12 replies to this topic

[reflex_monkey]

forgive me for posting this fairly arbitrary topic, but it driving me nuts now i can t figure this out coz my brain's in knots

how would one go about working out the damping velocities for each wheel assuming that you have all necessary variables i.e. damping coefficients, input force onto the tyre, suspension travel, pushrod lengths, rocker configuration and assemblies, etc, etc.

please go slowly since i aint the sharpest tool in the shed and im a few steps short of a ladder

[phantom II]

I don't think that you should apologize for any inquiry into shocks. I await input from the usual suspects on this BB because inquiring minds want to know the answers. I only have questions also.
What is the application? Is this a downforce car? An offroad car, a street car. What suspension geometry?. Will you have rising rate in the geometry or the spring? Will you make your own shocks or purchase them? What is your budget?
This whole Renault MD fiasco is to do with limitations of the shock. It is just impossible to dampen every force input because you just can't create a system in a little tube to oppose everything that happens at the wheel. The softer the spring rate and shock valving you can go with the better. Even with the most highly developed active suspensions in Corvettes, Mercedes and Cadillac, they have their shortcomings and can't handle certain bumps.

The Baja off road cars have massive suspension travel and the long thin multi shock packs, some times 5 each side at the back, are all tuned differently. They are probably the most highly effective mechanical shock set up there is, but at what price? In all cases, unsprung weight must be as low as possible.
It is advisable to dampen the wheel rate instead of the spring rate, IE: placing the shock as close to the wheel as possible but that wont do for a F1 car. Where you mount the shock and how long they are and what angle is important .
The higher the velocities of the fluid in the shock, the more accurate the damping but the damping range is limited and temperature control is difficult. To get valves to work if the shock is too short and wide is difficult. You must know the shaft speed and travel in the shock which is calculated from the wheel velocity and begin from there.
I am at the mercy of the manufacturer when they decides what shock the spring needs that I have calculated the load and rates for. I never get them right.
I wonder if F1 would move away from their 13" wheels if they were allowed to. Low aspect ratio tires give a lot of lee way for pressure adjustment and they don't need that much damping and also reduce unsprung weight and can take massive impacts. I laugh at street cars with huge wheels and high aspect ratio tires. Only stylists think they look cool but you don't see them on race cars. Choose you tire carefully.
Corvettes have a leaf spring with constant rates and the loads are taken up at the center of the chassis. They have huge ARBs and dive and suat geometory.
The spring is not taken into consideration when they designed the shocks. The wheel rate is where it's at. C4 Vettes had an ingenious Bilstein set up. Each shock mounted at the corners had a little electric motor ($500 ea.) placed on top of the shock that rotated a shaft that had a spiral groove machined into it. Depending on the position of the shaft, an appropriate oil passage or orifice would control velocity thru a single valve. The motor would move the shaft thru 200' at 100 times per second in real time clockwise and anti-clockwise. It worked really really well within reason. A big wheel and tire and brake, etc will do what it wants to do and no little shock will tell it different.

Originally posted by reflex_monkey
forgive me for posting this fairly arbitrary topic, but it driving me nuts now i can t figure this out coz my brain's in knots

how would one go about working out the damping velocities for each wheel assuming that you have all necessary variables i.e. damping coefficients, input force onto the tyre, suspension travel, pushrod lengths, rocker configuration and assemblies, etc, etc.

please go slowly since i aint the sharpest tool in the shed and im a few steps short of a ladder

[gbaker]

I'm not sure what you mean by "damping velocities". If you are asking about how the wheel motion is effected by the damping coefficient and the other variables mentioned, you'd have to run the math.

Starting with simple linear motion you can determine the damper reaction/wheel speed from the wheel acceleration--given that you know the input force on, and mass of, the wheel. If that's not your question and you want to design a suspension from scratch, then the answer gets a bit more complicated.

[sblick]

I bet they stick an accelerometer on a suspension arm and find the frequency of interest of the tire and then adjust the mass damper accordingly. I sit here in a small company that does noise and vibration and when we tune mass dampers we drive the car and figure the frequency of interest from our customers complaint. We then adjust weights and durometer of rubber until we get the right frequency. We have a small MatLab program that will tell us the approximate size of the weight that will dampen out that particular acceleration on that particular part. I am oversimplifying to some extent but the whole story is there.

[Greg Locock]

If you know the profile of the road, and you have a very good tire model, and an unbelievably good shock model, then the shock velocity for a given setup can be predicted from a full vehicle model. The problem you've got is that the shock velocity will be strongly affected by the suspension calibration, including the shock valving. So its a bit of a chicken and egg situation.

However to be honest I doubt anybody does it that way, much better is to measure the real shock velocity, with a stringpot, as you do a lap. Then you know what velocity range you are interested in.

The other problem with modelling shock velocities is that if you look at real data for shock force vs speed from a real event it bears only a vague relationship to the nice little curve you measured on the shock dyno.

