First of all sorry for the rash of posts today. I have had a million questions and ideas bouncing around my head and I finally got a chance to get some of them down on paper today. I also saw that endre posted something similar to this a while back but I wanted to see if there is any renewed interest.
The two most important factors affecting how a ski performs are its shape and its flex (feel free to dispute). For the most part communicating the shape online is easy. Sure, some of the non traditional sidecuts that are being developed make it more difficult but even when a simple 125/102/115 won’t do a picture or cad drawing does a good job of getting the message across.
Flex is a different matter altogether. There are expensive machines to evaluate flex, individual companies have a flex ratings that work to compare skis made by that company (igneous), there have been relative rankings scales made like this one. Despite all of this there is not an easy, and accurate way to compare the flex of skis without feeling the all skis being compared. Here is my idea:
A single ski is placed with each of its ends on a support (box, bench whatever). The supports terminate in a 90 degree square and are positioned so that they end at the un-flexed contact points (corresponds to where the camber mold meets the tip and tail molds in the press). In other words all of the effective length is hanging free between the supports. The distance from the ground to the “middle” of the ski is then measured and becomes Z. A 50 lb weight is hung from a short piece of 2x4 placed flat and perpendicular to the ski on ski middle. The distance from the middle of the ski to the ground is again measured and becomes A. This process is then repeated with the weight being increased to 100 and then 150 lbs. These measurements become B and C respectively. The rating of a skis flex looks similar to traditional shape measurement and is expressed Z-A/Z-B/Z-C.
The reason for multiple weights is that different skis flex different progressively. As an example when comparing a ski with flex 25/45/130 and one with flex 25/40/50 (all measurements in mm) we could say that they both start out fairly soft but the second ski becomes much stiffer under load.
While I have giving this a bit of thought I still have a ton of questions:
1. What is the “middle” of the ski? Midpoint between the supports? Chord center? Recommend boot center?
2. What about zero camber and reverse camber skis? How do we standardize where the supports are put?
3. Does some sort of compensation need to be made for the length of the ski, or is this implied?
4. Are the weights I suggested the right amount? Too much? Too little?
5. Since many skis have asymmetrical flex (tip is softer than tail) do we need to include a measurement that represents a ski with the center and forward clamped to a bench and the back end hanging free with a weight and vice versa?
And finally the most important question:
Is this something anyone else even cares about. I have yet to finish my press but I already wish I could know how stiff some of the skis I see on this site are. As a consumer of skis I am frustrated by constantly hand flexing skis, forming an opinion, and then having friends completely disagree base on how they remember the ski. The only real way to know how a ski flexes is to ride it, but since that is often impractical a system like this might really help. I also think that it would be a feather in the garage builders' cap if we could cooperate and develop a system while the big boys failed to do so. However if the whole idea seems a little harebrained let me know.
Accurate, replicable method of measuring ski flex (long)
Moderators: Head Monkey, kelvin, bigKam, skidesmond, chrismp
powdercow,
Oooh, this is a good one. Okay, grab a ski and find the narrowest point on the ski... that's the sidecut center. In theory, this is the point where, if weighted and edged on snow, the ski will form a sidecut arc radius that is equal in front of and behind the weighted point, so the front of the ski and the rear of the ski are headed for the same destination in the turn... which is a good thing. Mark that spot on the ski. Next place the ski in the measuring device that you describe, and weight as you suggest. However, don't be concerned about the depth of the arc, just try to find the point on the ski where the 'flex arc' in front of the weighted point matches the flex arc behind the weighted point. I'll interject here that I have a device very similar to the one that you describe. But between my two upright supports, I have a piece on wood that has been cut in a perfect circular arc (19 ft radius). I adjust the arc and the weight applied to the ski to find where the flex arc center is on the ski. Okay, in theory, this is the point where, if weighted and edged on snow, the ski will form a flex arc radius that is equal in front of and behind the weighted point, so the front of the ski and the rear of the ski are headed for the same destination in the turn... sound familiar. Mark that spot on the ski... and yes, you're hoping the two marks on the ski line up with on another. If they don't, the ski was not designed very well and it isn't going to turn very smoothly.
