Jerry will never be able to satisfy you, the ABS will never satisfy you, you will always comeback with something else that you don't like.
Uffda, I made some (IMO valid) comments on the ABS method of testing, in that it is used to measure the skill of the applicant in designing and making a knife that passes the test. Not to make the perfect work knife, which may or may not be different.
It doesn't say that they don't satisfy me. Any testing improves my knowledge of the knives in question and is therefore satisfactory.
This thread was meant to make knives comparable (on the issue of lateral stress at least) from a scientific viewpoint. This is quite possible, both with Busse's setup as well as that from Fallkniven.
Zius, I think this is terrific. But, again, this is only two knife brands going head to head and, we don't know if Fallkniven is agreeable to the test. This is where a lot of the testing goes to hell. The one maker or another making a complaint that their knives were tested under conditions they did not agree to, did not witness, and with a blade of unverifiable condition prior to the test.
Spearhead,
The data comes from Fallkniven's own site, and I will not be adding any data of my own. Just a calculation (which I will publish in full, so anybody with a calculator and some knowledge of mechanics can check me for any errors) to make them comparable with Busse's data, for which I alone am accountable. I will however ask permission to Fallkniven before posting results, and post them at Fallkniven's forum as well.
Thanks for the other comments. I agree completely.
If the tests are so far removed from actual use then they are meaningless.
No. Testing of car crashes is VERY useful, even though they are based on lab-testing and therefore fail to include many conditions that may apply in real life. They are simplifications of real life, therefore they are able to be modelised, and calculated. Car crash testing can be done quite well using calculations of theoretical models only (using the Finite Elements Method). Car crash testing in lab conditions is an even better approximation of how a car will react to a collision. The real thing is obviously the true test, but unfortunately, impossible to measure because the measuring equipment is not in place when the car crashes
The first thing I would like to see are materials tests, impact toughness both notched and not, tensile and yield point, ductility, wear resistance and corrosion resistance.
This data along with geometry specifics makes a *huge* jump in killing all hype. The next step would be designing a simple few stock tests which would give a decent picture of overall performance. I have been doing the latter off and on for some years.
Agreed. But based on the Fallkniven test setup it is possible to approximate the ultimate stress point.
Zius, Thanks for the offer to plug the numbers in to your formula, but the introduction of "stress risers" that are the result of using a sharp cornered vise are immeasurable. If I introduce rounded micarta jaws, rounded wooden jaws, or in the case of the Lulea tests round dowels to bend around, then the near complete lack of stress risers involved in the test will yield greatly differing results.
Would you see any opportunity to set this up in the near future? Rounded jaws of hardened steel (to make sure the material doesn't dent and therefore don't affect the performance of the blade) would IMO be the best choice, as well as the choice probably being used by Lulea University.
Evolute,
Sorry mate, but this is the normal science/engineering process in any capital intensive object, whether in the fields of mechanical engineering, aerospace engineering or marine engineering:
A) Make a conceptual model of a certain object
B) Build a model in a laboratory
C) Test the model, to see if the conceptual model approaches reality in laboratory conditions
D) Improve the conceptual model, and repeat the cycle until it is accurate enough to use. All within the same design parameters of course
E) Publish the conceptual model, backed by the lab testing
Than engineers start to use these models in the following way:
1) Make a conceptual model of the object you are designing
2) Calculate using the methodology published
3) Test the object on a model scale if applicable. Compare the results to the conceptual model and amend the conceptual model if neccessary.
4) Construct the real object
5) Test the object in laboratory conditions if possible, or (with large objects, such as airplanes and ships) in real life approximating laboratory conditions as good as possible.
6) Compare the results to the conceptual model and amend the conceptual model if neccessary.
7) Measure user opinions when the object is in use. Don't amend the conceptual model using these opinions, but rather use them as a guide how to improve the conceptual model and the laboratory testing.
As you can see, it is a cycle, during which the conceptual model and the calculations based on it get better all the time. In aerospace and marine engineering, we have *almost* reached a point where model testing or lab testing isn't neccessary anymore, since everything can be almost perfectly modelled using Computational Fluid Dynamics. For mechanical engineering, those Finite Elements Method programs are getting better all the time too. We will see the time where the reality can be near perfectly modelled.
This comes pretty close to what I meant to say in a much shorter way. Your comments are only valid for point A), where a new research is started. But for simple problems like pure bending as in this case, point A) is pretty easy.
as an engineer who spends much of his time these days working on automated systems, it would be relatively easy to produce a slicing machine, clamp a knife onto the end of a pneumatic piston, with a compression/strain gauge in the fixture and watch the compression go up as the knife gets blunter with so many slices of rope or whatever. get two machines in shiny aluminium and perspex boxes, one for you and one for the competition, and run 'em side by side. have you ever seen those mockups of seat testing apparatus at ikea? the mechanical side of it is very easy to build. i don't know so much about the monitering of the strain gauge, datalogger and realtime displays on a PC though. perhaps you could just use a spring balance, if the pistons were moving slow enough for folks to see them.
That would IMO be ideal too. But don't use rope, but a material that has a more homogeneous structure. Some sort of plastic maybe.
Edited for spelling. Forgot to preview...
