which direction do you cut cardboard

hardheart

hardheart

Joined
Sep 19, 2001
Messages
8,968
When slicing one of those cardboard trays for holding six-packs, I noticed how much difference there was in force needed to cut parallel to the ridges of the middle layer, or perpendicular to them. I was using my paramilitary, and after 100 cuts it still shaved decently, so it wouldn't seem that the corrugated layer was bunching up ahead of the blade when slicing perpendicular to them (cut through a ridge, valley, ridge, etc). But I don't see why else I would have to push harder than when I cut in the direction that the ridges/valleys ran (cut through one ridge or valley of the middle layer along its length).

I hadn't really noticed the difference before, but it was significant. I was wondering which way was used by those who measure the forces needed.
 
I've noticed this too, and I think it may have to do with the adhesive that holds the corrugated layers together. When you cut parallel to the ridges of the inner layer, there's a good chance you're cutting the strongest part of the adhesive bond most the time. I think it's also possible that the cardboard tends to wedge more when cutting this direction.

I've only used cardboard for testing edge retention, just the amount cut before a certain amount of dulling occurs, never tried to measure actual force required. I think if you were measuring continuously you'd find a lot of variation in force during a single cut, whether cutting parallel or perpendicular to the ridges.
 
Same as Cliff and Knifetester. The main issue would be when testing knives to make sure you cut the same way with both, or all of the knives being tested. Kind of an obvious point.:)
 
What "grade" cardboard is being used?

Have you tried antifreeze boxes? They are about the toughest thing I've run across short of double corrugated.
 
Try some of the large (1.5 ft by 3 ft) fruit boxes at Costco. I have to chop to get through them. For grading, at least "single ply" is what I think is cardboard, otherwise, it's just paperboard.
 
Hardheart - I wrote a lengthy instructional at one time about Cardboard / Corrguated material. Check http://www.bladeforums.com/forums/showthread.php?t=336580&highlight=corrugated there for my brief review of that or you can do a Search and locate the original full length.

Good for you for asking questions. Most testers use this material without knowing anything about it and many will further try to "scientifically" compare different knives without insuring the testing medium is absolutely consistent throughout their tests thus at best making any data from the tests suspect.

Just be consistent and happy cutting.
 
Hillbillenigma said:
Most testers use this material without knowing anything about it and many will further try to "scientifically" compare different knives without insuring the testing medium is absolutely consistent throughout their tests thus at best making any data from the tests suspect.
Click to expand...

If you random sample from cardboard you will simply be introducing a random uncertainty in your results. If you do this on the same type of cardboard (boxes from the same company) the deviation is so small it isn't significant. You can basically looking at the standard deviation in the mean behavior of a large amount of cardboard. Even if the deviation was so huge that any one cut was 100% different from another, then the total run of even a short comparison would only have a 10% variation. If you average several runs or do long runs you can reduce this to a small percentage. The deviation due to the behavior of the steel itself is actually much greater, Landes work shows this in great detail. The main point is to not use different card stock, but that would be obvious. It would be like using one knife on pine and another on spruce and judging the blades accordingly. I only use boxes from the same company and always benchmark them with a standard blade. It is more important to make sure the initial sharpness is consistent as different blade steels don't respond the same and that you account for the effect of geometry if you are trying to make inferences on steels.

-Cliff
 
Cliff Stamp said:
If you random sample from cardboard you will simply be introducing a random uncertainty in your results. If you do this on the same type of cardboard (boxes from the same company) the deviation is so small it isn't significant.
-Cliff
Click to expand...

Rediculous. Using boxes from the same company is in no way reducing the amount of variables introduced into the "making" of the corrugated nor will it reduce the percentage of variability of the data gleaned from its use in Cutting tests. Corrugated Manufacturers utilize Paper stock (for both Liners and Mediums) from many different sources with infinite amounts of Recycled material in each Slurry of material and thus each Roll of paper. Throughout each day and even within the same Lot (Box Run) Manufacturers will change Rolls, splice Rolls, change Weights of paper within an Order, etc. Higher amounts of Recycled material (and sources of this Recycled Paper) will have varying percentages of post (and some pre) consumer waste in each Roll, some as high as 50%. This material could include (but is not limited to) soft metals (such as Foil Gum wrappers, Aluminum Foil etc.) harder Metals such as Paper Clip chips or shavings, Dirt, Rocks and anything else that could have been introduced during the Recycle dropoff, pickup, compacting...... it is rediculous to believe that these materials would not effect cutting tests.

