What do you learn from destruction tests?

[Rick Marchand]

I haven't read through all of the responses but I'll give you my thoughts (as a maker) on the subject of testing.

I think it is the responsibility of the maker to do destruction testing. No customer should have to throw away money to see if a knife meets their expectations. The problem is that not enough makers test their knives and folks have different criteria by which they judge a blade. In a perfect World, the maker can tell you exactly when the knife will fail. Unfortunately, most of our testing methods are not perfect. A flex test in a vice is different than prying apart a dead stump... chopping through 2x4's is different than punching through a car door and slicing paper is not the same as carving a spoon. So we are left with good folks spending hard earned money, to destroy a knife, that might fit the bill.

I just talked in a big circle, right there.... lol. In the end, I want to know what my knives can and can't do... but I test them in a manner that meets MY expectations as a hard use knife. If a potential customer wants something more, I'd hope they would come to me, first. Production/factory knives are a bit different in that respect but they don't dent the budget as much, either.

Rick

 

[maximus83]

If you insist on modern scientific testing knives… Hire an engineer specializing in materials testing. Give him a tenth of what the Large Hadron Collider costs. He can buy and make testing equipment, and score a hundred of each knife to be tested. Let him report his work in peer reviewed journals. To make it really scientific, give matching funds and the same assignment to another engineer so he can replicate the tests. As soon as those results are in, get back to us.

It is an imperfect world. We do what we can with what we have. Some information is better than no information. Imperfect knowledge is better than no knowledge.

This is pretty much the right response, IMO. It's not practical to expect somebody on their own nickel, and their own time, without external funding, to do tests with all the scientific rigor that gets put into testing new cancer drugs. He's not going to be able to use multiple samples of the same knife, for instance. However, in a BASIC sense, he does follow aspects of the scientific method as pointed out: it's repeatable, he records his results with video documentation, he runs the same tests on each knife, etc. Are there various flaws and gaps in the testing? Sure! Does that mean the entire testing is useless? Not at all. As pointed out, we learn some useful things in these tests. And the tests help to confirm things that we think we already know, like the relative toughness of carbon steel versus stainless. As Raymond says, some information is better than no information. It doesn't have to be a 100% foolproof, "perfect" scientific test to add some value and knowledge. It just means that we have to be more tentative about the conclusions drawn from the testing. For instance, it's useful data to know that one s30v knife by a prominent maker broke under testing. However, from that one knife, it would not be fair to conclude that ALL knives in that steel lack toughness. However, the question has been raised, and it definitely is something to keep in mind and should be investigated further. I would love to see Noss do more tests on s30v survival blades, and s35v as well for that matter, to see how tough they are in extreme conditions relative to carbon steel.

 

[unit]

I haven't read through all of the responses but I'll give you my thoughts (as a maker) on the subject of testing.

I think it is the responsibility of the maker to do destruction testing. No customer should have to throw away money to see if a knife meets their expectations. The problem is that not enough makers test their knives and folks have different criteria by which they judge a blade. In a perfect World, the maker can tell you exactly when the knife will fail. Unfortunately, most of our testing methods are not perfect. A flex test in a vice is different than prying apart a dead stump... chopping through 2x4's is different than punching through a car door and slicing paper is not the same as carving a spoon. So we are left with good folks spending hard earned money, to destroy a knife, that might fit the bill.

I just talked in a big circle, right there.... lol. In the end, I want to know what my knives can and can't do... but I test them in a manner that meets MY expectations as a hard use knife. If a potential customer wants something more, I'd hope they would come to me, first. Production/factory knives are a bit different in that respect but they don't dent the budget as much, either.

Rick

Rick, those are some fantastic thoughts and I was really taken back by what I bolded above.

I am not a maker, nor do I consider myself a "tester" or even a reviewer, but when I get a new knife I intend to use in the woods, those are two of the first things I do to start my learning process for the new knife. I find that these two tasks offer me a good standard by which to judge the knife and its abilities to do some of the tasks I will call on it to do. How I came up with these tasks is not clear to me...I think I just sort of arrived at them.

I guess I need to check out one of your knives at some point...it seems (to me, at least) that we may hold similar expectations for knives

 

[l2lku2]

I do not agree that some knowledge, or a single data point, is better than no knowledge or multiple data points. "Some" or little knowledge can be misleading if not understood within perspective of a larger body of knowledge.

