Dyno data –And what it tells us about how to tune a shim stack and control the shape of the damping force curve

Dyno Data on TT

You'd think dyno data would be all over the place in suspension forums comparing shim stacks, bike setups and year-to-year suspension tuning changes. Turns out dyno data is pretty rare. If you dig around there are a few dyno runs scattered through the TT suspension forum.

This thread is an attempt to use the collective knowledge of TT to evaluate that dyno data and then try to figure out what the dyno data tells us, if anything, about how to tune a shim stack.

Dyno data found so far:

If you know of any other dyno data here on TT? Post a link!

Edited by Clicked

The dyno data on TT covers a wide range of valve port geometries:


1a-TT-dyno-valves.png


 


... Damping force values from the dyno tests cover a factor of six in tuning range. There is data for crf, kx, rmz and yz shocks dyno tested up to shaft velocities of 120 in/sec.


2-dyno-range.png


... and a bunch of different shim stack configurations.


3-stacks-all.png


The goal is to use the collective knowledge of TT and try to figure out:



  • How to use shim factors to evaluate a shim stack




  • If shim factors have any relationship to damping force measured on the dyno




  • How different styles of crossovers behave




  • How changes to the high speed stack effect damping




  • How to control the shape of the damping force curve




  • How damping of a straight stack compares to a tapered stack



Got any other dyno data? Post up a link!


Shim Factors and How To Use Them

  • Ever heard of shim factors?

  • Know how to use them?

  • Do they actually work?

MXScandinavia had those questions and ran a series of dyno tests to figure out if the shim factor tables on the MX-Tech web site were accurate and if those shim factors had any relationship to damping force measured on the dyno.  

Shim factor theory comes from the Almen and Laszlo Belleville spring equations. The equations were developed by the two General Motors engineers back in 1936. The Schnorr Handbook shows the full version Almen and Laszlo equations. The equations show shim stiffness has a complex dependance on the shim OD, ID, edge lift and thickness. 

1-belleville.png

That theory is simplified by shim factors by assuming the stack clamp diameter, shim diameter and edge lift are the same for all of the shims in the stack. That simplification makes the stiffness ratio (k.2/k.1) between two shims a simple function of thickness cubed.

2-shim-factor.png

The thickness cubed theory used by shim factors figures out the stiffness of a shim relative to the stiffness of a 0.1 mm thick reference shim. Using the equation above you can make your own shim factor table and check it against the values on the MX-Tech web site.   

3-sf-table.png

For shim factor theory to work all of the shims in the stack must have:

  • The same ID

  • The same OD

  • Be made out of the same stuff. Steel and stainless steel have a different stiffness.

  • And, have the same edge lift in terms of the dimensionless ratio h.1/t.1 = h.2/t.2

That edge lift requirement is tricky. Shims that are twice as thick have to be lifted twice as high to have the same edge lift (h) to thickness (t) ratio. Doubling the edge lift means the fluid flow area at the stack edge is larger so the damping force is going to be a whole lot lower. For shim stack comparisons you really want the same edge lift, like h.1 = h.2, not at the ratio of edge lift to thickness (h.1/t.1) required by shim factors. So, the limitation is confusing.

Almen and Laszlo use the edge lift ratio (h/t) and the OD/ID diameter ratio to define non-linear shim stiffness effects. Shim factors ignore those effects and only use thickness cubed. That simplification leads to errors when comparing shim factors for different stacks. MXScandinavia wanted to figure out how large those errors were and if it made any differences in damping force measured on the dyno.

Edited by Clicked

Shim Factor Theory Testing

To test shim factors MxScandinavia replaced a stack of 14x40.2 face shims with a theoretically equivalent shorter stack of 0.3 mm shims and then dyno tested the stacks to see if the real world damping force was the same or not.

  • If you use shim factors to replace a stack of 14x40.2 mm face shims with a theoretically equivalent stack of 40.3 mm shims how close will the damping force be?

The first step is to figure out the shim factor for the baseline stack of 14x40.2 face shims MXScandinavia used. The shim factor of the stack is simply the number of shims (14) times the shim factor for a single 0.2mm thick shim. That works out to 112 for 14x40.2 baseline stack.

