Spark :
Look at his Strider review - to paraphrase a portion he says "Hmmm, the ergonomics of the inner wrapped & bare steel tang aren't very good". Duh! Taking the handle off the knife and remarking on the ergonomics is beyond me.
This is a promoted aspect of full tang vs partial tang knives - functionality in extreme cases of handle failure.
Spark :
When you start tossing up impressive looking graphs and such, without addressing the underlying methods behind those numbers ...
Some of the older reviews like the one referenced didn't get into those aspects as no one expressed an interest at that time. I have been meaning to update them to deal with the current questions. There are indeed many factors which can influence results both in a random and systematic manner. The random aspects tend to average out, however the systematic ones need carefull study. The blade evaluation page will be where this will be discussed. It is still in its early stages and many aspects are still not covered.
I have been meaning to add to that page specifically comments on the hemp cutting as I have been doing a lot of it lately and found that it is really method dependent. Some factors are obvious but some are not and were surprising. I am also still working out how to turn the numbers into simple meaningful statements which depends on coming up with a model that reproduces the results to a decent degree which means understanding the process in more detail than I currently do. Unless you just want to use an empirical model which I don't, but may have to later on.
So yes, even simple things like cutting cardboard or rope can indeed be effected by method, factors such as speed, backing material, where on the rope you do the cutting (how far from the free end etc.), tension, etc. . As an example, the speed in which you cut cardboard will significantly influence the rate of blunting, one of the old reviews commented on this in some detail. Thus you have to cut at a constant rate, again small random changes will average out, so just take care not to rush with one blade to finish up for example. I would however not agree with the proposition that these issues are so very complicated that they can't possible be understood. We are talking about very basic issues and how they apply to a simple wedge.
Of course, some aspects can't be eliminated and have to be treated in separate ways. For example chopping comparisons will be influenced by swing technique. There are two main techniques; a very hard drive with *heavy* wrist torque, and a very fast, snappy swing with light wrist torque [usually with a draw]. Which technique is optimal depends on the physical characteristics of the user and the blade (mass, balance, length and grip). This is something else I have been meaning to look at in more detail. The results I quote for chopping comparisons are for the first method unless described as otherwise. Someone of different physical characteristics could indeed get different results, depending on how their body type favoured the blade, this has been discussed in the past. Wood type can also influence it by effecting penetration and binding. In detail, the greater the penetration the greater the influence of the primary grind, and thus the lower the effect of the edge geometry. I have been meaning to bound this spread, but unfortunately don't have access to any of the very hard woods.
Not2sharp, in regards to Ron's quote, what I estimated was the impact energy not the force, two very different quantities. The full details are in the review if you want to discuss the details fire away. Basically to estimate the impact energy, you need the time of the swing, distance covered and impact location. These can be well estimated by simple means. To reproduce the force of the impacts you need this estimate and the time of the impact which depends on the compressibility of the metals (mainly the bar) and the strength of the person holding the knife. To put an upper limit on the strength of the person you would just have the knife fixed in place, or could use some specific tension (spring) and state what it was. The looser the spring the lower the force from the given impact energy. The bar was probably just mild steel, which could easily be determined and reproduced.
In regards to machines being necessary, this argument doesn't even make sense from a basic perspective. It implies that people could not do informative work before we had such machines. Yes machines allow a higher precision which is always good as it allows examination of finer details. However, first of all, there is absolutely no correlation between precision and robustness (repeatability). Secondly, I would argue that performance which is significant enough to be noticed by a human user, can be estimated by simpler means (this should be obviously be readily self evident) However machines have one huge benefit which is they would allow you to do things much faster and thus would be of great benefit as you could look at more aspects.
For example I keep meaning to look at the slicing aggression of an edge as a function of grit size in micron and check if it is linear, the edge retention influence (which is nonlinear) and angle influence on both which I think is quadratic. The time it would take to get a solid conclusion is about 150 hours to do all the cutting (a lot of rope), and about 10-20 hours to log all the data and look at it in detail. With a CATRA-type machine it wouldn't take a day. The main cause for the large times when you are doing something by hand is repeating it to refine the measurements and eliminate the variance due to sharpening effects, consistency of materials, cutting technique etc.. Even when you do that you have to take care as experience can influence the results. For example with the hardwood dowel cutting, the more of it I do, the stronger my wrist gets and thus I have to keep rescaling the older results every few months (I have yet to update those
webpages).
To be clear, machine work would not replace the hand work just as stock testing doesn't replace normal knife work, it would just be complementary. The other reason I have not done the grit thing is that the all the sharpenings (about 75) would change the blade profile far too much. I have solved this of late with an Olfa Extra-heavy Duty cutter. The replacement blades are cheap and thus I can easily keep the profile constant by changing blades for every grit. It also allows a consistent reference for subjective sharpness comparisions (ease of slicing paper, shaving etc.) .
As an example of "human" testing, lets assume you want to see how tree type effects chops needed. You do this by keeping note of the chops required for specific types of trees. This is where the above posters would jump in exclaim "Aha, that is unscientific. You are ignoring random swing factors and not even taking into account the size of the tree and location - duh !" [light and soil type will effect density *greatly* even in the same wood class]. This is all true, however these are random factors and will average out in the long term. Results would not be very accurate if you compared a young and soft 4" spruce to a 12" seasoned pine, however after chopping through 1000 of each (this isn't an active years worth of wood chopping), the average size and nature of each type will be very consistent and allow a very accurate estimate.
Now a more relevant point would be : "What about if the Spruce trees in your area are consistently bigger or smaller than the Pine". Lets assume for example that a generation ago wood of one type was selectively cut (this was the case for Birch by the way). In this case the wood size would not be evenly distributed between Pine and Spruce. Same goes if one one type grows faster than the other, or reaches a larger fully grown size. To eliminate this factor you could go around the area in which you were cutting and estimate the size of the trees and see if there is a consistent difference (just keep a running sum).
Of course you could also just keep track of the size of the tree and the number of chops at the same time and compare scaled results, this would allow a much more precise work (which would mean you didn't need such a large sample). But this isn't even necessary if you do enough work. You can see this quite clearly by asking an exerienced axemen how many chops it will take him to cut through a tree if you allow them to do a test cut to check the wood. The best can tell you to within a chop or so. They have not even bothered to do the above counting by the way, its just the effect of a much larger sample size requiring a much lower precision necessary in the sampling.
Ray, thanks for the comments, but I made the decision to do the block chopping, so the responsibility for it is mine. I stand behind it for reasons as I have noted in detail before. It isn't even what I would consider one of the hardest tasks for a heavy use knife, lots of wood work is more difficult on a blade.
Mike, you have made a solid point in that the most valuable results for any individual with a knife they own is what they can do themselves. However there are lots of reasons to read what someone else has done. They might have knives you don't, have access to materials more readily, are more experienced with a certain technique or you just want to look at QC issues and manufacturer responce. For example in the recent thread on the Valiant Golok I was discussing various aspects of vegetation with Jimbo. What he describes doesn't grow around here so I have no way to get any experience with it directly. I also don't have access to trees of the size that he does, nor of the extensive leans that populate his vegetation.
Thanks for the support for those that offered it, and what is more did so clearly without attacking anyone. We are after information exchange here and that is the right path for that goal.
-Cliff
). It is this other portion that truly concerns me. Here are several reasons:
