Damper Rates - I give Up!!

[Matt22]

Ok, trying to spec some damper curves for FFord.

Understand that Critical Damping is 2*sqrt(k x M) and that we may want 0.7 x Cr for Rebound and less for Bump.

Just want to check what the k and M are! I'm using k as wheel rate (N/m) and M as Sprung corner weight (kg), this gives 2600 Ns/m but the current dampers have twice this rate.

Help, my head hurts....

Matt

[phoenix]

Originally Posted by Matt22

Ok, trying to spec some damper curves for FFord.

Understand that Critical Damping is 2*sqrt(k x M) and that we may want 0.7 x Cr for Rebound and less for Bump.

Just want to check what the k and M are! I'm using k as wheel rate (N/m) and M as Sprung corner weight (kg), this gives 2600 Ns/m but the current dampers have twice this rate.

Help, my head hurts....

Matt

k is the energy in the spring after it has been compressed by any increment - i.e. spring rate x compression (including any pre-load). M should be the mass the spring/damper sees through any leverages and angles.

[phoenix]

Bye the way, to get 0.7CR damping, use 1.4*sqrt(k x M)

[silente]

Hi Matt!

First of all why you want exactly C/cr=0.7?

I had the same problem some time ago working on a formula 3 car...i found values of c/cr which sometimes arrived to 2.5-3!

I read probably the same papers you did on internet and they say to stay around 0.7..but normally with the same damper you control not only heave but also pithc and roll...and to calculate the damping ratios in thid phase you need to know the inertia momentum of the car. I wasn't able to calculate it.

The problem is to find a reference value od C/cr but i am not sure this is 0.7...

[phoenix]

Originally Posted by silente

Hi Matt!

First of all why you want exactly C/cr=0.7?

I had the same problem some time ago working on a formula 3 car...i found values of c/cr which sometimes arrived to 2.5-3!

I read probably the same papers you did on internet and they say to stay around 0.7..but normally with the same damper you control not only heave but also pithc and roll...and to calculate the damping ratios in thid phase you need to know the inertia momentum of the car. I wasn't able to calculate it.

The problem is to find a reference value od C/cr but i am not sure this is 0.7...

With a C/CR = 1 the damper will never overshoot the spring length at ride height and will take a long time to return to it's free length. The next bump may be encountered before the car has returned to it's correct ride height. The next and successive bumps will cause the damper/spring combination to get shorter and shorter. This effect is known as 'jacking down'. It can mean you soon end up on the bump stops with no room for further bump.

This is why a C/CR of less than 1 really has to be used. 0.7 tends to let the car settle after 0.75 cycles following a bump input. Any C/CR < 1 is under damped and as the ratio gets lower the number of cycles and therefore the time to return to a stable state increases. In the UK a car will fail the MOT, I believe, if the car has not settled by 1.25 cycles following a bump. This occurs a C/CR = 0.5

'cos I can't do drawings, look here:

http://hyperphysics.phy-astr.gsu.edu/hbase/oscda2.html

[silente]

i know what you mean phoenix, and i agree...

but i have seen a lot of succesful teams using C/cr bigger than one with succes...i am not able to understand exactly why these cars work so fine with this kind of damping forces but they do!

As i said before, probably the key is in the possibility to have a better control also on pitch and roll.

What is the wy you would use to design a damper curve from a blank sheet knowing more or less which range of stiffness to use for springs and bars?

[flavorPacket]

don't take this the wrong way, but you should really get someone else to spec these for you unless you really want to learn a lot.

the problem with damping on a normal car is that you have one damper to control the behavior of two masses. (Actually you have two dampers, but nobody knows much about tire damping that they're willing to share). Luckily, the behaviors of the two masses often occur at different frequencies and wheel speeds, so you can try to tailor your damping forces at different speeds to do different things.

Yes, any basic vibrations book can tell you that c=.7 will get you decent response without overshoot, but you need to translate that into an overall effect on the vehicle. Maybe you need more force to get faster weight transfer, or properly damp one DOF of the sprung mass more than another (they ALL have different inertias and stiffnesses, therefore they all have different natural frequencies). Maybe you need less to improve normal force variation on a tire.

It's not that clear cut, and when people do it for real, they use a lot of information and tools that it sounds like you don't have.

[Matt22]

Thanks all

I know this is A LOT more complicated than it first seems. However, I know the recommended damper rates, springs, etc. of my older car, so I'm trying to understand these in order to choose dampers for my newer car. Some of the suppliers will help you out with the spec. but the affordable (high quality) supplier needs a defined damper curve.

Matt

[silente]

i have to say, this topic is really interesting.

