Real Heel vs 2nd Heel Apples to Apples

[jasonq]

I finally got around to doing and apples to apples compare of the exact same binding with both a real heel and 2nd heel attachment to settle for myself how different they really are vs how different the binding designs that use one vs the other are. Recall that i also wrote a rather simple model to graph binding activity given spring rates, attachment points, etc. My goal was not to figure out which boot attachment point was better or worse, just how they differ in a direct compare, and to help validate my model. The binding parts i used allowed for a moving the cable exit, and i tried a forward and a backwards position. See pictures. I ensured the exact same preload on the springs in both configurations. for those that have been around a while, yes that is a rotte/chouinard 411 front throw.

here is what i felt. pivot forward, not very different with the pivot forward. more different with the pivot backwards. In both cases, with the 2nd heel, more active (as defined by moment required to raise increase angle of boot to ski) with the heel low (~0-40 degrees), less active with the heel high (~>50 degrees). this was much less noticable with the forward pivot than the rearward. I think it is this more initial activity, less final activity that people like.

this sensation match up well with my model. The model says forward pivots in the the 40's that i prefer, they are nearly the same. in the 50-60mm range it seems a lot of people like, it is a noticable, but not huge difference. As the pivot moves back into 60-70 mm that some prefer, the difference become much larger. see graph. each pair of red and green lines is the exact same everything, except for where cable attaches to the boot.

does this mean that the same more initial activity, less final activity cannot be done with a real heel attach? yes, no, maybe. it means 2nd heel naturally falls more in that space. But using caming surfaces (lower initial cable pivot, higher final cable pivot), higher rear attachment, and spring rates, i am able to tune a design that can behave the same. However, 2nd heel has a wider design space for that profile. Also, i used a 26.0 TXP, which is the largest small size binding. A smaller size boot would trend more towards the heel attach since the 2nd heel is relatively closer to the real heel. Same deal with a 26.5 boot, crossing over to the smallest large size binding would put the 2nd heel closer to the real heel.

so after all this, yeah, i don't think i will be doing any 2nd heel designs since for my prefered activity curve, it is basically identical to the true heel, and with the true heel i can run F1, F3, or NTN in the same binding.

[jnicol]

Love the science jasonq!

Maybe I’m not like most people, but I do prefer the initial engagement to be softer, and the binding easier to walk around in. And then I like it to ramp up and 'more to be more' in the business end of the binding’s travel when your knee is bent. So, with your examples I see this is more possible with the attachment at the heel.

Meidjo seems to have achieved a nice progression that I liked (it has been a while since I skied it though). The unique thing about the setup there is the rods are curved and not straight. Perhaps this creates this effect?

[Allan Fici]

Jason, This is very good work and I appreciate all the testing and analytics that you do. I like you ski in 25.5-26 MP boots and this explains how I’m able to get a very similar feel from my Meidjo, Outlaw and TTS bindings. The Meidjo being most active with a 56mm TV pivot vs 48mm for both the Outlaw and my TTS bindings. The Meidjo and my TTS have a slightly higher initial activity off the deck and the Meidjo is more progressive in ending range activity with the 56mm pivot.
Thank you for this data!

[Charley White]

Little help, Jason.... I don't see the adjustability of the pivot point. Seems like it would require un-pictured groove-plates to bolt behind the toe-claw, sending the up-bend of the cable more or less back. Also curious what the units of the vertical axis are and what/how you measured them. (How did I miss this!!)

EDIT---thank you for the explanation here:

https://www.backcountrytalk.com/forum/bc-talk/backside-bar-grill/98215-it-s-really-hard-to-beat?p=98427#post98427

[xmatt]

jasonq Thanks for these graphs.

I am curious about something. The moment should be proportional to the spring constant. So, can't we multiply the y-value of any of these curves by choosing different springs? If that is done, how much difference is there between different locations of the pivot point? That is, if all the curves were normalized, by choosing spring constant, to have the same moment at 20 degrees, for example, how much difference would there be at 40 degrees? It seems like the differences would be a lot smaller in what the curves looked like, especially for the red curves which all seem fairly linear up to around 30 degrees deflection or so.

Also, how did you deal with the effect of preload? The spring exerts a force that is proportional (constant of proportionality being the spring constant) to its initial deflection (which is determined by preload) plus any additional deflection caused by raising the heel (increasing distance from pivot point to either 2nd heel or true heel). So for example one could increase the preload but reduce spring constant to get more initial resistance for a given resistance at higher angles.

Do I understand that the curves you show here are real world data, but you also had some model? In the model, you modeled some choice of preload on the spring?

[xmatt]

Maybe I’m not like most people, but I do prefer the initial engagement to be softer, and the binding easier to walk around in. And then I like it to ramp up and 'more to be more' in the business end of the binding’s travel when your knee is bent. So, with your examples I see this is more possible with the attachment at the heel.

Yes, that is what I like too, though I don't want it to be too much when the knee is bent. So, even though I might describe that as a "neutral feel", maybe that's actually a more "active" pivot point, combined with not too much preload? Except there's a certain minimum required preload to keep the heel throw on, so......maybe not too "active" a pivot point or else everything becomes too stiff?

That is, maybe the best way to achieve a "progressive" feel, meaning less initial engagement for given force at higher boot angles, is to have soft springs combined with a further back pivot point? And what do the curves on 75mm bindings look like?