A basic example. Lets say one specimen of a product is tested and the first one fails, and only that data point is shared, then many will conclude that this is a bad product and not buy that product.

However if another 9999 other specimens of that same product are then tested and none fail, that provides a more accurate picture of the failure probability of that product. (For sake of example, ignore for the moment the probability of the only failure being the very first one in a very large sample set.)

In this case, limited knowledge of the first test can therefore be more misleading to those who do not consider the larger perspective.

While larger sample sets are not conclusive, understanding of the probability of failure becomes more accurate as the data set becomes larger.

A lot of newcomers to any hobby or body of knowledge, can be easily misled by limited information and that can sometimes mean they lose out on many opportunities. It is hard to know what you don't know until you know it.

That is a problem for our hobby, because many times the only data available is a single data point, or very small sample sets. I would also be very leery of buying a knife if I see just one fail, provided it is a test I consider relevant. I don't buy a knife based on a single review either. We all like to buy "peace of mind". That is one of the reasons why until now I have a disproportionate ratio of production vs. custom knives, because as a newcomer to a hobby, there is simply so much more data available for production vs. customs. It takes a long time to become knowledgeable about custom makers.

In our hobby, there is much subjectivity and variances of the testing procedures and results, and different opinions on the relevance / applicability of different types of tests. What is relevant and applicable is subjective and personal and we will never reach agreement. And that is where much of the controversy arises. Because of such limited data, those data points can be very influential, especially to newcomers and people have a strong reaction to what they feel is misleading to those who do not know better, yet, which in turn affects companies and makers, sometimes "unfairly".

 

[maximus83]

If you take these comments to their logical conclusion, you could never test anything at all--informally--and learn anything useful about it. So, if we make this kind of thinking the new standard, then when a user buys a new Spyderco in S35v steel, cuts something hard, and his edge chips or rolls, we should ignore that information because it wasn't learned via a sufficiently "scientific" test, performed with 20 samples, a control group, and a mechanical device used to do the cuts to ensure the same cutting motion and foot-pounds of pressure are applied to each knife.

Why not just be more practical and admit what we KNOW to be true in cases like this? It's a small sample, it needs to be verified in a wider context, but it is, or may be, useful information to add to the pool of knowledge. It is worth knowing that the edge rolled when he cut a certain kind of hard substance. Now if it turns out later that he just got a bad knife from the factory, that he was cutting something he shouldn't, etc., those are all mitigating circumstances. But I still want to know about it when something like that happens. Just like I want to know about it, if an owner of the type of car I drive has a brake defect that leads to their death. My reaction on reading something like that in the paper is not going to be "Hmmm, that's irrelevant, they didn't establish that the brakes failed by using scientific testing." Instead, my reaction is that I'd better keep an eye on that. It's not a proven pattern of brake failures yet, but it could be. If more of these start to happen, I'm going to need to call the dealer or maybe even get my brakes checked.

 

[l2lku2]

Your S35VN example can be used to illustrate how a single data point can be misleading without proper context because of how that single data point can be (and often is) extrapolated to reach inaccurate conclusions and be misleading.

There was a test of a S35VN Sebenza where the edge "flattened" and different people with various levels of knowledge could have been, and might have been influenced in various ways by that test.

With various levels of knowledge and experience in this hobby, some could conclude that they would not buy a Sebenza in S35VN (or any Sebenza) or any CRK in that steel (or any CRK).
Others could conclude that S35VN is a bad steel and they would not buy any knife from any manufacturer that used S35VN. And that actually happened and I saw such statements. And that is not an unreasonable position to take if that was the only data point available.

Only when different tests were posted did I see some people change their opinions of the steel. In that case, the original postulations (however premature and incorrect) were clearly misleading if seen in isolation. And there is still a lot of different opinions based on still limited data and different opinions of testing procedures, subjectivity etc. that will continue for some time.

I don't think people really care what anybody does with their knives in their own garages or kitchens. But I have seen a culture on many knife forums where the ones with tenure don't like to see newcomers being taken advantage of, whether by unscrupulous dealers with ridiculous prices, or manufacturers with BS claims and super secret steels or those who just spread a view that is vastly different from their perspective, for example that beating a knife with a hammer until it breaks is a representative test of a good knife.

