Testing Protocol

My background is in engineering, new product development. Electrical/electronic products have to go through a variety of tests. Some of these tests involve other materials and it is important to standardize those tests. For example, if your product has a lightbulb larger than, as I recall, seven Watts, and if that bulb is not fully-enclosed by a suitable enclosure, then you have to undergo a test that involves placing the product on a sheet of paper, heating the lamp to full temperature, and then triggering a lamp explosion (there are several ways to do this). Any glass shards that fall onto the paper must not cause burn marks on the paper. That's the test.

Obviously, the goal of this test is to reduce the risk of the lamp igniting a fire. Is this test perfectly valid? No. If some idiot wants to put his lamp on a highly-flamable surface, then a lamp that passes this test could cause a fire. But the paper test is the test that is standardized and everyone uses. It's not perfect, but at least it's uniform.

Or is it? Paper isn't paper. Not all papers discolor or even ignite at the same temperature. Thicker paper is harder to burn than thin. Obviously, the exact type of paper is specified. In fact, one buys the special paper from UL.

If your product has an intake fan, you may be required to undergo dust incursion testing. The test specifications obviously call out the exact nature of the dust involved and there's a company that sells jars of standard dust.

Standardizing the materials used, especially materials that will be cut or chopped, seems absolutely essential to me. Will it be perfect? Does a lumber 2x4 exactly mimic the stresses that might be encountered chopping an actual tree branch? No. But just as there are in infinite number of surfaces that a lamp might be set on, some more flamable than others, there's an infinite number of tree branches and each one will chop or cut differently. For standardized testing, we need a standard material that is representative of real applications but which is highly repeatable.

Lumber 2x4s are very inconsistent as has been pointed out. But graded lumber 2x4s are much more consistent. This is lumber that's been inspected for large knots and other imperfections and which has been graded by a machine. It comes in three different grades. It is very consistent from New York to Los Angles. This is what allows an architect in NY to design a home that might be built in LA, or Dallas, or Seattle, or anywhere. That architect can specify graded lumber in the critical areas and know what that material will be. Fortunately, graded lumber is available anywhere building supplies are sold including Home Depot.
 
Gollnick - Yes and no. 2x4'S do come in grades, but as in most grading the standard is based on a minimum. So as long as the grade meets the minimum then its fine. That leaves considerable room for exceeding the minimum, and that is the problem with standardization. As Cliff pointed out the only way to overcome the variation is to use large numbers of a certain grade. This tends to level out the variation. However, it takes a lot of work and time, and that's something most knife testers either don't know or are unwilling to invest. As a result, tests can be skewed by a low sampling rate. This testing stuff is not easy.
 
Cliff Stamp said:
You could see it in the chops, some axes sailed through it like butter, the others just rang in the wood (same axes, all similar guys).-Cliff
Click to expand...

BTW Cliff, do you know where these guys get their superb axes? Are they custom axes? What steel is used, any idea?

These axes look so great with their large head/cutting edges...
 
Their axes are custom, they cost $500 for the head. The guys sell them, you can find them on the web. Carson in particular I recall runs a store where you can buy new and used ones. Note that these are *NOT* and I repeat *NOT* axes you would use to actually working cut wood, they are cast stainless.

My point about the variation is that even stock lumber has variation and just by picking material with less variation you don't eliminate it and thus ignore it, you just reduce it, and you can reduce it by simply using a larger sample size.

To get specific, one guy cutting random spruce all day long has a much more precise mean effect than another cut cutting one stock piece of 2x4. Not to mention the first guy has a hell of a lot more info on edge retention, handle ergonomics, edge durability, etc. and directly translates into users actually working with the knife.

The reason in research that you buy better materials is that labor and time costs are much more expensive than materials, for a dude using a knife, this isn't the case. I have to cut down 15-20 truck loads of wood each year, it doesn't cost me anything to use the knife to limb and buck up the smaller stuff.

However in the lab, if I have to buy research grade materials, or buy the cheaper and then purify it on my own, or take enough readings and baselines to work out the impurities, I will just buy the more expensive gas as it is cheaper than the time I would spent. Note of course even with the research grade gas there are still impurities that you have to deal with, just less of them.

