Accurate, replicable method of measuring ski flex (long)
Moderators: Head Monkey, kelvin, bigKam, skidesmond, chrismp
endre,
Got the pics this time. Thanks. Very nice work... both on the testing apparatus and on compiling the test results. After seeing your tester, I don't think I'll post a pic of my flex analyzer... well, maybe for a good laugh.
I'm sure that we all probably agree (and as justin56 pointed out), there is also a great deal of value in empirical testing methods... and it's all that many of us have to go on. Like justin56, I also used to have a 6'2" Berzerker tester, but I eventually had to divorce her because I could not afford to feed her.
G-man
Got the pics this time. Thanks. Very nice work... both on the testing apparatus and on compiling the test results. After seeing your tester, I don't think I'll post a pic of my flex analyzer... well, maybe for a good laugh.
I'm sure that we all probably agree (and as justin56 pointed out), there is also a great deal of value in empirical testing methods... and it's all that many of us have to go on. Like justin56, I also used to have a 6'2" Berzerker tester, but I eventually had to divorce her because I could not afford to feed her.
G-man
Shure:
The thicknesses are for the total ski (with base and top) and only the extremes: thickest part and thinnest at both ends back tip-middle-front tip
Atomic Big Daddy: 6-16,8-5,8
Atomic Thug: 6-15,2-6
Nordica Blower: 5-13,6-5,1
Dynastar legend pro xxl: 6-14,8-6,6
Goode Carbon 115: 5,8-12,2-5
4FRNT EHP193: 6,5-16,1-5,9
Movement Goliath: 6,2-17,6-5,7
K2 Pontoon: 6,8-15,4-6,8
K2 Maid'n AK: 6,7-13,3-6,45
Elan M111: 6,5-15,3-7
Völkl Sumo: 6,4-13,2-6
The thicknesses are for the total ski (with base and top) and only the extremes: thickest part and thinnest at both ends back tip-middle-front tip
Atomic Big Daddy: 6-16,8-5,8
Atomic Thug: 6-15,2-6
Nordica Blower: 5-13,6-5,1
Dynastar legend pro xxl: 6-14,8-6,6
Goode Carbon 115: 5,8-12,2-5
4FRNT EHP193: 6,5-16,1-5,9
Movement Goliath: 6,2-17,6-5,7
K2 Pontoon: 6,8-15,4-6,8
K2 Maid'n AK: 6,7-13,3-6,45
Elan M111: 6,5-15,3-7
Völkl Sumo: 6,4-13,2-6
I'm sorry, I did not think of putting in those numbers, that illustration was intended for less dedicated people than you guys.
The y axis is stiffness, calculated from deflection of a 10cm span of the ski.
The results from the deflection test are in mm, then I divide the force I use (25kg/249N) with the deflection of the ski (mm). this number gives an indication of the stiffness of the ski. The stiffness varies between 10-1000 N/mm, depending on where you measure. The middle of the skis vary between 300-1000 N/mm, but in the diagram i put out here the softest mid-part is around 450N (Goode) Of cource the middle of the ski is not the most important part.
You can see three marks on the y-axis. the one at the bottom is 0N/mm, the one at the middle is 500N/mm, the one at the top is 1000N/mm
Here I got a scetch of the principle of the flextester, it can be done really easy with cheap tools, you don't have to make a machine like mine (would be cool if you did though, that would make us able to communicate flex geometry)
The y axis is stiffness, calculated from deflection of a 10cm span of the ski.
The results from the deflection test are in mm, then I divide the force I use (25kg/249N) with the deflection of the ski (mm). this number gives an indication of the stiffness of the ski. The stiffness varies between 10-1000 N/mm, depending on where you measure. The middle of the skis vary between 300-1000 N/mm, but in the diagram i put out here the softest mid-part is around 450N (Goode) Of cource the middle of the ski is not the most important part.
You can see three marks on the y-axis. the one at the bottom is 0N/mm, the one at the middle is 500N/mm, the one at the top is 1000N/mm
Here I got a scetch of the principle of the flextester, it can be done really easy with cheap tools, you don't have to make a machine like mine (would be cool if you did though, that would make us able to communicate flex geometry)
by the way, don't bother the 500N mark on the drawing, you can use any wheight, the N/mm number will come out pretty accurate. For accuracy though I allways use the same wheight (and around room temperature), 25 kg (249N)
Here is another picture which shows a less accurate and slower but much cheaper way to measure flex:
Here is another picture which shows a less accurate and slower but much cheaper way to measure flex:
thanks! If we could make a standard and easy way to measure and communicate flex we could start finding out the best characteristics, what's good/bad about different skis, etc.
