To learn and share about dampers / shock absorbers

[mlubbock]

Dave,
exactly. One minor addition...
The reservoir in the assumed architecture (3-Way) is also complicated by by two shim stack arrangements. The secondary preloads the primary via the 'high-speed' adjuster. In practice the primary shim pack has all kinds of wacky deformation modes when under preload (due to the adjuster cage preload acting outboard of the hydraulic ports). Coupling series compliance with interleaf friction returns the hysteresis displayed.

As Dave mentioned, it is advised to optimise damping tracjectories on the main piston (keeping system pressures low) and use the remote reservoir for subtle balance tweaks (wet/dry or subtle changes to transient response etc etc).

[Eager Learner]

Hi, I've heard from a certain well known racing damper manufacturer that has had good results for many years in high level GT racing, that absolutely minimizing hysteresis is not always a good thing and that one of the main reasons people like to reduce hysteresis is that it makes damper simulation much easier if hysteresis is assumed to be 0.

Is there any truth to this statement at all? Or is it just an excuse?

[Eager Learner]

"The solution to your question, assuming you want to minimize hysteresis (also a good idea, I think), is do most of the damping on the piston, leaving just enough adjustable damping for your needs."

Hi Dave, sorry for the basic question but I'm a beginner. How does one do more or most damping on the piston instead of at the reservoir valve? If a damper has no low or high speed compression adjuster at the piston like say Motons or MCS, doesn't this mean that all compression damping is being away from the piston? Does this mean that dampers of this design are inferior?

[Jersey Tom]

Hi, I've heard from a certain well known racing damper manufacturer that has had good results for many years in high level GT racing, that absolutely minimizing hysteresis is not always a good thing and that one of the main reasons people like to reduce hysteresis is that it makes damper simulation much easier if hysteresis is assumed to be 0.

Is there any truth to this statement at all? Or is it just an excuse?

Is a damper without hysteresis easier to model or simulate? Sure. That's common sense - less complexity makes things easier, right? Is a damper with hysteresis "better" than one without? I don't think there's a clear cut answer to that.

As a preface to this next bit, let's just put out there right now - no racecar is ever "optimized." Never the best it can ever be. That's why you have practice sessions every weekend, and why cars get better over the course of a season. You'll never have a 100% complete understanding of everything there is.

Let's say you're a race team, and you've got two options for dampers. One is a pretty simple design - maybe it's linear with little or no hysteresis. The other is more complex, non-linear, with hysteresis. And let's say for argument's sake the more complex one has the most ideal world potential to be best.

What's going to be the best option for you - the damper with 100% most potential performance that you only understand 70% of... or the damper with 90% potential that you understand 95% of?

Now maybe your main competitor has a few super shock gurus on their team and they can understand 95% of any option out there. Best choice for them might be different than the best choice for you!

It's like that for a lot of things. Dampers, springs, tires, aerodynamics, kinematics and compliance, etc etc. Many different options on how to approach each of them, and there's no globally best choice. That's what makes engineering so interesting.

[Eager Learner]

Thanks Tom. I just can't imagine a situation where one would want any hysteresis at all. Can anyone describe briefly how hysteresis might be used to an advantage?

[DaveW]

Hi, I've heard from a certain well known racing damper manufacturer that has had good results for many years in high level GT racing, that absolutely minimizing hysteresis is not always a good thing and that one of the main reasons people like to reduce hysteresis is that it makes damper simulation much easier if hysteresis is assumed to be 0.

That damper manufacturer is probably trying to sell you something....

"Hysteresis" is used to describe compliance and shortcomings in the design of a damper that can be modeled as a series spring acting on a "pure" damper. Zero hysteresis implies infinite stiffness, and perfect switching, etc. Even if that ideal was achievable, mounting the damper in a vehicle would (inevitably) introduce additional series compliance. It follows, I think, that the simulation of a suspension must include a series spring if the simulation is to be representative.

Damper hysteresis simply consumes some of the stiffness "budget" a good designer will (or, at least, should) specify.

