To learn and share about dampers / shock absorbers

[Belatti]

I´m creating this thread to hopefully learn and share info about the called "black art" of damper tunning and development.

Yes! I know nowadays important racing teams have post rigs and stuff, maybe even they are applying CFD to hydraulic/gas dampers design and tunning, too.

But down here there are still newbies like me or DIY racers that want to know more about the ways to engineer your own dampers, using very limited resources like a damper dynamometer and the will and time to work.

I guess the first (and bold?) step would be to make a simple 1/4 car model, where we will input motion ratios, suspension geometries, masses, hopefully a guess of tyre stifness and we will have an initial aproximation of the damping curves we want to have, hopefully aided by longitudinal and lateral accelerations measured by a datalogger that could help us to imagine how fast does the weight transfers occur.

Just then we could go to a dyno to try to get what we want.

But there are other things to consider!
In the dyno curves, wich graph do we have to pay attention to and for what?

These are some examples of what we can see in some machines:



Then there are other (the mechanical) considerations we have to consider. What about hysteresis, dry and fluid frictions contribution, gas preloads, temperature build up, oil seals, etc.

I let the thread for all of you to give and take, but please try not to fill it with Red Bull dampers in Q3 speculations, magneto rheological witchcrafts, Mosley´s m...ass damping or why did Alonso and Hamilton fight each other over Schumacher´s girlfrien Corinna, yes, the ex of Heinz Harald. Keep it real guys

[Belatti]

and lets start a database:


The ABC (so the really really newbies dont flood the thread with Qs like "Whats a damper?"):
http://en.wikipedia.org/wiki/Shock_absorber

Info about springs:
http://www.eibach.de/

Racing Damper manufacturers:
http://www.penskeshocks.com/Home.php
http://www.ohlins.com/
http://www.koniracing.com/
http://www.multimaticmotorsports.com/dy ... press.html

Shock dynamometer manufacturers:
http://www.pronello.com.ar/english/english.html
http://www.roehrigengineering.com/

[747heavy]

there are other ways to display damper data, which include frequency and/or acceleration.

On an standard damper dyno (yoke drive, sine wave input), I think a force vs. displacement graph (called football or potato graph by some) is the most useful - IMHO.

As you see, that in your example the damper reacts different at the some speed for accelaration vs. deccelaration, leading to a quite high hysteresis in the upper part of the diagram ( I assume it´s the bump side ? ).

So you may have a stiffness "issue" here, can be air inside the damper or insufficient gas pressure compare to the bump damping forces, flex in the damper dyno loadcell or flex of the connection hose, if we talk about a remote canister damper.
It´s basicly a typical graph for a "spring-force" on top of your "damping-force".

This does not need to be a bad thing, on the track, but you will need to keep it in mind, and it makes it a bit difficult to control this characteristics.

some examples for alternative damping graphs.





some reading if you like:
http://www.autofocusasia.com/engine_cha ... tuning.htm

[747heavy]

if you want to model a damper / damperperformance mathematical, or built/tune your own dampers, you may find some of this interesting:

http://repository.tamu.edu/bitstream/ha ... hoades.pdf

http://deepblue.lib.umich.edu/bitstream ... report.pdf


not 100% related to your question, but maybe you find it useful nonetheless:

http://www.theoryinpracticeengineering. ... 04v001.pdf

[Jersey Tom]

You could write a simple vehicle dynamics sim to assess the impact of low speed damping on transient maneuvers...

[strad]

On an standard damper dyno (yoke drive, sine wave input), I think a force vs. displacement graph (called football or potato graph by some) is the most useful

I thought you wanted a "D" shape....

[xpensive]

You could write a simple vehicle dynamics sim to assess the impact of low speed damping on transient maneuvers...

It's been many moons since university, but can someone give a brief introduction to damping theory to begin with?

You know, x-dot and x-dot-dot, differential equations and that kinda stuff, jon-mullen perhaps?

[747heavy]

On an standard damper dyno (yoke drive, sine wave input), I think a force vs. displacement graph (called football or potato graph by some) is the most useful

I thought you wanted a "D" shape....

Hi Strad,

I was talking about the general graph (force vs. displacement or force vs. position)
this graph gives you the most comprehensive imformation, as it shows the transient behavior of the damper, and not just the peak values.

What the graph looks like, will depend on the characteristics of the damper. e.g. linear, digressive or progressive damping.
It can look like a "D", if this is good or not will depend from the vehicle, and what you would like to achieve with your choosen damping.

a force vs. displacement diagram for two different dampers.




I guess the first one looks more like a "D", if this is what you mean
and the second one more like a "football/potato" cut in half.
This is why some damper/racing people call it "the football/potato" graph/diagram
I did not wanted to make a reference to the specific shape, just to the way the forces are represented/shown

[RacingManiac]

On an standard damper dyno (yoke drive, sine wave input), I think a force vs. displacement graph (called football or potato graph by some) is the most useful

I thought you wanted a "D" shape....

All application dependent and depends on the style of the design intent....

Most passenger car damper would tend to have asymmetrical "football" shape, biasing more rebound than compression. But that probably won't work if you are designing something to do jumps and stuff...

