, you may need to test 50 or more specimens to achieve an accurate representation of typical cutting performance.

What about different lot runs of the same model knife? What about multiple manufacturing locations? What about the human effect on hand finished/assembled knives?

MAYBE then you could compare the nominal performance curves of one model of knife to another model of knife. At that point you could rank models, not manufacturers. I don't see how you can ever rank manufacturers using a method like this. You may be able to say that Brand A's Boning knife cuts better than Brand B's similar Boning Knife, but that leads nowhere into an overall ranking of the brand.

You might want to consult with a researcher about steps you can take on the frontside to make this effort have more meaning. The data, as you're proposing to gather it, can only lead to some invalid assumptions.
Click to expand...

All that are the knife manufacturer's tasks.
When end-user buys a knife, he expects the standard quality in each knife he buys.
We sample 2 Chef's knives of a given brand and put the averaged number in the report.
If the knife manufacturer cannot provide for standard quality, there is a problem with this manufacturer.

When I look at the list of owners of the CATRA tester, I see that the major knife makers have it. I believe they use it to control quality in the batches of knives they produce.
They've tested their own and competitors' knives but will not disclose the information because it is business sensitive.
However, there is nothing preventing public from doing this testing and revealing the facts.
 
Last edited:
I think you missed the meaning of my entire post but good luck anyway.
 
DTN27wT.png


Just factory edge testing?

wootzblade said:
All that are the knife manufacturer's tasks.
When end-user buys a knife, he expects the standard quality in each knife he buys.
We sample 2 Chef's knives of a given brand and put the averaged number in the report.
If the knife manufacturer cannot provide for standard quality, there is a problem with this manufacturer.

When I look at the list of owners of the CATRA tester, I see that the major knife makers have it. I believe they use it to control quality in the batches of knives they produce.
They've tested their own and competitors' knives but will not disclose the information because it is business sensitive.
However, there is nothing preventing public from doing this testing and revealing the facts.
Click to expand...
wootzblade said:
Australian Knife Magazine
is embarking on our most ambitious article to date! A community-based project led by Knife Grinders and Australian Knife Magazine aiming to rank the cutting performance of World's leading knife brands by means of independent CATRA testing.
Click to expand...
 
Yes the factory edge, because we want to check quality of the sharpening process the knife maker employs.

We use for that the CATRA ICP Initial Cutting Performance number - indicator of the edge "softening" by overheating the edge by agressive dry power sharpening techniques. ICP indicates quality of the sharpening process.

If ICP is bad, the buyer of that knife will have to manually re-sharpen it 6 times till he gets to the clean steel.

Having obtained the ICP number, we continue the CATRA test, and get the Knife Life TCC - a trusted indicator of the overall knife cutting performance and edge retention. This one depends on the steel, heat treatment and the edge geometry.
 
Last edited:
For tangible donations to the CATRA testing, we ship a Kangaroo tail as our thankyou. You will receive a processed Kangaroo tail, ready to cut out a strop.
We can even cut out a hanging strop for you, and you can assemble it yourself.
Details on our website.
Read more about the Kangaroo tail >>
 
Last edited:
This is my first post as I joined to learn about sharpening my Globals I just got (more on the way).
I naturally searched Global esp. as the NI (and SAI - I don't have any) knives are supposed to be 12.5 (per side) bevel. The others are 15?. The NI are the two with the different handles and tip. There are 3 flexible blades there too.
I would expect the NI's and the standards and flexy to test differently.

Two weeks ago I had not heard of Global. While researching a gift my daughter will give her boyfriend (KAMIKOTO - she likes the box), I found Global and reviews were how they were thin, light and sharp. I bought one of their a chef's knives (3rd from left). I was so impressed I bought the others inc. some flexible ones. I have a utility on order.

CutterClutter20201130_184406.jpg
 
Here you go.
https://www.bladeforums.com/threads/i-tested-the-edge-retention-of-48-steels.1726279/ :)

I'm one of those who doesn't base a purchase on the blade steel. In all honesty, blade steel is the last thing I am concerned about, at least 99.999999982% of the time.
Really, the only time I even consider blade steel is if I'm looking for a "woods" knife. In which case, I want a plain old time proven carbon steel, or possibly 440A/420HC.
No super/ultra hard 60 plus Rockwell "super steels" need apply. If the edge rolls, I can fix that in a couple minutes using just the leather upper of my boot, or my belt, or worse case, my porcelain on steel coffee cup.
If the edge chips or the blade breaks, I can't fix that in the field ... a knife with a chipped blade is worse than useless to me. By definition, a knife with a broken blade is unusable and hence worthless, regardless of what the blade steel is.

Usually when I'm afield/I was afield, I'm not/I wasn't in a position (or location) where I can/could just pop on over to the nearest Walmart/Target/Kmart/Blaine's Tractor Supply/Hardware/Sears or sporting goods store to replace a knife - or anything else - even if I have/had the required funds available.

My current woods/hunting knife? Choice of MoraKniv Number 1 with "Carbone" blade; Western L66 (1095 blade); Buck 110 and/or 877 (420HC) Old Timer 7OT with 1095 or "Schrade +" (I have both), Old Timer 6OT with "Schrade +" blade, and a 1800's style Russell Green River Sheath Knife with 1095 blade... oh and a older Cold Steel Finn Bear, with I believe a 1.4116 blade. :)
Honestly, I don't think I need another woods/hunting knife in this lifetime ... though I would like a Russell Green River, or Old Hickory "Buffalo Skinner".... :)
 
Looking forward to the project. As Vadim has shown, deburring makes a huge difference to the sharpness and longevity of the edge. I’ve heard a rumour that Henckels finishes their edges on a 120 grit belt. That could explain why a factory Henckels edge is usually around 250-300 BESS? About the same as sharpening with a Ken Onion belt grinder with no honing?

Shuns meanwhile are around 100 BESS out of the box. It seems they do a better job deburring than Henckels does?

Again, looking forward to it!
 
You must log in or register to reply here.
Back
Top