So when those who have used S35VN with good results see it being discarded as a bad steel based on a single data point, or those who have used a CRK GB with good results being discarded as a bad knife because of a single data point, they will naturally voice a strong opinion.

And we will never reach agreement, and we will have these same debates over and over and over ad nauseam. And while most with tenure know that participating in these threads can be distasteful and make more internet enemies than friends, they still sometimes feel compelled to participate, I suspect not to convince those with strongly held opposing opinions, but hopefully to provide some perspective to newcomers or those with an open mind. And that holds true for all sides of the issue of course.

 

[Mike-The-Knife]

Wow . Some of you guys should be CEO of some of the major companies here in North America . You can make a simple thing so complicated that both Canada and the U.S. would have ZERO unemployment .

Lets start with something simple . Most of us can agree that the major knife companies who have been around a while and stood the test of time so to speak "should" have their manufacturing process and heat treat down to where it is consistant . Especially if they rank themselves as making some of the BEST knives available on the market . And really if any company CAN'T keep their knives to a close standard , why would you even consider buying one in case you got a bad one , right ?

Next we have the tests themselves that Noss does . Not very scientific perhaps , but he does follow the same steps in the same order on each test . I know this because I HAVE watched several . Before any comment is made on that , it is MY time to waste , weather you would waste your's or not . So since the tests are performed in the same way and since the better knife companies tend to produce consistant quality then it IS fair to say that if one manifacturer comes out consistantly ahead of the others , it's knives MUST be tougher .

If you look on the knifetests.com site you will see that they HAVE tested 2 or 3 knives (all be it of different models) of several companies . Now if you look at the ratings of all the knives from one cmpany , you'll see that each knife does seem to score fairly consistantly with the other knives from it's company . To me that's a fair indication of what company produces the toughest knives . And THAT is what I learned on knifetests.com . Which by the way was the original queston .


HOLD MY BEER AND WATCH THIS !

 

[Ankerson]

Wow . Some of you guys should be CEO of some of the major companies here in North America . You can make a simple thing so complicated that both Canada and the U.S. would have ZERO unemployment .

Lets start with something simple . Most of us can agree that the major knife companies who have been around a while and stood the test of time so to speak "should" have their manifacturing process and heat treat down to where it is consistant . Especially if they rank themselves as making some of the BEST knives available on the market . And really if any company CAN'T keep their knives to a close standard , why would you even consider buying one in case you got a bad one , right ?

Nest with have the tests themselves that noss does . Not very scientific perhaps , but he does follow the same steps in the same order on each test . I know this because I HAVE watched several . Before any comment is made on that , it is MY time to waste , weather you would waste your's or not . So since the tests are performed in the same way and since the better knife companies tend to produce consistant quality then it IS fair to say that if one manifacturer comes out cinsistantly ahead of the others , it's knives MUST be tougher .

If you look on the knifetests.com site you will see that they HAVE tested 2 or 3 knives (all be it of different models) of several companies . Now if you look at the ratings of all the knives from one cmpany , you'll see that each knife does seem to score fairly consistantly with the other knives from it's company . To me that's a fair indisation of what company produces the toughest knives . And THAT is what I learned on knifetests.com . Which by the way was the original queston .


HOLD MY BEER AND WATCH THIS !

I just think some are making much more out of this thing than it really is in the end, not really worth getting all carried away over.

But then that's just my opinion.

 

[Maddogg774]

I haven't read through all of the responses but I'll give you my thoughts (as a maker) on the subject of testing.

I think it is the responsibility of the maker to do destruction testing. No customer should have to throw away money to see if a knife meets their expectations. The problem is that not enough makers test their knives and folks have different criteria by which they judge a blade. In a perfect World, the maker can tell you exactly when the knife will fail. Unfortunately, most of our testing methods are not perfect. A flex test in a vice is different than prying apart a dead stump... chopping through 2x4's is different than punching through a car door and slicing paper is not the same as carving a spoon. So we are left with good folks spending hard earned money, to destroy a knife, that might fit the bill.

I just talked in a big circle, right there.... lol. In the end, I want to know what my knives can and can't do... but I test them in a manner that meets MY expectations as a hard use knife. If a potential customer wants something more, I'd hope they would come to me, first. Production/factory knives are a bit different in that respect but they don't dent the budget as much, either.