Typically for work here, the 2 graph that we usually look for is the Peak force vs Peak Velocity graph. There are some weird speed that seems to be the standard that all automotive dampers are tested and benchmarked at. This is typically the graph you can see the "knee" at the damping curve, which shows the separation and transition between high speed and low speed damping. We also use the football curve, which can be useful to identify issues with damper, in lags, cavitation or over/under fill of fluid.

None of the dyno at work are "yolk" drive now, hydraulic actuator gives you more flexibility, and also allows for non-sinusoidal testing. A typical dyno run here will also measure gas force and damper friction at different point of the travel, and the program can process them out of the data if you choose.

[DaveW]

Books have been written about this subject, Belatti. Well... one that I know of, anyway: "The Shock Absorber Handbook" by John Dixon ISBN 978-0-470-51020-9.

747 has already posted good advice & useful references. I wish to add (controversially, perhaps) that a peak force-peak velocity trajectory (your plot number 7) compiled from a conventional dyno test generally conveys little useful information about a real damper - as your example demonstrates - despite the fact that it is, I'm told, an "industry standard".

[747heavy]

depending on the type of damper you use (not the manufacturer, but the working principle), you will need to keep this in mind:

http://www.penskeshocks.com/files/TSPre ... ng2-04.pdf

You need a certain relationship between the valving of your piston vs. the valving of your external canister/reservoir or head/foot valve in your damper.

As in how much force/damping you create around/with your piston vs. your external bump adjuster valve.

The amount of gas pressure you have to use, will depend on the amount of bump damping you want to use.
If the gaspressure is too low, you will end up with cavitation inside your damper at high shaft velocities/high bump damping forces.

[marcush.]

gas pressure needs are an indication of the systems design ..you have to avoid cavitation as this destroys the damper quite quickly and of course you will have not the damping as well.
A high pressure will of course reduce damper sensivity as the stickslip will be increased.So the gas pressure is used to keep the damper from cavitating not to tune the car as some people suggest .

some product (from the blue yellow country) is able to use very low pressure without cavitation.

[DaveW]

Some musings on dampers:

Damper "hysteresis" is usually caused by compression of the fluid & damper structure as the fluid pressure changes. Many authorities state that the compliance of the damper structure can be neglected, & my experiments suggest that is the case, unless an "active" remote reservoir is connected to the damper body by a hose (in that case all bets are off). It follows that fluid compression is the principal culprit, even though fluid bulk modulus must be reduced by at least an order of magnitude (I recall) in order to replicate observed levels of damper hysteresis with any accuracy. That observation suggests that gas, either as a discrete inclusion or suspended in the fluid, is the major contributor to hysteresis, as 747 suggested.

Damper "hysteresis" can be modelled reasonably accurately by a spring placed in series with a "pure" damper. The stiffness of the spring is usually direction-dependent. For a given "build", the deflection of the series spring will be proportional to the change in pressure required to achieve a damper force. In other words spring stiffness will be proportional to piston cross-sectional area. Typical (average) series spring stiffness values for dampers are 6.5 KN/mm for a 44mm diameter piston, falling to around 2.5 KN/mm for a 29mm diameter piston.

747 has stated that hysteresis is not necessarily a bad thing, & he is correct. However, vehicle designers like dampers to be small & light, & they sometimes achieve that by choosing a small diameter damper & installing it with a high motion ratio (MR) to reduce required damper stroke (MR defined as wheel velocity per unit damper velocity). If the MR is 1.5, say, then the effective series spring stiffness for a 29mm piston damper reduces to around 1 KN/mm.

Now, damper compliance is not the only flexibility in a suspension. The "installation" stiffness of an open-wheeled vehicle by itself can be as low as 1 KN/mm at the rear axle. In that case, a small damper & high motion ratio can reduce the effective stiffness of the series spring to around 500 N/mm. Alarm bells should be ringing by now.

The effect of series compliance is to reduce the "efficiency" of a damper, and the reduction will increase with increasing frequency. It follows, both logically & by observation, that the signature of a low efficiency damper is an uncontrolled hub mode. The lowest value of stiffness I have estimated on a rig was around 350 N/mm at the rear axle & that vehicle had no hub control. On-track, lack of hub control caused "snap oversteer" events. "Junk" dampers? Perhaps, but they were only obeying the laws of physics....

[marcush.]

i think this covers a lot of what you need to know .

http://www.optimumg.com/OptimumGWebSite ... _Tip_1.pdf

http://www.optimumg.com/OptimumGWebSite ... _Tip_2.pdf

http://www.optimumg.com/OptimumGWebSite ... _Tip_3.pdf

http://www.optimumg.com/OptimumGWebSite ... _Tip_4.pdf

http://www.optimumg.com/OptimumGWebSite ... _Tip_5.pdf

http://www.optimumg.com/OptimumGWebSite ... _Tip_6.pdf

[DaveW]

It is possible to extract reasonable guesses of installed damper trajectories from a whole vehicle rig test.

Here are sample load-displacement trajectories for the rear axle of an open-wheeled race vehicle. That shown in red includes (overall) "hysteresis". That shown in green is the same thing, after it has been "corrected" by removing the effect of a 1.3 KN/mm series spring (Kis). Here are the corresponding load-velocity trajectories. Note the difference in estimated damping coefficient (Coef) for the two versions - hence "efficiency". The observable "glitch" in the green trajectory at low velocities is probably caused by compliance/backlash in the various suspension bearings.