-Cliff
 
2x4'S do come in grades, but as in most grading the standard is based on a minimum. So as long as the grade meets the minimum then its fine. That leaves considerable room for exceeding the minimum, and that is the problem with standardization.
Click to expand...

That's why I'd suggest picking the lowest grade. It's good enough to be in that first grade, but not good enough to be in the second grade. At least you have it confined to a range. That's about as good as you're gonna do for whole wood.

You could switch to a product such as MDO board, but that's not really representative of chopping and cutting actual wood.

I agree with testing multiple times and also testing multiple samples whenever possible. Many of the aforementioned tests for new products have to be repeated several times and/or with multiple samples.

No "laboratory" testing is ever fully realistic. Consider, for example, crash testing for cars. We've all seen the pictures and even video and we've all seen a "crash test dummy." Well, there are an infinite number of ways to crash a car. And people come in all sorts of different sizes and shapes and builds. People sit in cars in all sorts of different ways. They sometimes wear big buffy coats and they sometimes wear nothing at all. So, how do you come up with a small set of tests that can be practically done and that can be economically done (remember, most of these tests are quite destructive so it costs $30,000 per test nominally just for the car itself) that can allow us to evaluate a car's safety and that are consistent so that you can compare the performance of one model to another or so that you can compare from version of a certain model to another (Answer questions like, "did the new bumber design actually make the car safer," for example.). The answer is that you can not reasonably nor economically crash the car in every possible way with every possible person in it. So, standard crash test dummies were developed that attempt to represent a reasonable cross-section of people, standard crash protocols were developed that attempt to represent a reasonable cross-section of crashes, and even standard mechanisms were developed to impliment those crashes so that they happen in a consistent, repeatable way.

The same thing needs to be done with knife testing. We can't test every knife cutting every kind of possible material with every possible user. We have to develope our own set of crash test dummies and our own set of crash protocols. We may even need our own set of standard equipment to impliment the tests.

It's a big task.
 
aside from testing knives to destruction (and that will only let you know about that particular knife, since there is variation in lots of materiel, heat-treat, and quite a few other variables whether factory or handforged, which is why factories statistically sample and have an acceptance standard of mean failure rate), it seems to get lost in the quest for details, but a standardized test tells you nothing about a knife except how it will perform in those limited tests...a knife which cuts 2x4's strictly means only that, it is good at cutting 2x4's, and in no way guarantees how the knife will perfom on saplings, bone, etc., any more than shooting into wet newspaper or gelatin gives you any indication of how a bullet will perform on living flesh...it serves only as a baseline of comparison between different test subjects when those tests are rigorously kept as free of variables as possible...for instance, clamping a blade in a fixture which slides down a pole with a certain weight tied to the fixture so that the same force is applied to every blade in an equal manner....what you will learn from that is only that some knives perform better sliding down firepoles than other knives do....otherwise, the only true way to know how a knife will perform is to use it as it is intended to be used, and see how it performs...the same way the only way you will ever know how a rifle performs as a hunting weapon is to take it hunting, anything else being "scientific" data collection which in the end tells you nothing.....
 
boy, I'll bet there's a few "endorphin junkies" on these forums ;)

Medium is quite important.

On the CATRA abrasion resistance testing, the medium is CATRA specific with a particular chemical analysis requirement. It's quite expensive, but very consistent.

It's not that manufacturers won't share information, it's that sharing information creates a lot of controversy that the manufacturer would be taking the heat for.

We've found that heat treat, geometry and edge angle make for big differences in results. We use the equipment to determine how to best serve our customers, not marketing.

We've always provided as much information as we could provide without getting into argument. When asked, I will usually say, "in our testing", etc.

For corrosion resistance, we use "Q-Fog". It's reliable and calibrated.

For toughness, we use manufacturers specs (until our Charpy rig is set up).

Sharpness and abrasion resistance are the toughies because of all of the variables.