Arguably, hysteresis increases dynamic spring rate, and decreases effective damping coefficient. It therefore makes the damper less "efficient". That will not necessarily cost performance is moderation, but it will ultimately (and perhaps unexpectedly) effect vehicle control and "grip". As such hysteresis is never a "good thing".

[DaveW]

How does one do more or most damping on the piston instead of at the reservoir valve? If a damper has no low or high speed compression adjuster at the piston like say Motons or MCS, doesn't this mean that all compression damping is being away from the piston? Does this mean that dampers of this design are inferior?

A shimmed damper with all compression damping generated at the reservoir will create large and unsightly compression side hysteresis as the damping is increased. Much more, with examples, has been covered in this thread....

A solution would be to run with rebound biased damping, increasing static ride height to avoid hitting end stops. Back-back tests a few years ago indicated that the solution, compared with a slightly compression biased alternative cost about 1.6 seconds a lap at Mugello.

[hardingfv32]

Definition of rebound/compression biased shock solution, please.

Brian

[DaveW]

Definition of rebound/compression biased shock (damper) solution, please.

This plot shows the rear damper force/velocity trajectories. The sign convention is positive compression. That shown in red is rebound biased, whilst that shown green is (slightly) compression biased.

The corresponding damper position time histories are shown here. Again, positive displacement is compression. Hopefully, you can see that the rebound biased damper used a maximum compression deflection of 16mm. with this input, whilst the compression biased damper used a maximum of just over 6mm.

[MadMatt]

Like I said in the other post... these values, be it % Rebound vs % Comp, or % High Speed vs % Low Speed.. are just identifiers and indicators. They are useful because they are simple and don't require any in depth system knowledge.

The idea is you identify what range of parameters work well, be it for your car, your driver, or a given track. Then the next time you go somewhere, you know "Ok my combination of things likes 50/50 Comp/Reb and 70/30 Low/High speed.. I should aim for these splits in my setup". No real science behind it.

I suppose you could say there is SOME logic... ie matching reb & comp so that the car doesn't "jack down" over bumps, etc. However, there are some cases where having the car ratchet itself down is a good thing...

Had to dig this because I was looking for an answer like this with which I agree 100%. However, I still think there is room to quantify what a damper should be at a given race track and condition.

[theTTshark]

Had to dig this because I was looking for an answer like this with which I agree 100%. However, I still think there is room to quantify what a damper should be at a given race track and condition.

If we lived in a world where we could understand ever minute detail of the race car/track combo plus have drivers that all drove exactly the same every single time, we MIGHT be able to say there is one damper per corner that will be perfect. The problem is we aren't close to any of these. The most realistic way of quantifying it is that there is a rough area that your dampers should operate in for a car/track combo. But then things like weather, driver, tire inconsistencies, other setup changes, etc all through a big wrench into things. I've seen in all forms of racing two drivers with two very different damper settings go and produce a lap time almost identical to one another. Motorsports is constantly trying to head towards perfection the problem is that the human behind the wheel will likely never allow you to believe you're there.

[Yawpower]

I believe that "negative hysteresis" was mentioned earlier in this thread. Possibly this image will spark further conversation.

[Greg Locock]

I think that's friction/compliance.

Your plot is very low velocities, the behaviour between + and -2 inches per second is often grossly non linear.

[Yawpower]

I think that's friction/compliance.

I agree that below 1.25 ips the spread can be attributed to static friction, but what about above 1.25 ips? The damping force at a given velocity is greater while accelerating than while decelerating. The opposite of what you would expect from compliance.

[Greg Locock]

I don't build super detailed component based models of shocks, but I see a very similar effect when plotting steering system 'compliance', from models that are assembled part by part.

<deleted bit that was misleading>

I think the most likely reason is backlash, whether physical or hydraulic (gas), but it could be more complex things like an interaction between non linear springs and friction. I can get the same effect with a simple model if I put a negative stiffness in series with the damper, but that is a bit silly.