Okay, now regarding varying the applied weight on your device. You mentioned that a ski's flex varies progressively as the weight is increased... commonly true, but not necessarily a good thing. What you really want is for the ski's flex arc to remain equal front and rear, no mater how much weight you apply to it... that way the front and the rear of the ski are always headed to the same destination. But just try to get an equal front/rear flex arc on a factory built cap ski... no way because the waist of the ski is too stiff... and the more you weight it, the worse it gets. A sandwich ski that is not profiled properly will give a similar result. There are dozens of interesting patents that are intended to mitigate this problem. If you look at current production skis, you'll see all kinds of design interventions that are aimed at reducing the negative effects of a waist that is too stiff as compared to the fore and aft bodies... very thin waist profiles (getting so thin that it's difficullt to get a binding screw to hold) top sheet ridges/elevations that are designed to stiffen the tip and tail sections to more match the waist. So, when it comes to ski 'flex', I only have one personal criteria... that the flex arc is equal in front of and behind the weighted point of the ski no matter how much weight is applied, and that it forms a true circular arc. More weight just means a tighter radius turn.
As far as overall ski 'stiffness' goes, I just say soft/wide skis for soft conditions, stiff/narrower skis for firm conditions, and lots of variation in between. A firm ski in soft conditions just isn't going to get enough resistence from the snow to get it to flex and perform. A soft ski in firm conditions isn't going to get any pressure to the tip and tail edges, and that'll suck also. And, yes, I agree that a ski's 'stiffness' varies progessively with an increase in load, primarily because the ski reaches it's elastic limit.
So, in answer to some of your questions, I personally don't think in terms of a 'middle' of a ski. The sidecut arc/flex arc center on my templates is labled boot center, mostly because I haven't come up with a more creative lable. I think that an assemetrical flex is pretty undesirable... it just means that you'll be riding a ski that is trying to go to two different destinations. After sking on my own skis, when I go back to a production ski, I can easily feel that happening, especially in varied snow conditions.
"Does anybody else even care about this?" Hell, yes!
Oh, by the way, the above is just all my own personal opinion and may be complete garbage to someone who really knows what they're talking about.
Cheers
G-man
Oooh, this is a good one. Okay, grab a ski and find the narrowest point on the ski... that's the sidecut center. In theory, this is the point where, if weighted and edged on snow, the ski will form a sidecut arc radius that is equal in front of and behind the weighted point, so the front of the ski and the rear of the ski are headed for the same destination in the turn... which is a good thing. Mark that spot on the ski. Next place the ski in the measuring device that you describe, and weight as you suggest. However, don't be concerned about the depth of the arc, just try to find the point on the ski where the 'flex arc' in front of the weighted point matches the flex arc behind the weighted point. I'll interject here that I have a device very similar to the one that you describe. But between my two upright supports, I have a piece on wood that has been cut in a perfect circular arc (19 ft radius). I adjust the arc and the weight applied to the ski to find where the flex arc center is on the ski. Okay, in theory, this is the point where, if weighted and edged on snow, the ski will form a flex arc radius that is equal in front of and behind the weighted point, so the front of the ski and the rear of the ski are headed for the same destination in the turn... sound familiar. Mark that spot on the ski... and yes, you're hoping the two marks on the ski line up with on another. If they don't, the ski was not designed very well and it isn't going to turn very smoothly.
Okay, now regarding varying the applied weight on your device. You mentioned that a ski's flex varies progressively as the weight is increased... commonly true, but not necessarily a good thing. What you really want is for the ski's flex arc to remain equal front and rear, no mater how much weight you apply to it... that way the front and the rear of the ski are always headed to the same destination. But just try to get an equal front/rear flex arc on a factory built cap ski... no way because the waist of the ski is too stiff... and the more you weight it, the worse it gets. A sandwich ski that is not profiled properly will give a similar result. There are dozens of interesting patents that are intended to mitigate this problem. If you look at current production skis, you'll see all kinds of design interventions that are aimed at reducing the negative effects of a waist that is too stiff as compared to the fore and aft bodies... very thin waist profiles (getting so thin that it's difficullt to get a binding screw to hold) top sheet ridges/elevations that are designed to stiffen the tip and tail sections to more match the waist. So, when it comes to ski 'flex', I only have one personal criteria... that the flex arc is equal in front of and behind the weighted point of the ski no matter how much weight is applied, and that it forms a true circular arc. More weight just means a tighter radius turn.