I will post the rest of my results after it has been published in the magazine, 40 pairs in a couple of weeks (jibskis, fatskis and skis for girls) then 40 pairs one month after that. (carving skis and bc jib skis)
I will post the rest of my results after it has been published in the magazine, 40 pairs in a couple of weeks (jibskis, fatskis and skis for girls) then 40 pairs one month after that. (carving skis and bc jib skis)
Thanks for posting your methods Endre, I was wondering how the measurements were done in the last issue of fri flyt...
It seems you have developed a very precise technique and it yields data that is completely coherent and reprodusable.
I have an idea for a testing apparatus that involves the same type of idea that powdercow suggests at the beginning of this post. The idea is however to add a digital camera into the equation. By taking a picture of the ski from the side during loading, you will get an idea of how the ski flexes under load. By specifying the placement of the camera as well as the focal length, uniform results should be possible. The curve of the ski could then be traced using a cad-programme, and correct scaling of the curve would make all skis measured in this way directly comparable.
This technique however involves lots of possibilities for making mistakes, but doing a few measurments and then comparing them with your curves generated from testing might give us a chance for calculating the flexcurve of the ski under any given load.
Comments?
By the way: Endre: Have you talked to anyone about simulating the properties of a given layup using a finite-element analysis programme? I know this is a complex task, but finding a way to do it would eliminate a lot of trial and error to find out how a ski is going to behave after changing any given parameter.
It seems you have developed a very precise technique and it yields data that is completely coherent and reprodusable.
I have an idea for a testing apparatus that involves the same type of idea that powdercow suggests at the beginning of this post. The idea is however to add a digital camera into the equation. By taking a picture of the ski from the side during loading, you will get an idea of how the ski flexes under load. By specifying the placement of the camera as well as the focal length, uniform results should be possible. The curve of the ski could then be traced using a cad-programme, and correct scaling of the curve would make all skis measured in this way directly comparable.
This technique however involves lots of possibilities for making mistakes, but doing a few measurments and then comparing them with your curves generated from testing might give us a chance for calculating the flexcurve of the ski under any given load.
Comments?
By the way: Endre: Have you talked to anyone about simulating the properties of a given layup using a finite-element analysis programme? I know this is a complex task, but finding a way to do it would eliminate a lot of trial and error to find out how a ski is going to behave after changing any given parameter.
Thanks Svimen!
I have done a lot of testing of ski shapes with FEM, mostly in Catia. Pretty complicated stuff though, with all the fibre directions, different material properties etc. It is fun and creates nice animations of theoretical ski flexing. But in real geometry research I think it is more fruitful to actually test the skis on real snow.
The method you describe is a pretty normal method of testing flex, but they do it a little bit differently. if you check out the telemarktips video from the karhu factory you can see that they trace a flexed ski with a measuring device. This generates a curve that tells something about the flexing properties of the ski. The problem with this method is that the camber of the ski affects the results, and you have to compensate for this first.
here is the video:
http://www.telemarktips.com/video/factory9.mpg
I have done a lot of testing of ski shapes with FEM, mostly in Catia. Pretty complicated stuff though, with all the fibre directions, different material properties etc. It is fun and creates nice animations of theoretical ski flexing. But in real geometry research I think it is more fruitful to actually test the skis on real snow.
The method you describe is a pretty normal method of testing flex, but they do it a little bit differently. if you check out the telemarktips video from the karhu factory you can see that they trace a flexed ski with a measuring device. This generates a curve that tells something about the flexing properties of the ski. The problem with this method is that the camber of the ski affects the results, and you have to compensate for this first.
here is the video:
http://www.telemarktips.com/video/factory9.mpg
Thanks for the video, seeing it done like that makes it seem easy...
Still some questions about the flexpatterns and ways of testing it. (Hopefully my illustration is visible.) Testing and calculating the flex in the ways discussed above have the limitations of the third drawing in that it is not really consistent with the way in which a ski is flexed (especially not in loose snow)
The forces are not applied only to the tip and tail of the ski, and therefore the flex pattern is not correct. (A flex pattern measured in the correct way with a more or less uniform mass being pushed down by the ski will tend to more flattended in the middle, because this is where most of the forces from the skier are taken into the snow.? I assume this to be correct simply based on inuition.)
However, using your method of measurement should allow us to calculate the exact bending pattern of the ski, because you measure the stiffness over the whole lenght of the ski in discrete intervals. (10cm) If we could then find the formula describing the flexing of a uniform board (ski) being pushed into a uniform mass (snow) by a force applied onto the middle, it should be possible to come up with a crude approximation of the actual shape af the ski in action. This could then be coupled with your measurements to generate a model of the varying stiffnes of the ski, which could the again be used to calculate a more accurate flex pattern.
Or we could just take a ski, force it into a large block of some uniform material (oasis or something similar) and then compare the shape to the results from both powdercows suggested method and your data. Probably they are not far off...