Rick

+1 especially for customs...I might add there are plenty of people who have more money then sense, they can buy the knives for testing

 

[AttilaTheHun]

I learned that there's a guy in his garage who scientifically tests knives. I say and stress "scientific" because:
- All knives are submitted to the same tests
- The tests are documented and repeatable by anyone. Noss uses common equipment that is cheap and readily available.
- The tests are run in a specific order every time, the same order for each knife. Start with easy like cutting cardboard, which every knife can do, but do it anyway. Then webbing. Then wood. Then bend at same angle for each knife, measured.
- Each time, Noss voices observations. Don't let the cussing or emotional tone fool you, there's valuable information. It sounds like a surgery, followed by an autopsy (plus slang).
- Noss is not rushed by time. There is no 5 minute format to match the viewer's attention span like other Youtube videos. If the video gets too long, he breaks it down into several videos but does the job right no matter what.
- The intent is find out what breaks the knife. Once the breakage point is characterized (like hitting concrete for 1 hour), then we have proof that the knife is good for everything in between cutting paper and whatever broke it.
- That's where science meets engineering. Of course no-one will cut concrete with a knife in the real world. In the dream world of science why not? But applied to reality it tells us where breakage is likely to occur and gives us a domain within which the knife can operate without breakage, regardless of what the manufacturer claims or the knife's datasheet/pedigree would have us believe.

This is called "load testing", as in load testing a bridge. In the old days, before computer-aided finite-element-analysis, a bridge would be load tested before opening it for traffic. Engineers would line up 50 full cement trucks on top of the bridge and see if it would hold, knowing that in reality it would never happen. If it held, then the bridge was deemed good enough for carrying day-to-day traffic, which would never even get close to 50 cement trucks at once.
Noss' destruction tests are just like that. Incidentally the tests are identical for each knife and give us a way to then compare knives' relative toughness since they were all subjected to the same tests. It was not the point of the videos really IMHO - in real life do we go around claiming that this bridge is better than that bridge because it took 60 cement trucks? There's other things to consider when comparing. But we can't help it, we like to simplify. As a result it upsets some people who have a strong allegiance to a brand, model, or country of origin when some other knife scores better.
Just like a bridge is easier to cross or handier or more functional or cheaper, these load tests aren't meant to account for this sort of things, they are just load tests to characterize the maximum load of the device.

My earlier comments were taken out of context. So here is some context.
Science doesn't mean complicated. 1+1=2 is science. Science doesn't have to involve a computer when a pen and paper suffice as it has for thousands of years. Noss' vice, 2x4s, concrete blocks and 3lb mallet are adequate and appropriate for the science experiment at hand.
Submitting all knives to the same simple tests (that anyone can repeat) in the same order without shortcuts and documenting it on a video, like Noss did, is science.

The only manufacturer that I've seen publicize some test videos is Cold Steel. Most of their marketing videos kind of have the same "slash manilla ropes" test. Or hanging from the locking mechanisms of folders (which indicates that CS DID do destructive testing of their locks, more power to them). Now granted these are marketing videos, the knives never break and we don't know how many takes were made but the spirit is there nonetheless and they do prove their pieces are not there just to be pretty.

If knife manufacturers willingly did destructive tests and backed whatever figure/number with a warranty, more power to them. I would trust the results more if done by 1 or more 3rd party on YouTube with no meat in the game though. And because seeing the average Joe breaking stuff in his garage makes me chuckle.

 

[Mike-The-Knife]

Although I have never been to one of the "Blade" shows , I believe that the Busse groupe (Busse , Scrap Yard & Swamp Rat) actually repeate some of their tests at the show so that people can see them in person . I don't know if any of the other companies do this , but prehaps some of the other folks on here would know and could add to this .


HOLD MY BEER AND WATCH THIS !

 

[Rick Marchand]

Although I have never been to one of the "Blade" shows , I believe that the Busse groupe (Busse , Scrap Yard & Swamp Rat) actually repeate some of their tests at the show so that people can see them in person . I don't know if any of the other companies do this , but prehaps some of the other folks on here would know and could add to this .


HOLD MY BEER AND WATCH THIS !

There are lots of makers doing live performance tests at Blade. One company that stands out is RMJ Tactical. They bring out 50gal barrels and beat the hell out of them. I feel sorry for the adjacent booths... lol.