4-stack-sf.png

To keep the same stiffness MXScandinavia needed to find a stack of 40.3 face shims that matched the 112 shim factor for the baseline stack. You can't get an exact match as shown in the above figure. Four 40.3 shims is the closest with a shim factor of 108. That is about 4% softer then the 112 shim factor for the baseline stack of 14x40.2 face shims. If you add another 0.3 face shim the shim factor goes up to 135 and that's way too stiff.

So, the closest match using shim factors is 4x40.3 shims. That stack is theoretically 4% softer then the baseline stack of 14x40.2 face shims. What does the dyno say?

Edited by Clicked

Tapered Stack Shim Factor Theory

The replacement stack of 4x40.3 shims is about 4% softer then the baseline stack. But that is only for the face shims. To get some idea of the damping force difference you need some way to estimate the stiffness of the entire stack, not just the face shims.

The Almen and Laszlo equation gives a way to estimate the effect of both shim diameter and thickness. The fraction of the shim under the clamp does not bend. So you need to subtract that portion of the shim deflection. The Almen and Laszlo equations also compute stiffness at the shim edge, but we need the deflection at the valve port edge to estimate damping force. A simple diameter ratio gives an easy way to estimate deflection at the valve port edge.

6-almen-laszlo.png

Substitute in those relations and you get an estimate of the shim deflection at the valve port edge. That stiffness is a spring constant that give the force needed (lbf) to deflect the shim edge 1 mm.

7-tsf-work.png

Sum up the force contribution from each shim in the stack and you get an estimate of the overall stack stiffness. To get even numbers out of that equation the whole thing is multiplied by a million (10^6). That relationship is called the “Tapered Shim Factor” (tsf) for estimating stack stiffness.

8-tsf-eqn.png

The above equation does not work very well. errors are in the +/- 30% range for estimating shim stack stiffness. Hopefully someone will post a better method.......

Applying that equation the baseline stack of 14x40.2 face shims give a tapered stack shim factor (tsf) of 241, the replacement stack of 4x40.3 face shims has a tapered stack shim factor of 237. So the replacement stack is 2% softer. The dyno tests from MXScandinavia is going to show the difference was larger than that.

10-stack-tsf.png

Edited by Clicked

MXScandinavia Shim Factor Dyno Data

From shim factory theory the difference in face shim stiffness is:

  • SF14: Baseline stack of 14x40.2 face shims with a shim factor of sf=112

  • SF4: Replacement stack of 4x40.3 face shims with sf=108

  • The replacement face shims are about 4% softer

 

To figure out the change in damping force you need some way to estimate of the stiffness of the overall stack. Using tapered stack shim factors:

  • SF14: The baseline stack with 14x40.2 face shims has a tapered stack shim factor of tsf= 241

  • SF4: Shim factor scaled stack with 4x40.3 face shims and a tapered stack shim factor of tsf= 237

  • The replacement stack is about 2% softer in overall stack stiffness

Here are the dyno results for the two shim factor scaled stacks from MXScandinavia:

11-sf-dyno.png

 

The replacement 4x40.3 stack was expected to be softer and the dyno shows that. To get the same damping force the 4x40.3 stack needs a shaft velocity that is about 8% higher. That 8% shaft velocity difference implies the stack was about 8% softer, not the 2% expected from shim factor theory. Shim factors missed damping force change by a factor of 4.

Hack around on the above analysis curves and that 8% change in damping force turns out to be around 5 clicks on the bleed screws in terms of a “real world reference” of how those stacks are going to feel on a test ride.

The above dyno tests changed nothing but the face shims on the stack. Shim factors should have nailed it. MXScandinavia posed this question: Why where shim factors so far off for these two stacks?

MXScandinavia ran some more dyno tests to try and figure out why...........

Edited by Clicked

Sweet stuff!  :thumbsup:

 

 

Can't wait 'til I have time to chew on this (that's when the wife walks by, looks over my shoulder at what I'm reading, shakes her head and mutters something that sounds a lot like "geek"  :smirk: ).

 

:cheers:

And not to hijack the thread, but speaking of John Curea, does anyone know what happened to him? He suddenly disappeared some years ago.