As far as overall ski 'stiffness' goes, I just say soft/wide skis for soft conditions, stiff/narrower skis for firm conditions, and lots of variation in between. A firm ski in soft conditions just isn't going to get enough resistence from the snow to get it to flex and perform. A soft ski in firm conditions isn't going to get any pressure to the tip and tail edges, and that'll suck also. And, yes, I agree that a ski's 'stiffness' varies progessively with an increase in load, primarily because the ski reaches it's elastic limit.
So, in answer to some of your questions, I personally don't think in terms of a 'middle' of a ski. The sidecut arc/flex arc center on my templates is labled boot center, mostly because I haven't come up with a more creative lable. I think that an assemetrical flex is pretty undesirable... it just means that you'll be riding a ski that is trying to go to two different destinations. After sking on my own skis, when I go back to a production ski, I can easily feel that happening, especially in varied snow conditions.
"Does anybody else even care about this?" Hell, yes!
Oh, by the way, the above is just all my own personal opinion and may be complete garbage to someone who really knows what they're talking about.
Cheers
G-man
About measuring the stiffness, I do in the following way: I put one ski between two wooden blocks, at the same distance from the ski centre, then I stand on the ski (I weight about 70Kg) and I measure the deformation (see figure below). I do that for three distances, D=70 cm, 90 cm and 110 cm.
In this way I can measure the stiffness during the fabrication steps, and adjust the construction to get the right flex.
Here are some results: http://tromso.isuisse.com/aeg/table.jpg
I did few more test since then: I can say that the stiffest ski around is about twice stiffer then the softest one I made. Probably the stiffess of most of the skis lay in a 20%-30% range.
In this way I can measure the stiffness during the fabrication steps, and adjust the construction to get the right flex.
Here are some results: http://tromso.isuisse.com/aeg/table.jpg
I did few more test since then: I can say that the stiffest ski around is about twice stiffer then the softest one I made. Probably the stiffess of most of the skis lay in a 20%-30% range.
You guys write such long posts and I have so little time, but this is VERY interesting, thanks for bringing the subject back on track.
The last year I have been testing skis together with a norwegian ski magazine (fri flyt). I have been frustrated about the lacking info of flex characteristics of skis, so as a part of my masters degree last spring i designed and built a flextester. The basic principle of the test is to make a testing method which is easily repeatable by everyone, so that we can communicate these things over internet.
The flextester i built is not simple at all, but the plinciple is very simple. I wanted to find stiffness in separate parts of the ski, so i chose a distance of 10cm. I get stiffness results in as many parts of the ski as i am interested in. After testing i get a curve of the whole ski. The results are very accurate, I can even see traces of the binding screw holes (i have to remove the bindings)
When I measure the ski, I clamp the ski sideways (so gravity don't affect the result, apply a known force (wheight distributed by wire transmission). This force is applyed more than 10 cm. from where the ski is clamped. Then I measure how much the ski has flexed at 10 cm. from where the ski is clamped. The result i want is a common term of stiffness, force/length, I use Newton/mm.
I will put some pictures og the testing and graphs of the 135 pairs i have tested so far when i get time.
The last year I have been testing skis together with a norwegian ski magazine (fri flyt). I have been frustrated about the lacking info of flex characteristics of skis, so as a part of my masters degree last spring i designed and built a flextester. The basic principle of the test is to make a testing method which is easily repeatable by everyone, so that we can communicate these things over internet.
The flextester i built is not simple at all, but the plinciple is very simple. I wanted to find stiffness in separate parts of the ski, so i chose a distance of 10cm. I get stiffness results in as many parts of the ski as i am interested in. After testing i get a curve of the whole ski. The results are very accurate, I can even see traces of the binding screw holes (i have to remove the bindings)
When I measure the ski, I clamp the ski sideways (so gravity don't affect the result, apply a known force (wheight distributed by wire transmission). This force is applyed more than 10 cm. from where the ski is clamped. Then I measure how much the ski has flexed at 10 cm. from where the ski is clamped. The result i want is a common term of stiffness, force/length, I use Newton/mm.