 

[maximus83]

Tests performed by makers themselves are not that useful--too likely to be subject to manipulation by selecting the circumstances of the tests and excluding tests that would tend to reveal the weaknesses of the blades, etc. Independent testing is far preferable, even if it is informal, as long as the work done is documented in some way and is repeatable. Noss's work may not be perfect, but it qualifies.

Now, if the makers would submit a few blades to an independent tester who would test according to an agreed upon, repeatable method, that would be ideal.

 

[Rick Marchand]

Tests performed by makers themselves are not that useful.

I suppose you're right, if the maker is a douche and only worried about hyping knives to sell them. Some of us really care about quality and test blades under harsher conditions than most users would care to. Many makers (myself included) take performance VERY seriously. I don't produce enough to sit back on the virtues of a waranty. I don't want to replace blades. Of course, I would replace/refund a damaged product.... but that shouldn't happen if I've done my homework as a maker. There is no single knife design that can do everything. Likewise, not every knife can be a hard use tool but if a maker chooses to market as such, it better be damn tough. Every knife has its breaking point and it is the responsibility of the maker to have a general idea of where that is and present it with honesty for those who care to know.

For example.... I would be reluctant to submit a knife for a cinder block chopping test. Because, I know that the edge will roll, deform or possibly chip out. The edge is not meant to chisel through brick. Flip it over and use the spine or exposed pommel to destroy the block.... I can break bricks all day like that. Now, if you want to cut through a chain link, yes, use the edge. The spine will do you no good. The edge WILL get damaged but you will accomplish your task. If you called me up and said you used your knife to cut though a chain in order to free your dog or get into a locked cabin to escape the elements.... I would first want to know if you succeeded, then how the knife fared. Of course, it would be replaced.

That being said, if you called me up and said my knife failed your backyard 90deg flex test... I'll tell to to go piss up a rope. If you want a knife that'll pass a 90deg flex test, I'll make you one, but in my opinion, it will make a poor "hard use" knife because other attribute will suffer for the sake of having a bendy knife.

 

[Ankerson]

I suppose you're right, if the maker is a douche and only worried about hyping knives to sell them. Some of us really care about quality and test blades under harsher conditions than most users would care to. Many makers (myself included) take performance VERY seriously. I don't produce enough to sit back on the virtues of a waranty. I don't want to replace blades. Of course, I would replace/refund a damaged product.... but that shouldn't happen if I've done my homework as a maker. There is no single knife design that can do everything. Likewise, not every knife can be a hard use tool but if a maker chooses to market as such, it better be damn tough. Every knife has its breaking point and it is the responsibility of the maker to have a general idea of where that is and present it with honesty for those who care to know.

Rick

If you are forging your blades you would really have to test them before you sent them out anyway to make sure of the process.

However it's always good to test the blades anyway before shipping them out to confirm quality.

 

[Maddogg774]

We're also talking about finished and supposedly tested and refined products. Also for me when I saw how the CRK GB performed in Noss' test I wasn't gonna spend the $300+ and hope my experience would be better. I was willing to spend $119 on a Scrapyard SOD and hope Noss was right..he was! I'm in love with Scrapyards. Now the 511 is about $90 and 711 about $100. To me the more you ask me to pay for a finished product knife, the more I expect from it...sillyme!

 

[maximus83]

I think maybe we're talking about two different things. A company testing its own products for quality assurance purposes is both good, and necessary. I'm glad there are companies with integrity, including knife makers, who will do that. But I think independent COMPARATIVE testing--where you take blades from a number of different companies and subject them to the same tests, side-by-side, is a different kind of testing and is very valuable to buyers of the blades. For instance, think about the kind of comparative, independent testing that Consumer Reports does on many different products. I think that is the kind of thing that Noss is trying to do, albeit in an informal way. Bottom line: there is a difference between internal testing for quality assurance, versus external/independent and comparative testing.

 

[cziv]

I have a friend who's father has become a very wealthy man over the past 25 years by doing independent "destructive testing" for major automobile companies. I used to know some of the stuff they did but I honestly can't remember. They really torture tested the vehicles though - I do recall that.

I don't think the car companies reported this man's results anywhere but at lest they were/are involved in "independent testing." I think this is a good thing and may save lives as compared to applying "The Formula" as given in the movie Fight Club.