I don't exactly understand one of the things we said...Starting from zero to design a good damping curve for given suspension (fixed geometry, springs rates, motion ratios, roll bars...), which value and which path you would use for C? would you consider only the heave movement to fix it or you would need inertia momentum data to fix also pitch and roll damping?

Which way you would use looking data from data logging to understand if your damping is too high or too low? I read a lot on Claude Rouelle site about balancing rebound and bump, but which way to understand if C is correct?

[Matt22]

silente, as far as i can work out we need to compromise on all of the motions. The highest number for c comes from roll (for my car anyway - non aero) so this is what I have used. Won't know if I'm right until next season......

[flavorPacket]

For most cars, the highest inertia is in pitch, so if you try to properly control that motion with a conventional one damper per corner setup, then you'll end up overdamped in roll, and really really overdamped in warp. If you run a 3rd damper, then you can get around this compromise: you can get stiffer damping in pitch without affecting roll and warp (at least theoretically).

You can see that with a normal suspension, you'll have to make compromises between properly controlling each mode. The best way figure out what to do is through testing and advanced simulation. At the end of the day, the laptime tells all.

But, keep in mind that this is only regarding controlling the behavior of the sprung mass. You also have to worry about the tires, specifically the vertical load on the contact patch, and that's an entirely new discussion...

[phoenix]

Originally Posted by flavorPacket

But, keep in mind that this is only regarding controlling the behavior of the sprung mass. You also have to worry about the tires, specifically the vertical load on the contact patch, and that's an entirely new discussion...

If you know how a damper can control the spring in a tyre - write it here. It will be a first!

[sloracer]

Firstly, the units should turn out to be Nm/s.

Secondly the damping "rate" requirements are different at high speed vs low speed and again for bump vs rebound.

Also, my thoughts about high damping ratios on aero cars are that you have to add the aero load into the "mass". I think that once you recalculate the damping ratio with downforce you'll find a much smaller (reasonable) damping ratio.

[Matt22]

N per m/s, so Ns/m

Yes agree with your points, sloracer. Since I posted my first question I've got hold of some more data from the car manufacturer. Reverse-engineering this data I've come up with my curves (based on roll damping as I said).

Interestingly, the data I received suggests that the car is question is very very overdamped in pitch (non-aero, without a third damper).

[flavorPacket]

Originally Posted by Matt22

Interestingly, the data I received suggests that the car is question is very very overdamped in pitch (non-aero, without a third damper).

If it's a road car, that's not really surprising. Production car suspension designers also have to deal with steady-state vibration/excitation issues that race car guys don't really care about. In a typical ride situation, this will normally lead to an overdamped transient pitch mode.

and phoenix, that's not really that tough. It's a simple 4 dof non linear base input vibration problem with parametric excitation...Unfortunately, ideal normal force variation also depends on wheel angular velocity, among many other things, and the damping in the carcass itself, which isn't exactly easy to measure.

[silente]

Matt22, so you were able to measure the inertia momentum in pitch??

how did you do it?

[Matt22]

Model, not measure

Like lots of people before, I've modelled the car using the known masses, motion ratios, C of G heights, etc. From this I can work out the required spring rates, ARB rates, ride height, etc.

Although it doesn't do dynamic simulations, at least I know how far the dampers move under each g-loading and from the data-logger I know how long those loadings take.

[silente]

yes, you mean the calculation on the suspensions...

but what i mean is the calculation to know how big is your inertia momentum..this is necessary to calculate damping ratios in pitch and roll...

[sloracer]

Ns/m... sorry, I don't know what I was thinking.

Do you guys know of any good literature on damping the body motions (ie pitch and roll). All of my knowledge/experience is based around the mechanical grip areas.

[flavorPacket]

I'd recommend the shock absorber handbook by Dixon. it has a decent 4 dof pitch/heave analysis, which you could then apply to other modes.

Also, what knowledge do you have regarding tuning dampers for mechanical grip? That's the most difficult part to analyze. Do you just mean experience on track?

[sloracer]

I worked for Claude Rouelle for a bit and learned a few things... it's really hard to say "OK this is a new car, and these are the damper numbers I need." Most of what I do is based around getting information from a data system and making changes from there. A dyno chart def helps but it still takes time to tune on the track.

[flavorPacket]

Originally Posted by sloracer

I worked for Claude Rouelle for a bit and learned a few things... it's really hard to say "OK this is a new car, and these are the damper numbers I need." Most of what I do is based around getting information from a data system and making changes from there. A dyno chart def helps but it still takes time to tune on the track.

well, sloracer, that's certainly true if your models aren't good enough. A talented professional team can get pretty darn close with good simulations and a couple hours on a 7 post.