 

I'd heard his wife was battling cancer but can't confirm that. If it's true, I hope she got better. John was always a classy guy and willing to help.

 

 

:cheers:

Cool thread. 

 

Okay...I want to make sure that we divide what is real dyno data to that of what is produced with ReStackor. 

 

Are we cool with that?

 

Actually, I don't mind either way. We can call software simulation or math a dyno if we want. But...I just want to make sure that we understand the difference. 

 

Additionally, dyno data is going to be limited not because of the short supply of dynos, but because of the fact that most dynos can't replicate displacement and velocities like what we see in the real off-road world. 

 

So...I hope we have other "dynos" to consider....right? 

We use Butt Dynos down here in the south:)

The above equation does not work very well. errors are in the +/- 30% range for estimating shim stack stiffness.

...

Applying that equation the baseline stack of 14x40.2 face shims give a tapered stack shim factor (tsf) of 233, the replacement stack of 4x40.3 face shims has a tapered stack shim factor of 229. So the replacement stack is 2% softer.

...

The replacement face shims are about 4% softer

... 

The replacement 4x40.3 stack was expected to be softer and the dyno shows that. To get the same damping force the 4x40.3 stack needs a shaft velocity that is about 8% higher. That 8% shaft velocity difference implies the stack was about 8% softer, not the 2% expected from shim factor theory. Shim factors missed damping force change by a factor of 4.

(Snipped for brevity, not my intention to twist words)

If the margin of error on the shim factor equation is 30% (plus possible other sources of variation), should we be analyzing predicted stiffness differences of 4% and 8%? Or do we simply chalk it up to the fact that it's (well) within the error band of our tool, and that as such the predictions were accurate?

With regard to the 4X error in damping force change, I am concerned about the validity of this measure due to the very small nature of the numbers involved (and my previous comment about the accuracy of the shim factor equation). When you consider the variability in all of the individual components, perhaps we can't know for certain exactly how far off we are; in other words, is this amount of error "in the noise" of our measurement technique? For example, variations in shim thickness, diameter, preload, friction, variations in the damper build, and then variations in the dyno runs themselves. Even if these are all individually very accurate/repeatable (within 1% or less), it would not be surprising to see a 5-10% variation in the overall result.

From a statistics point of view, I'd think about it as "would this data pass a T-test" or "what is the p-value for these data sets?", or think about it as a gauge R&R. Of course the data sets are limited and so doing these de rigueur isn't really possible, but I think the questions still express my intent.

It would be interested to test the predictive power of the shim factors for stacks with a more significant change, so that we could neglect the possible variability of some of these items as inconsequential.

Good info on the shim factor equations and this applications, regardless of the aforementioned caveats that's good and useful info, especially the tapered stack stuff. :thumbsup:

And not to hijack the thread, but speaking of John Curea, does anyone know what happened to him? He suddenly disappeared some years ago.

 

I'd heard his wife was battling cancer but can't confirm that. If it's true, I hope she got better. John was always a classy guy and willing to help.

 

 

:cheers:

 

Agree with that. John Curea made a lot of great contributions that helped to build the ThumperTalk site.

 

Same goes for the rest of the guys listed at the top of this thread (Valving Logic, Push Ind, MXScandinavia and Kawamaha). Dyno data from those guys has done a lot to change the discussion on how to go about tuning a shim stack.

Cool thread. 

 

Okay...I want to make sure that we divide what is real dyno data to that of what is produced with ReStackor. 

 

Are we cool with that?

 

Actually, I don't mind either way. We can call software simulation or math a dyno if we want. But...I just want to make sure that we understand the difference. 

 

Additionally, dyno data is going to be limited not because of the short supply of dynos, but because of the fact that most dynos can't replicate displacement and velocities like what we see in the real off-road world. 

 

So...I hope we have other "dynos" to consider....right? 

 

The dyno data we have here on TT simply is what it is. The goal is to try and figure out what that dyno data means in terms of the way clamp shims, crossover gaps and stack taper behaves.

 

If you any ultra high speed data you want to throw into the mix post up a link. Either way I hope you can stick around to help us figure out what the dyno data means in terms of controlling the shape of the damping force curve.