Another Variable is in all Corrugated is the adhesive (Starch), this changes from Batch to Batch and the amount used is adjusted throughout productions of each Order produced. The amount of dried Starch effects Rigidity and more causes a rigid liner to be stronger, the fibers more tightly woven together, making cutting much more difficult.

Another Variable in Corrugated are coatings, many Corrugated materials have coatings of various types to reduce moisture penetration (wax) or color enhancement. This effects the Coefficient of friction of anything touching the exterior or interior of the corrugated, perhaps allowing a blade to pass through the material with less resistance. Of course this varies throught an Order and widely varies throughout a day or week.

The variations contained in any single corrugated box (inch by inch) are significant but the variables from different corrugated materials (even from the same Manuf) are huge. Simply using Corrugated from one Company is like saying that all Soda made by Pepsi should taste the same... cause their all made by Pepsi.. right?

I have not exhausted my ability to disprove your theory, but I have exhausted my patience. You may quote other theories from other Authors but I will stick with a simple proven principle of analysis - perhaps published by Demming if I remember correctly - paraphrased - variables exist, you must find them, if you ignore them you do NOT have true scientific results.
 
If you are cutting parallel to the ridges I think you are simply cutting less cardboard per inch, than if you cut across the ridges. Do to the wavy piece in the middle.
 
Hillbillenigma said:
Using boxes from the same company is in no way reducing the amount of variables introduced into the "making" of the corrugated ...
Click to expand...

In fact you have by basic statistics as you have constrained to a specific subset and thus the variation is less than of the entire population because you have removed a systematic variance. This would also be further true with any additional constraints. The only time this would not be true would be if all the subsets were identical in behavior which would mean that there were no systematic behavior at all and that all cardboard, regardless of manufacturer or type all behaved exactly the same. Now if that was true then your situation is even simpler because you don't have to subset any more.

.... it is rediculous to believe that these materials would not effect cutting tests.
Click to expand...

All measurement is random to a degree and at some level everything effects it if you want to go to a fine enough scale. What matters is how much and if this deviation is so large that it swamps out what you are actually trying to measure. There is a large amount of statistics designed to handle such issues.

Lets assume I have five boxes and that some are twice as abrasive as the average and some are half as abrasive as the average. Take these boxes and cut them into sections which are 10 cm wide and then do your 10 cm cuts through a random sample. You have now removed any systematic variance and the remaining problem is simply making sure your sample is large enough so that the basic root square error reduction gives you the required precision. Median or similar robust means of analysis would be preferred.

If you take a large enough sample then you can work with really uncertain data. I have had to do it on many occasions when studying stellar spectra. I was using dry ice to provide a thermal point and it was producing large thermal gradients in the media being studied, the spectra were more than ten times as noisy as the room temperature ones. All this meant is that I had to take more of them to get the required precision.

Always repeat your work to confirm behavior and if possible do the cutting blind (at least once) to check for personal influence. This just means you don't know how much cutting was done and thus the stopping point is obviously honest. Also do the work on unknown knives and see just how accurate the results are when again you have no idea what they should be and thus can't influence the results to a personal favorite.

This is why it would be best if all reviews were done on unknown steels which were only revealed after the work was finalized.

You may quote other theories from other Authors ...
Click to expand...

I didn't quote a theory, I cited published work.

-Cliff
 
thx, Hillbillenigma

I see where both you and Cliff are coming from. Depending on the cardboard used, maybe a particular knife will make 100, 150, 200, or 250 cuts on a run before being reduced to the same level of sharpness. So, the constitution of the test material has had an effect on the results. Yet, you won't narrow those results down anymore for any other knife, either, so you can still compare results when you have made enough runs. It doesn't matter how many different types of wood are used to make a comparison test, a chainsaw is going to beat a butter knife every time. But, you can't say that a chainsaw or butter knife cuts through 'wood' for a certain measure, you will need a range, and it may be broad, especially when you have not done a better job in defining your test medium.

Absolute precision isn't absolutely necessary, imo, since we cut much more than cardboard, and we all cut in different ways in different circumstances. But just to know that a chainsaw outperforms a butter knife consistently is pretty good. Sure, S30V vs ZDP-189 is going to have somewhat closer results, you need to cut more cardboard. With enough runs, the difference in performance will be shown, though there could be overlap which would be harder to attribute to the differences in steel or cardboard. Then I guess even more runs would be needed, or just different tests.
 
hardheart said:
Depending on the cardboard used, maybe a particular knife will make 100, 150, 200, or 250 cuts on a run before being reduced to the same level of sharpness.
Click to expand...

It is actually way more than that. I have seen a factor of ten easily and this is on the same thickness, same type, just different boxes/containers. Having a constant reference knife helps a lot because you can judge from it.