I will put some pictures og the testing and graphs of the 135 pairs i have tested so far when i get time.
Powdercow,
I think the reason there is not a standard flex measurement is that it is too complicated to translate the data into a purely quantitative useful number. While I think your method of using multiple weights to explore the skis deflection is better than a single weight there are still a few factors that can affect your results in a big way.
First, the method does not account for where the deflection is occurring. You could have 2 skis that show the same deflection numbers with one having a very uniform curvature and a second one that is stiff under foot and soft in the tip and or tail. In other words, 2 skis that would ski very differently. To account for this you would have to take measurements along the entire length of the ski. So now you have gone from 3 data points to closer to 100.
The next issue is the effect of length on perceived stiffness. Take 2 skis of different lengths. Under the same weight they could bend to identical curved shapes but because one ski is longer, its displacement values will be higher.
In writing this I have realized that perhaps a numerical value for stiffness would have only minimal value, but a graph of a ski's deflected shape for a given load/loads, that is normalize for length, would be a useful way of comparing skis.
edit: Looks like endre types faster than me.
I think the reason there is not a standard flex measurement is that it is too complicated to translate the data into a purely quantitative useful number. While I think your method of using multiple weights to explore the skis deflection is better than a single weight there are still a few factors that can affect your results in a big way.
First, the method does not account for where the deflection is occurring. You could have 2 skis that show the same deflection numbers with one having a very uniform curvature and a second one that is stiff under foot and soft in the tip and or tail. In other words, 2 skis that would ski very differently. To account for this you would have to take measurements along the entire length of the ski. So now you have gone from 3 data points to closer to 100.
The next issue is the effect of length on perceived stiffness. Take 2 skis of different lengths. Under the same weight they could bend to identical curved shapes but because one ski is longer, its displacement values will be higher.
In writing this I have realized that perhaps a numerical value for stiffness would have only minimal value, but a graph of a ski's deflected shape for a given load/loads, that is normalize for length, would be a useful way of comparing skis.
edit: Looks like endre types faster than me.
by the way, after testing a lot of skis i have not found very much progressive flex charactreistics, a little bit in some skis with very mixed materials, but to be shure about the compareability of my results I allways test at room temperature and the same load, 25kg. The results I get vary from around 30-1000 N/mm, depending where on the ski i flex.
Some pictures of the machine I built to test flex patterns
Last edited by endre on Mon Oct 23, 2006 11:44 am, edited 9 times in total.
this diagram shows 12 pairs of next year's superfats. These curves was printed in a skitest in the norwegian ski magazine "Fri Flyt" last month.
Last edited by endre on Mon Oct 23, 2006 11:42 am, edited 3 times in total.
i have all thicknesses to davide, i'll publish it in a while. by the way:can you see the pics? they don't seem to work on my computer anymore
Last edited by endre on Mon Oct 23, 2006 7:04 am, edited 1 time in total.
I think the previous posts highight some very good ideas and they are techniques I've seen used in the industry.
In the heat of welding, bolting and working with springs its easy to forget what you want to test for. So, take breaks and rethink as you build.
Building solid and acurate tesing equipment is within the ability of the average thinking man or woman.
In my experience, most machines only attempt to duplicate what you can figure out with your own senses, arms, hands and legs. But, they allow testing which is probably more repeatable over time--at least people like me that love empirical data like to think so. Usually, that is good enough.
In my own experience, the best testing apparatus I've used to test for flex to failure was a 6'2" Berzerker---he could break anything you handed him. So, it made my machine building skills a bit obsolute in that respect.
With him I not only got a good failure test I got a qualitative one.
In the heat of welding, bolting and working with springs its easy to forget what you want to test for. So, take breaks and rethink as you build.
Building solid and acurate tesing equipment is within the ability of the average thinking man or woman.
In my experience, most machines only attempt to duplicate what you can figure out with your own senses, arms, hands and legs. But, they allow testing which is probably more repeatable over time--at least people like me that love empirical data like to think so. Usually, that is good enough.
In my own experience, the best testing apparatus I've used to test for flex to failure was a 6'2" Berzerker---he could break anything you handed him. So, it made my machine building skills a bit obsolute in that respect.
With him I not only got a good failure test I got a qualitative one.