Edited by Clicked

the "valve dynos" I made are a great tool to change a shim stack in the direction you want it to behave.

In summer I don't want to test - I want to ride! In winter I have a lot of time to design stacks - but cannot test. So its great to simulate them

and often the dyno run destroys the illusion of what I was hoping I would get :rant:

 

IMO step one would be to analyze velocities and forces that we have in different situations (bumps, sharp obstackles, jump landings...), then

use the dyno to match the insight we got from the measurements.

 

the good thing - I have a data acquisition unit. the bad thing - ist not portable (to be honest, thats also a good thing, so I am not forced to spend a lot of time

in recording data)

. So I still wait for telemetry data... :ride:

(Snipped for brevity, not my intention to twist words)

If the margin of error on the shim factor equation is 30% (plus possible other sources of variation), should we be analyzing predicted stiffness differences of 4% and 8%? Or do we simply chalk it up to the fact that it's (well) within the error band of our tool, and that as such the predictions were accurate?

With regard to the 4X error in damping force change, I am concerned about the validity of this measure due to the very small nature of the numbers involved (and my previous comment about the accuracy of the shim factor equation). When you consider the variability in all of the individual components, perhaps we can't know for certain exactly how far off we are; in other words, is this amount of error "in the noise" of our measurement technique? For example, variations in shim thickness, diameter, preload, friction, variations in the damper build, and then variations in the dyno runs themselves. Even if these are all individually very accurate/repeatable (within 1% or less), it would not be surprising to see a 5-10% variation in the overall result.

From a statistics point of view, I'd think about it as "would this data pass a T-test" or "what is the p-value for these data sets?", or think about it as a gauge R&R. Of course the data sets are limited and so doing these de rigueur isn't really possible, but I think the questions still express my intent.

It would be interested to test the predictive power of the shim factors for stacks with a more significant change, so that we could neglect the possible variability of some of these items as inconsequential.

Good info on the shim factor equations and this applications, regardless of the aforementioned caveats that's good and useful info, especially the tapered stack stuff. :thumbsup:

10-stack-tsf.png

These two stacks are the same. They have the same clamp, the same taper and the same crossover. The face shims were set by the thickness cubed rule to be the same. With everything same, same, same you don't need shim factors to figure out the stiffness of those two stacks should have been the same.

 

They weren't

 

You could dismiss that as dyno measurement error, but here's the deal. The ReStackor curve below shows the same thing. There is something about that SF14 stack that makes it stiffer then expected and MXScandinavia set out to figure out why.

 

The difference is not large, it's about five clicks. You could just say man-up, twirl the clickers and get on with it. Don't know what MXScandinavia was thinking but I imagine, as a professional tuner, he just got sick and tired of missing his setups by five clicks every time and set out to figure out what in the heck was going on so that he could get stuff right. Ya know, learn someth'n.

 

It's cool he posted his results on TT to share some of his suspension tuning secrets. Not sure anybody caught on to that reading the MXScandinavia thread. I think there is a language barrier thing going on there.

 

So here is the question: What is it about that SF14 stack that makes it stiffer then expected? Got any ideas..............

 

11-sf-dyno.png

Friction

kawamaha, welcome.   :thumbsup:

 

We are going to look at some of your fork hydraulic dyno data latter on.

Friction

 

The ultimate black box.

The SF14 stack has a lot more shims, more surface area and that gives it more friction. Friction is definitely a part of it. There is another reason but we have to look at a few more of MXScandinavia's shim stack configurations to get an understanding of what in the heck that is.

 

 

BTW: For all of the ReStackor analysis curves in this thread the shim friction coefficient has been set to one. Loose that parameter and you can pretty much match anything.

Edited by Clicked

So friction on these stacks is the same,hmm I'm not sure then

SF14 stack has more shims and more surface area so it had more friction. I was agreeing with you.

 

I was just trying to point out the analysis curves used the same friction coefficient for both cases. That does not mean the value of friction was the same. That depends on the surface area and surface loading and other stuff. Friction is a wacky subject. The Schnorr handbook goes into that a bit.

Edited by Clicked

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!


Register a new account

Sign in

Already have an account? Sign in here.


Sign In Now