On our first cutting machine that we built, we used an insulation material that was very consistent. The CATRA was 4 times the cost, but more accurate, so we bit the bullet and purchased one (one of 17 worldwide) and we use it regularly.

We'll CATRA test anything you want, for a fair fee, but the results cannot be made public. We are not "accredited" and that's not our job. We would do it for individuals or companies wishing to learn more about a material.

On the other hand, sweet spots are more art than science. Ergos can only go so far in the lab, but for real ergos, "hot spots", etc, nothing beats "street testing".

We'll give Cliff anything he wants to spend his time testing because, we found him to be generally impartial (the "generally" is to avoid argument). We know he'll probably destroy it. Heck, we do. "You have to exceed the limit to find the limit". It also gives us the opportunity to compare his results to ours. The goal is to build a better, safer product.

sal

------------------------------------------------

The entire "business chain" from; raw materials, manufacturing, marketing, selling, distributing, shipping, warrantees, insurance, credit, etc. ALL EXISTS to service the ELU (End Line User). Remove the ELU from the equation and the entire business chain falls like a house of cards. We all work for you!
 
R.W.Clark said:
Argue knife use in another thread. This one is about testing media.
Click to expand...


Intended knife use has everything to do with what test media or method is employed....an advertised hard-use knife to stake your life on if you or a buddy are trapped in a burning HUMVEE might be fairly tested by prying, chopping into cinderblock, glass or other hard objects to simulate incidental contact while using the blade in an urban area emergency, extreme flexing, and suspending minimum adult male weight from the handle....I know that sounds "abusive" because it is...it also gives a soldier or civilian a rough idea as to what his blade would be capable of when called upon.....because he can't always make the call about what use a blade will be put, nor will he always have a bolt-cutter, crowbar, or power-saw on or about his person...he will be carrying a knife because he can....
 
Why does it always come back to "hard use / survival" type knives?
Now we are chopping glass?
Suspending the minimum weight of an average male? From what continent?
(I saw The Dirty Dozen too)
:yawn:

There are no perfect test mediums.
What we can try to achieve is consistancy in the tests, with mediums that represent what will be encountered in the real world.

The rest is just marketing.
 
I do carry smaller sharp knives for smaller chores, but also like having a rugged knife for the chores I fear might break the smaller blade, and it is a legitimate quest for many folk...there are several hundred thousand US servicemen who might particularly NEED a good knife like that currently, and I am sorry you find that boring...be glad they are there and you are here being bored ....but from what I see in comments here by many makers, apparently most advertising claims ARE hype....while in the Army, my knife did indeed act as a handhold while jammed in a granite rock face, it also levered AC units into MILVAN cutouts, perforated rusted-shut drums, chopped through ammo crate banding, chopped through roots hitting the occasional rock...those uses are not hype...they are a day-to-day reality for many military units....so I think testing should be allowed to simulate such conditions, especially if the knife is sold as something up to those fun and games....
 
Perhaps medium and function are not the same?

Wouldn't the medium first be established before the function testing could begin?

sal
 
Sal Glesser said:
Perhaps medium and function are not the same?

Wouldn't the medium first be established before the function testing could begin?

sal
Click to expand...

Wouldn't the function be established first, which would decide the media...I wouldn't use a skinner to chop through 2x4's....
 
Ebbtide said:
Why does it always come back to "hard use / survival" type knives?
Click to expand...
Because they are knives, and when you talk about evaluation of knives in general you should not restrain the testing to media which are specifically focused on one group and exclude others.

Any test should be accompanied by something that tests pretty much the opposite to show that all aspects of materials and designs are compromises. Ask yourself are you doing a review to sell the knife or actually provide information to the user.

No one is saying that all you should do is harder work, or that such work should be a large focus of the review of every knife. Simply that lots of knives are designed to do such work and the ability to do so leads them to being optimal for various situations and this should obviously be considered.

mtngunr said:
...chopped through roots hitting the occasional rock...those uses are not hype...
Click to expand...
I am very careful when chopping wood, still not a year has gone by when I have not accidently smacked the axe into a rock, usually on a bad glance. It doesn't happen very often, but it only needs to happen once if the steel can't take it. Similar for the large blades I use for limbing and clearing the brush from around the trees, usually there is hits a rock that is covered over.