(I am thinking about this a lot recently, as while I can ski my super-light TTS skis fine, whenever I look at myself skiing them it looks a bit weird compared to what I look like on my Meidjo setup, and I am wondering if some is a different feel in the binding which I am not used to as I don't ski them as much)

[jasonq]

jasonq Thanks for these graphs.

I am curious about something. The moment should be proportional to the spring constant. So, can't we multiply the y-value of any of these curves by choosing different springs? If that is done, how much difference is there between different locations of the pivot point? That is, if all the curves were normalized, by choosing spring constant, to have the same moment at 20 degrees, for example, how much difference would there be at 40 degrees? It seems like the differences would be a lot smaller in what the curves looked like, especially for the red curves which all seem fairly linear up to around 30 degrees deflection or so.

Also, how did you deal with the effect of preload? The spring exerts a force that is proportional (constant of proportionality being the spring constant) to its initial deflection (which is determined by preload) plus any additional deflection caused by raising the heel (increasing distance from pivot point to either 2nd heel or true heel). So for example one could increase the preload but reduce spring constant to get more initial resistance for a given resistance at higher angles.

Do I understand that the curves you show here are real world data, but you also had some model? In the model, you modeled some choice of preload on the spring?

lets see, this is a rather old thread, but...

yes, all the curve will scale with spring rate (but max deflection of the spring will likely change too).

yes, in this apples to apples, if we varied the variety of apples by mixing different spring rates, we could probably get a better match, by whatever measure you wanted to use (mse between curves, area between curves, etc). but the shape of the curves would still be different between real heel and 2nd heel if the rest of the geometry was held constant. for example, i the 82.5mm pivot case, the red curve could use a softer spring and thus cross the green curve at a lower angle, or vice versa.

however, i do also believe that it is possible to design a real heel binding to match a given 2nd heel binding feel, and vice versa. at least within reasonable limits. but the each attach location naturally favors a given curve shape. If you really hunt there is a POLR post somewhere about how i designed the heel throw such that the initial cable angle was the same as if it was a 2nd heel attach.

the model has preload built in. that is why the moment at 0 degrees is not zero. pre-load was a constant deflection since it was apples to apples. And yes, if spring rate was to then change, pre-load could change too. that has the effect of sort of shifting the curve up or down. but there is a limit to preload, either because of mechanics of entering the binding or because of lost functional deflection since you used that spring deflection for preload.

the curves shown were from a simple model, but the model matched with what my sensation was when carpet testing a real world binding. yes some choice of preload which i do not remember, but probably like 5mm.

someday i want to write a more complicated, ideally, more realistic model. but don't really have a need to do so, so keep not doing it.

[xmatt]

lets see, this is a rather old thread, but...

yes, all the curve will scale with spring rate (but max deflection of the spring will likely change too).

yes, in this apples to apples, if we varied the variety of apples by mixing different spring rates, we could probably get a better match, by whatever measure you wanted to use (mse between curves, area between curves, etc). but the shape of the curves would still be different between real heel and 2nd heel if the rest of the geometry was held constant. for example, i the 82.5mm pivot case, the red curve could use a softer spring and thus cross the green curve at a lower angle, or vice versa.

however, i do also believe that it is possible to design a real heel binding to match a given 2nd heel binding feel, and vice versa. at least within reasonable limits. but the each attach location naturally favors a given curve shape. If you really hunt there is a POLR post somewhere about how i designed the heel throw such that the initial cable angle was the same as if it was a 2nd heel attach.

the model has preload built in. that is why the moment at 0 degrees is not zero. pre-load was a constant deflection since it was apples to apples. And yes, if spring rate was to then change, pre-load could change too. that has the effect of sort of shifting the curve up or down. but there is a limit to preload, either because of mechanics of entering the binding or because of lost functional deflection since you used that spring deflection for preload.

the curves shown were from a simple model, but the model matched with what my sensation was when carpet testing a real world binding. yes some choice of preload which i do not remember, but probably like 5mm.

someday i want to write a more complicated, ideally, more realistic model. but don't really have a need to do so, so keep not doing it.

Thanks. Actually, my apples-to-apples was not so much to compare red to green but rather to note that red 82.5 mm at a soft spring constant would probably not be all that much different from red 57.5 at a large spring constant. I kind of always thought "pivot point back means it is hard to lift the heel!", but it seems like one could maybe make a fairly "neutral" feeling binding with a back pivot point but a soft spring. At that point you are trading off spring travel for stiffness, i.e., the soft spring 82.5 needs a softer but larger travel spring. And also there are some limits as to how much you want to change the spring constant, as you want to have enough forward pressure at 0 heel angle to avoid pre-release.

[jasonq]

here is what i think you want.
green = real heel
red and blue = duckbutt

blue and green are the same spring (axl stiffy), preload, and pivot location. aka apples to apples real heel and duckbutt.

red = pivot farther back, then softer spring and more preload in order to match as close as possible.

things to note, the match ain't bad overall, but notice how the red line goes to zero. That means the spring bottomed out. granted i just assumed the deflection of the axl stiffy. but it still shows that to get the match in this case, it takes more spring travel. in this case, when the red line bottomed out, the other two had about 0.45 inches of travel left.

YMMV depending on what you like and are trying to match.