Friday, July 5, 2013

CATRA's TCC Numbers and Why They Matter

Andrew of Edge Observer called it "Bro Science" on the podcast, Episode 8, for those wanting to track it down, and that term is spot on.  Testing knives, evaluating their cutting abilities and resistance to damage or failure, is nothing like real science.  Instead it is guys in their backyards trying to break shit.  I love that.  Hell, roughly 50% of the allure of the Slingshot Channel is watching stuff go boom (you've never heard of the Slingshot Channel?  For shame, it is part Mr. Wizard and part Demolition Derby).  What exactly is a spine whack supposed to simulate?  What real world use is it an approximation for?  And why do I need a fixed blade that can be bent into a horse shoe?  What does that prove?  All of these tests are distinctly unscientific.

The hallmark of science, the scientific method, has very little to do with what is happening here.  There are no hypotheses being tested, no reliable and repeatable methods, and no systematic documentation or consistent measurements.  Sharpness comes down to cutting paper or blue jeans stuffed with sides of beef.  Durability comes down to being the last to break.  And the problem is, even if that particular knife did okay, because it is only one of hudreds or thousands in a production run there is no way to generalize its performance to the entire group (this is the small sample size problem you here so much about around election time and the beginning of the baseball season). 

This is not to say that it is IMPOSSIBLE to use scientific methods to evaluate knives.  It is.  Folks do it all the time.  They measure hardness in a scientific way using the Rockwell Scale.  And they measure sharpness and edge holding abilities.  We just don't have access to this last kind of information.  Virtually all of the major cutlery companies own or rent time on something called a "CATRA machine".  Here is more information.

The CATRA machine is a pretty ingenious device, in part because it is so simple.  First there is a mechanism that holds a knife.  Then there is a mechanism that hold the cutting medium (usually pieces of silica impregnated paper).  These two things are brought together and the knife's blade is drawn against the medium.  The machine controls the draw cuts allowing for incredibly precise slices.  These draw cuts are repeated again and again until the machine has performed 60 cuts.  Once this is complete, the machine counts the number of pieces of the medium that were cut in the 60 draw cuts.  The total is tallied and referred to as the Total Card Count or TCC.  A high TCC number correlates very well with good edge retention and high sharpness.  A low TCC number correlates very well with poor edge retention and low sharpness.  Here is a video of the machine in operation:

There are some quirks that can throw the numbers off, such as the use of ultra high hardness surface treatments (TiNC coatings, for example) and the organization of carbides, but generally the CATRA machine tells you how hard a knife's steel is and how well it cuts.  This, it seems to me, would be incredibly useful information, but we don't have it.

For reasons I cannot understand the cutlery industry has been very reticent to release this information. Spyderco, has, in the past released, TCC numbers, though they do not do this as much as they used to.  Bohler Uddeholm has released TCC numbers for their steels.  Global Knives, a kitchen knife company, has released TCC numbers, as has Cutco, another kitchen knife company.   The fact remains, however, that a majority of companies don't.  The numbers we have are either cobbled together from information from the past, leaks from companies, or tests done by knife magazines.  Here is the most comprehensive list of TCC numbers on the web, thanks to Cliff Stamp.   For more on Cliff's take on CATRA and TCC numbers, see Episode 9 of the podcast.

First, the numbers themselves are fascinating because, unlike the B-U list, these are steels as they are implemented by knife companies on the knives we buy, not theoretical numbers on steels under perfect conditions.  Second, they show that hardening and heat treat really matter.  Note the difference between Elmax hardened to 62 HRc (TCC of 959) and Elmax hardened to 60 HRc (TCC of 762).  Third, it is interesting to see how super-hyped super steels fare.  Steel junkies and numbers folks, rejoice, there is a lot of data here.

S30V doesn't look all that different from 440C when it is implemented by Spyderco or tested from the Dozier/Knives Illustrated sample.  Of course, hardness and edge retention are only one set of benefits for S30V, but it does make a point--steel performance is highly variable.  Also, take a look at the numbers for D2, an oldie but goodie steel and a favorite of custom makers just starting out.  The TCC of 666 at an HRc of 61 is awesome.  Not only is it quite easy to get D2 into that hardness range, but the cutting performance once you get there is great, on par with many newer, supposed super steels.  M390 and M4 really do seem to be worth the extra cost.  I was surprised at ZDP-189's performance, it was lower than I expected, but again we have little information about the sample used.  In fact, we don't even have the sample's HRc.  Steel aficianados, like Cliff Stampe, will tell you that the amazing thing about ZDP-189 is its performance under ideal conditions.  More so than other steels, it really benefits from best practices, rendering impressive results like those seen on the "no sharpening needed" Rockstead blade (Rockstead claims they only need to be stropped). 