This doesn't even mention the fact that there are times like you said when you actually are cutting roots or otherwise working in the ground. Or simply cutting used materials which have been left outside and thus can get contaminated with dirt and debris. Start a camp fire by chopping up some deadfall and you can easily chop into rocks in the bark.

-Cliff
 
lets chop bowling balls in half and put the old saying about soldiers to rest ;)

i agree with mtngunr on this. if a knife can chop thru rebar for instance with no damage will it not go thru 2x4 as well with no damage?

i think testing on cinder blocks is a good test media as it accelerates wear and tear and thats what people are looking for in a tactical/survival knife or at least i am.

as far as other types of knives it probably is a good idea to be realistic about the test medium and not try and skin a humvee with a swamprat safari skinner as it is not likely it will be in that kind of tactical situation.
 
The thing that cracks me up about the concrete block test is that it isn't even really that difficult. Busse has it done live with a Swamp Rat and you can't even see any edge damage in the pictures, the edge just dulled. But you have all these makers saying it can cause gross failure in their knives and voids the warrenty when all it does is dull a production knife costing a fraction of theirs.

And you can't bring up any arguement about overbuilt prybars because they are not running their edges significantly thinner, in fact I have seen lots of specs, and handled the same knives which are thicker and/or more obtuse. It is simply an illustration of just how brittle the high carbon stainless steels are when you can have an activity dull one knife and cause gross failure on another.

It certainly isn't the ultimate test of toughness either, most large knives will see use that is harder, an accident impact into a rock, or a hit into a nail is harder on an edge than chopping into a piece of concrete. Not to mention various other uses. The Solution that was broken earlier today would have not taken nearly as much damage on a concrete block chop as it did on the hardwood torquing.

I have seen lots of blades that would easily pass a concrete block chop but fail to be functional when limbing or trying to baton through a hard knot.

-Cliff
 
I saw Mr Clark's response on "U Haul packing boxes" and wanted to say something about Medium consistency in regards to "Cardboard". This is a reprint from a reply that I have written before on this subject.

Most everyone has their own vision of "Cardboard". The correct term used in the paperboard industry for paper that has at least two (and as many as six) Liners each seperated by a varying size of wavy interior Medium(s) is called Corrugated. This is what almost all shipping Boxes or Containers are made of, brown Corrugated paper - aka Cardboard.

However, there are other box materials that get called Cardboard that do not have Liners seperated by Mediums, it is one thickness of combined layers of paper tightly pressed together. This type of material is called by names such as "Fibreboard", "Pressboard" or "slickboard" - This material is usually used in items that require POPs (Point of Purchase Displays) or shelf presentation ie. Cereal Boxes, Laundry Detergent Boxes, Cig Cartons, etc.


Let's look at some other Terms of Corrugated/Cardboard that are used within the paper industry;

Liner - An exterior grade of Paper (usually Brown - Kraft, or White - Kla-White). This paper can be any number of Weights.

Weight - How much material was used to make approx 1,000 square feet. The weight can be semi-accurately determined by gauging the paper's final thickness using a mircrometer. The thicker the liner the Heavier the Weight (usually).

Medium - The wavy interior of the combined paper. This paper can also be of varying Weights. The "Waves" are called Flutes. There are several Medium gauges, the most common being "C" Flute, this one is about the size of a number 2 pencil. A smaller one is "B" flute, a Larger one is "A" flute. You can combine these flutes into just about any combination to make "Doublewall" (which is 1 Liner, a Medium, another Liner, a second Medium and 1 more Liner) or "Triplewall" or ....well... you get the idea.

Starch - All the Liners and Mediums are held together with an adhesive with a base of Corn Starch. This adhesive may be thick or thin, this will vary with each run of material and will effect the overall quality of the paper.

