Optimal damping 1

[RSam]

I read few books about vehicle dynamics and according to all I have seen (specially RCVD by Milliken) the damping have to be 2 times higher in rebound than in bump. I read that road have 2 times more bump than holes and also that the stiffer rebound allow a higher energy dissipation but some research paper usually describes road as a gaussian signal with 0 mean value. The other thing is that high speed bump and rebound must be softer than low speed. I guess the reason is related to resonance frequency of sprung vs nonsprung mass but I am not 100% sure. Can someone gives me a clean and scientific explanation about these 2 dogmes:

1- damper must have a rebound damping 2 times higher than the bump damping
2- High speed damping must be softer than low speed damping (both in bump and rebound)

Thanks in advance

Sam

 

[gt6racer2]

1) Low/mid speed damping is used to control sprung mass motion. ( 1-2 Hz). The jounce/rebound ratio is just preference for ride Vs handling. To control the sprung mass we have to take out a certain amount of energy ( damping ) the system doesn't really care if this is done in jounce or rebound. ( Let's neglect damper jacking for now ). In the US we used to have ratios as high as 6:1 for soft impacts but the handling was bad/slow responding as the lack of jounce damping allows the body to roll a long way before building cornering forces. Race cars commonly use 50/50 dampers, with especially tight low speed jounce as the comfort is not such an issue. With the trend toward "european" ride a ratio of 2:2 has become the more comon balance. The unsprung mass control is largely lower speed damping.
2) The second point is strange, as all dampers make more force at higher speeds. ( OK so there are exceptions- but they're rare). Typically high speed damping is generated by an orifice limit - which increases damping rapidly with velocity. The high speed damping is used mainly to control unsprung mass shake on live axles ( 10- 12 Hz ), or to prevent crash through.

 

[SusTestEng]

1) The main reason you see twice as much rebound damping is due to the fact that during rebound, you are trying to control the spring force AND the body mass, while in compresson, you are sort of only controling the body mass as the spring force is building and helping to slow down the motion. Higher rebound also helps flattness an sort of "pulls down" on the car.

2) At very low piston speeds, the damping is usually almost equal to slightly more rebound. At theese low speeds(just barely turning the steering wheel on the freeway, or very slight dips in the pavement) the damping is controled by slit valves(sometimes called notch leafs or bleed valves) where the oil is simply "leaking" throung the piston or bottom piece.
At High piston speeds (sharp impacts, or extreme transition limit handling) the damping force is controled by the size of the hole in the piston and how much oil can pass through it. I think what the most books are trying to say is that it is prefered if the slope of the force/piston speed graph is less than the mid speed slope. This is because the oil will reach a point where it can't flow any faster and the damping force will start increasing VERY rapidly(2-3 m/s piston speed). So basically they want to see as big of a hole as posible in the piston, so this sharp rise in damping force doesn't occur. But the trade off is that sometimes if the hole is too big, the damping for limit handling is not enough.

 

[GregLocock]

1) The main reason you see twice as much rebound damping is due to the fact that during rebound, you are trying to control the spring force AND the body mass, while in compresson, you are sort of only controling the body mass as the spring force is building and helping to slow down the motion. Higher rebound also helps flattness an sort of "pulls down" on the car.

Not only 'sort of'. If you measure the ride height over rough surfaces you'll actually see it suck down onto the springs as the rebound damping dominates.

Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[Fabrico]

Nearly every shock on the planet has what is called speed sensitive valving. Rebound has a definite and fairly slow maximum speed. On the other hand, compression speed can be almost that of vehicle speed. They are different functions working at different speeds. Comparing them on anything sportier than a big Cadillac seems like a moot point.

As far as rebound damping energy, shouldn’t body mass be subtracted from spring force instead of being added to it? How can rebound address anything more than spring force? The tires aren't glued to the ground.

 

[GregLocock]

On our ride and handling circuit shock velocities max out at +1500 mm/s and -1300 mm/s, and on our durability events the same ratio applies.

1) the maximum velocity is nothing like vehicle speed
2) upward and downward velocities are similar.



Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[SusTestEng]

"On our ride and handling circuit shock velocities max out at +1500 mm/s and -1300 mm/s, and on our durability events the same ratio applies."

You must have some pretty smooth roads We have a section or 2 that we can get the piston speeds up to 2.5 m/s. It is a section of very rough/broken asphalt with undulations that we drive at about 35-40 mph. It is used for flatteness evaluation and body control on rough surface. Mainly to see the limits of the high speed oriface of the dampers.

 

[GregLocock]

That's our limit handling track. Ride evaluations are done on a selection of local roads, mostly.

Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[SusTestEng]

It seems that some of you in this thread are not directly related to shock tuning, so I will offer some insight that might completly confuse you, or it may shed some better light on the subject. Shock tuning is a "black majic", you CAN NOT look at a force/velocity curve and make a judgement of the damper. The ONLY thing a force/velocity curve is good for is to compare the left and right damper to make sure they are built corectly, and to sort of gage the relative "stiffness" of a damper. The really important part of a damper is the reactions to accelerations and decelerations of the rod. A force/velocity curve only shows you the reaction force at an instantaneous velocity, and doesn't show you the whole picture. I can have 40 different dampers built with the SAME EXACT force/velocity curve that all feel VERY different. That being said, it sometimes is a little hard to explain to upper management "what you did all week" when the curve of the new dampers is almost right on top of the old ones. The biggest factor is how the three main element work together and transition to each other(on a twin tube design you also have a bottom valve that works with the piston in compression, but that gets complicated to explain). You have a bleed area, a main leaf valve area, and an oriface area that create damping force. All of these have to work seperatly AND transition to one another smoothly to produce a "quality damping feel". Usually, when I tune shock absorbers, I don't request "damping force changes", but I do request part changes and not worry about the resulting damping force. I usually talk in terms of slit/notch leaf valves, or the leaf stack combination(thickness, diameter, and order). There are alot of parts inside a damper, but once you understand the basics of what each specific part effects and you can visualize what is going on inside the damper and relate that to what you are feeling when you drive the car, shock tuning gets pretty easy for you. I don't mean to confuse any of you, but the actual damping force number is not an exact representation of what is going on.
A force/displacment curve(shaped like an American football), gives a little bit better picture of what is going on, but it still leaves out the accelerations. If there was a way to acurately measure and record a graph that shows exactly what you are feeling, that would be a great tool, but the current dyno's are just not very capable of doing that. It takes a large amount of energy to reproduce what is actually happening on the car.

 

[hutch325]

The "black art" of this is pretty interesting to me, as I have recently been specifying some dampers for some race cars. Basically I have done some simple calculations (damping ratios mainly, but a few simulations) to get an idea for how much force would be appropriate, then relied on an experienced damper builder to put the thing together so it works. So far, so good.

So my question is, what's the best way to learn the intricacies of this tuning? Build lots of dampers? It does sound like it's possible improve performance by trying to eliminate friction effects (maybe less shims rubbing against each other) and deflection distances (stiffer shims). Sort of the right idea?

 

[SusTestEng]

Hutch325:
The BEST way is to spend some time with your damper builder! Ask him to show you every part and how the damper goes together. Or, if you have a boken shock laying around, open it up and check it out...as an engineer, I'm sure you have taken things apart to see how the work. It is pretty difficult to explain in writing what all the effects of the internals are, that is why it is best to do a tuning session and build a bunch of dampers and see how the feeling changes. During a typical 1 week tuning session with a supplier, I will spend about 2 days on volume and 3 days on the "Black art"/feeling of the damper. I think you have the right idea about friction, deflection, but one of the most important parts is the smoothness (balance) of the damper, to eliminate the sharp increases in focre build up. This is the part of tuning that the dyno has really no use, and it becomes all subjective. A really good damper builder should "know" what is needed and be able to adjust the valveing based on your subjective comments about the stroke and balance. Like I said before, it is easiest to have him build a couple sets(of the same damping force) and tell you what is different between them, so you can feel the difference in the valve code. The hardest part is finding that "perfect" combination of the leaf valve stack, to get the stroke to feel linear, to have the right amount of body control, to have the right amount of "damping feel", and to make sure it transitions well with the bleed area and the hole in the piston. The other part is to balance the car front and rear, to make sure the dampers work together and not create any strange pitching feeling.

 

[NormPeterson]

Originally posted by SusTestEng

. . . The really important part of a damper is the reactions to accelerations and decelerations of the rod. A force/velocity curve only shows you the reaction force at an instantaneous velocity, and doesn't show you the whole picture.

What I'm getting from that is that a force/velocity plot essentially presents the response to some arbitrary constant rod acceleration rather than to whatever a real-world rod a(t) might be.


Norm

 

[SusTestEng]

Norm,

When a damping force machine takes data, it usually runs 3 or 4 strokes at each specific rod/piston velocity(0.02, 0.05, 0.1, 0.3, 0.6, and 1.0 m/s). It takes it's data point at the specific velocity at the peak of the force. The problem with that is that it doesn't show how it gets to that force. Is it a linear progression, is it a really sharp spike, or is it always just really high force??? That is the part that is all feeling, because the dyno's don't show this very clearly according to feeling. This is why you spend 2 days getting the volume (area under the force/velocity curve) correct and 3 days fine tuning the way the valving transitions between one another(what is not shown by the dyno)... to get that feeling correct.

 

[SusTestEng]

Hutch:
Here is a thread that gets into the "Black Art" portion of damper tuning.

http://www.eng-tips.com/viewthread.cfm?qid=107747

 

[hutch325]

Thanks, that thread was pretty interesting and the effects of using thin vs. thick seemed to make sense.

With regards to dyno testing, it seems like you could use a shock dyno that is essentially a linear servo to generate more useful data (i.e. feed it a waveform then measure the resulting force), then run that data through some sort of algorithm to process it into a more useful form, maybe a surface plot of force vs. both velocity and acceleration. I realize this itself would not be easy, then you would have to actually do something useful with the output.

I guess if you're not in F1, stick to simple calcs and testing.

 

[michael2]

"The other part is to balance the car front and rear, to make sure the dampers work together and not create any strange pitching feeling."

See, to me this adds another layer to the black art, and I completely get it. But should some cycling info be built into the damper before the black art portion of develop0ment begins? By cycling I mean how the front and rear work together based upon knowns such as front to rear weight bias.

Thanks, Michael