The problem is that this is just the tip of the iceberg.  Looking around you might be able to add a few steels to Cliff's list.  A thread on Spyderco's forum confirmed that BD-1 has a TCC of 570.  There are other nuggets elsewhere.  But why is this done piecemeal?  I had considered paying for ten blades to be tested to add to the information, but the cost was prohibitive, something on the order of $2,000-$3,000.  An email or two later to some of the blog's long time readers convinced me this was a waste of funds, but the idea still tempts me every once in a while.  HRc numbers are out there.  Why not TCC numbers?

Cliff Stamp thinks it is because manufacturers don't see a benefit from releasing this information.  There are too many ways to rig the system (like performing more than 60 draw cuts) for it to be reliable and there is no way to confirm what other manufacturers get for results.  He also thinks it would lead to an arm's race, something that is expensive for companies and potentially costly.   I am not sure this is what would actually happen, though I think Cliff is right in believing this is what knife companies fear.

They might be concerned because of the reductionistic nature of the TCC numbers.  They might think that if they release this information it will be the only think people look at when buying a knife.  This is just not true.  First, HRc numbers are released and I know of no one that bases purchasing decisions solely on these numbers.  Second, while there are ways to rig the system, I think people would quickly find out that this happened and even if it took a while, the backlash would be very bad. Third, I don't think the kinds of people, us, that would be interested in these numbers would be swayed by the numbers alone.  People that care enough to search out TCC data are the same kinds of people that understand that steels are complex and no one number is the end all, be all.

I also don't think it would lead to an arms race, per se.  Releasing the HRc numbers didn't do that.  I am not sure, but I think Victorinox has done quite well with soft steels.  People that are interested in this information have multiple knives and understand that different steels do different things.  H1 has relatively poor TCC numbers, but that's like worrying about a Ferrari's off road abilities.  Pinnacle edge retention is not why H1 exists and no one that understands knives enough to seek out H1 would be dissuaded by its TCC number.  There might be an arms race among steels that offer no advantage other than edge retention, steels like M4, M390 and the like.  I could see a race to Rex 121 and its bonkers high hardness happening, but people would quickly tire of trying to sharpen that steel and go back to a more balanced formulation.  S30V is popular not because it is the hardest steel on the planet, but because it is one of the most balanced steels on the planet.  

Finally, there is a compelling argument from the flashlight world.  Look at what happened when the flashlight industry adopted the ANSI standard cutting out or reducing the number of ridiculous spec claims?  Instead of hurting companies it only helped spur on innovation.  ANSI standards made emitter upgrades a real and important thing.  They became marketing tools and consumers, blessed with more information, have bought as many or more lights than before.  More information, especially this kind of information, is always a good thing.

Benchmade, release your TCC numbers.
Spyderco, go back to your old ways and release your TCC numbers.
KAI USA, release your TCC numbers.
CRKT, release your TCC numbers.
Cold Steel, release your TCC numbers.
Boker, release your TCC numbers.
Buck, release your TCC numbers.

Gerber, cares, your stuff is garbage anyway and something tells me 7Cr or "mystery meat" stainless steel won't do so well.

It comes down to this: would you buy a car if they refused to tell you the MPG?  Right.  This is useful information that knife companies possess.  It would be helpful for us the consumers to have this information.  And it wouldn't hurt their bottom line to release it.  So, show us the TCC numbers.


  1. So, who's going to start the petition on

    FWIW, I think you're right Tony. The ANSI standard is exactly what came to mind.

  2. Speaking of M4, I just came across a limited production run of Benchmade Mini Grips in M4 hardened to 62-64 RC (the M4 on your chart was only at 61 RC). I can't afford it at the moment, but I thought someone else might be interested.

    They don't list the price, but it's $104

  3. Spyderco has been pretty clear on why they stopped releasing their data. What really needs to be done is for a clear standard to be written and accepted so that the numbers actually mean something. As it is now you can take any knife and generate any TCC count you want, making 420HC be greater than S110V for example by changing how the test is done. Forget the manufacturers for a second, just think about that from the point of view of the consumer.

  4. I've reviewed your site to serve as a particular purpose. Your site has not featured much in my few they are my own blog, mirc indir , mobil sohbet are bizimmekan site.