You really need to repeat cardboard work a lot, even if you buy specific stock there is massive variation just in the blade itself because steel structure is pretty random, especially in the high carbide steels. Again always repeat.

Basically repeat the measurements until the mean doesn't change significantly, to a minium of 3-5 runs anyway. More runs never hurt if you have the time. Of course if you can only do one run then just do one run, it is better than doing no runs.

Essentially you compile all the data anyway so you one run just goes into the pool of every other piece of work and becomes just another data point for steel/knife comparisons. If you just do one run then that data point has a pretty large uncertainty is all.

-Cliff
 
variables exist, you must find them, if you ignore them you do NOT have true scientific results.
Click to expand...

What difference does it make?

It is a knife review. DO you think that in use you can tell a +/- 5% variation for most uses?

IF there is not stastical significance, proper sample sizes, proper bounding, etc, does that really diminish a reviews value in any way?

How about a review that says" Got the knife, cuts great, about the same as a NIB Endura, and holds and edge about like a Victorinox paring knife, opens like a Buck 110." IS that review worthless because it is not "scientific"?

Is measuring edge retention down the the .00000 really meaningful to anyone?

If one knife cuts +/- .6% better than another or +/- 1.4593% worse than yet another really significant in practical use to anybody?

How about, damn this knife feels good in my hand. Can you measure that? (some ergonomisist would say yes they can, but I can't and really would not care to)

Give me a good old subjective review based on some good use, by a person that has used a variety of knives. Throw in some cool pictures once in a while as a bonus.
 
knifetester said:
What difference does it make?

It is a knife review. DO you think that in use you can tell a +/- 5% variation for most uses?

IF there is not stastical significance, proper sample sizes, proper bounding, etc, does that really diminish a reviews value in any way?

How about a review that says" Got the knife, cuts great, about the same as a NIB Endura, and holds and edge about like a Victorinox paring knife, opens like a Buck 110." IS that review worthless because it is not "scientific"?

Is measuring edge retention down the the .00000 really meaningful to anyone?

If one knife cuts +/- .6% better than another or +/- 1.4593% worse than yet another really significant in practical use to anybody?

How about, damn this knife feels good in my hand. Can you measure that? (some ergonomisist would say yes they can, but I can't and really would not care to)

Give me a good old subjective review based on some good use, by a person that has used a variety of knives. Throw in some cool pictures once in a while as a bonus.
Click to expand...

Guess there are many types of reviews. I like them all. It lets me see through many people's eyes. I like Cliff's work because he does try to quantify what for many is a very subjective experience. And I like the folks who say, "By golly that knife is comfortable in my hand and has good workmanship." I find value in both types of review.

Somebody here started by talking about, "does it make a difference which way I cut cardboard?" That makes this more of a Cliff-type subject because it gets into things that are beyond a subjective feel.

But, "no harm, no foul." I have seen Cliff complement subjective reviews and I don't think he said in this stream that he found no goodness in such.

Vaya Con Dios.

=====================================
In God I trust.
All others bring data.
 
knifetester said:
What difference does it make?

How about, damn this knife feels good in my hand. Can you measure that? (some ergonomisist would say yes they can, but I can't and really would not care to)

Give me a good old subjective review based on some good use, by a person that has used a variety of knives. Throw in some cool pictures once in a while as a bonus.
Click to expand...

I've read too many reviews that are something like "Best knife in the World" or "All the knife you'll ever need, 'nuff said". Completely meaningless tripe. Of course measuring matters. No, you can't measure everything, but that doesn't mean we shouldn't try.
 
But, "no harm, no foul." I have seen Cliff complement subjective reviews and I don't think he said in this stream that he found no goodness in such.
Click to expand...

Please don't take my post above as being critical of Cliff's work, it was not intended to be in any way. Cliff simply does many, many things I am at a loss to understand, and am clearly not smart enough to comprehend (such as the math stuff, the squigilly line graphs, equations, physics, materials science and stuff).

My point was jus the opposite in fact, why ridicule Cliff for not being "scientific" enough, when it is not really important in this context. It's not like he wears a white lab coat while performing tests or anything. It is a knife review, not a search for a cure for cancer.

Say what you want about Cliff, but find another person that spends more time comparing knives on the internet, or has for so many years. Good luck.
 
I've read too many reviews that are something like "Best knife in the World" or "All the knife you'll ever need, 'nuff said". Completely meaningless tripe. Of course measuring matters. No, you can't measure everything, but that doesn't mean we shouldn't try.
Click to expand...

Just don't lose sight of the forest though the trees. . .
 
You must log in or register to reply here.
Back
Top