On the bottom of most every type of Corrugated Box or Tray is a Printed Stamp, a Box Maker's Certificate (or BMC). This is a Circle or sometimes a Diamond with a large number and other smaller info printed inside it. The number inside the Circle is the approximate ECT (Edge Crust Test) or the Mullen (Puncture resistance) estimation of that combination of Paper. The Higher the number in the BMC the tougher, more durable, more puncture resistant, the higher the stacking strength, etc.

NOW.... lets dig further. All Paper (whether it ends up being Liner or Medium) is produced at a Paper Mill. The Mill has huge Vats that create a Slurry of Paper fibers. Each slurry will have some percentage of Virgin wood material and some percentage of Recycled material (PCW - Post Consumer Waste). The normal amount of Recycled material is approx 15%, this number can increase to 85% or higher. With the decrease of Forests and the cheaper cost of Recycled material the Recycled content of all boxes has climbed dramatically over the past 10 years. The higher the Recycled content the higher the amount of foreign matter, tiny pieces of string, tiny metal pieces...etc...all wind up in the paper. There is no way to tell by looking at any given box how much Recycled material was in the mix when that paper was made, but the amount of foreign matter will greatly effect anything that is cutting it.


WHEW!! For those that are still awake I hope that helps with the understanding of the test material.... all Cardboard (or boxes) are DEFINATELY not the same. But there is certainly nothing wrong with testing out one knife to see how it will perform cutting lots of good'ol Cardboard.


For proper scientific testing the test material must be called out specifically and standardized as much as possible If you are serious about learning how different knives, steels or grinds compare to each other. However, one variable at a time should be considered (aka Apples to Apples).
 
Hillbillenigma said:
There is no way to tell by looking at any given box how much Recycled material was in the mix when that paper was made, but the amount of foreign matter will greatly effect anything that is cutting it.
Click to expand...
You can deal with this by pooling your cardboard and then drawing random samples from it, adding new material to the parent pool as it becomes available. The variance in the mean then is MUCH reduced from the variation from box to box. The larger the sample pool you draw from and the larger the sample pool you draw into, the smaller you can make this variance. You can evaluate it by simply repeating a set of work and checking the consistency and again, use benchmark knives.

There is also no need to constrain experiments to one variable, lots of research is done when dozens of variables change each time, this is where multi-variable techniques of analysis come into play you just need more complicated models. It is much easier when only one thing is being considered, however usually with knives you don't have this option unless you custom get test blades done. As a maker of course you should take this path as why make life complicated when you don't have to.

-Cliff
 
Cliff Stamp said:
You can deal with this by pooling your cardboard and then drawing random samples from it, adding new material to the parent pool as it becomes available. The variance in the mean then is MUCH reduced from the variation from box to box.
-Cliff
Click to expand...

Would it not be simpler to just get your cardboard from a single source that uses the same cardboard time after time? Like boxes from your pal who drives for Larry's House of Pizza and Kosher Supplies? Or from a grocery store with you specifying the crates eggs are delivered in? Otherwise, it would take a lot of random samples USING THE SAME KNIFE to get statistically valid results before you could even begin to test other knives...why average when you can start with an average of "1".....
 
mtngunr said:
Would it not be simpler to just get your cardboard from a single source that uses the same cardboard time after time?
Click to expand...
I buy stock materials all the time, rolls of thread for sharpness testing, rolls of hemp rope for edge retention and cutting ability and so on.

...why average
Click to expand...
Personally, I don't have a ready supply of cardboard so I pool a huge bunch from a large variety of sources. I have to average runs anyway to reduce the effect of user sharpening, method (speed of cuts mainly), and so on, so the cardboard variance is just one more that gets thrown in and then they all get reduced with cumulative runs.

But yeah of course if I had a steady supply of stock materials like graded 2x4, CATRA standard cardboard I would use it. I am just not going to spend $500 on such when I can buy a knife out of it instead. Any maker who wants to see me do stock testing with one of their knives can supply the material and I'll gladly do it. My point was mainly that there are ways to deal with materials variances, it isn't a significant consideration provided you take a few simple steps to reduce the effect.

-Cliff
 
You must log in or register to reply here.
Back
Top