Worth pointing out that a road car might see shock velocities as high as 6 m/s, a racing car will see less than 1.5 m/s.

[shaun979]

Originally posted by Greg Locock
Worth pointing out that a road car might see shock velocities as high as 6 m/s, a racing car will see less than 1.5 m/s.

Is it right to say that the low velocity in racing cars is due to the shorter travel, higher spring rate, and less rough of a surface (no major big bumps)?

In the chicken/egg situation you mention with shock velocities, how do you know you're close with baseline settings? I've always wondered this ever since seeing some ~100K USD shock dynos that are able to playback shock travel from datalogs. How do the engineers know whether the track profile they're recording and trying to optimize for, is right?

[Greg Locock]

"Is it right to say that the low velocity in racing cars is due to the shorter travel, higher spring rate, and less rough of a surface (no major big bumps)? "

Yes, but I haven't seen data for a circuit kerbing event, but that's the sort of thing we see all the time on durability. It's amazing how much you can limit the high shock velocities by making sure the shock is right at low velocities but then the ride guys whinge.

"In the chicken/egg situation you mention with shock velocities, how do you know you're close with baseline settings? I've always wondered this ever since seeing some ~100K USD shock dynos that are able to playback shock travel from datalogs. How do the engineers know whether the track profile they're recording and trying to optimize for, is right? "

Dunno. I've seen various proposals for setting approximate shock calibrations analytically, before the car is built. I don't think any of them has really shown a convinciing methodology. Whatever worked last time seems as a good a baseline as any!

[phantom II]

The moto cross guys set up at the event. I think that's about the only information that they download from the bike during practice. There are some skilled technitions that can read the terrain and build a shock accordingly on the spot. Are there any bike guys here that can tell the story of how they valve those shocks? Over damping is like having the springs too hard. How do you know what the correct restriction and velocity for the various events are? What are the typical designs for shocks? Are all F1 cars the same?

Originally posted by Greg Locock
"Is it right to say that the low velocity in racing cars is due to the shorter travel, higher spring rate, and less rough of a surface (no major big bumps)? "

Yes, but I haven't seen data for a circuit kerbing event, but that's the sort of thing we see all the time on durability. It's amazing how much you can limit the high shock velocities by making sure the shock is right at low velocities but then the ride guys whinge.

"In the chicken/egg situation you mention with shock velocities, how do you know you're close with baseline settings? I've always wondered this ever since seeing some ~100K USD shock dynos that are able to playback shock travel from datalogs. How do the engineers know whether the track profile they're recording and trying to optimize for, is right? "

Dunno. I've seen various proposals for setting approximate shock calibrations analytically, before the car is built. I don't think any of them has really shown a convinciing methodology. Whatever worked last time seems as a good a baseline as any!

[RDV]

reflex_monkey- how would one go about working out the damping velocities for each wheel assuming that you have all necessary variables i.e. damping coefficients, input force onto the tyre, suspension travel, pushrod lengths, rocker configuration and assemblies, etc, etc

Very good question... would like to know myself, as a cornerstone of the work I do... at moment is totally empirical. A good case of lemon engineering.. as in suck it and see, and the reason for seven post rigs.

Gregg-However to be honest I doubt anybody does it that way, much better is to measure the real shock velocity, with a stringpot, as you do a lap. Then you know what velocity range you are interested in.

Indeed, as Gregg mentions further on , rig measurements seem to correlate very little with what one sees at the coalface...

Originally posted by Greg Locock
Worth pointing out that a road car might see shock velocities as high as 6 m/s, a racing car will see less than 1.5 m/s.

Apart from kerbs, the velocities we are interested in , and which change car behaviour are more towards the 5 to 50mm/sec. On offroad cars have seen as high as 14m/s... there are no rigs that can measure that...at least to my knowledge... and I have used Kings, Donnerre's, Konis, Sachs, Ohlins, Bilsteins ,Armstrongs, Whitepower, Dynamic,

phantomII-I wonder if F1 would move away from their 13" wheels if they were allowed to. Low aspect ratio tires give a lot of lee way for pressure adjustment and they don't need that much damping and also reduce unsprung weight and can take massive impacts.

Yes please!!

phantomII-The Baja off road cars have massive suspension travel and the long thin multi shock packs, some times 5 each side at the back, are all tuned differently. They are probably the most highly effective mechanical shock set up there is, but at what price?

The price is not so good handling on paved roads.. that said they take like a duck to water white noise type of terrain... at the moment will unequivocaly state that WRC cars are at the pinnacle of damping understanding... they run low-profile tyres and rough ground, but also have to contend whith yumps, gravel plus dirt... track cars by comparison are quite crude as everything is tributary to the tyre carcass behaviour and even more so to aero control.

As for overdamping that is the plague that afflicts most F1 setups... it makes everything that little bit more linear and thus less complex, though not necessarily more eficient.

[Fat Boy]

Originally posted by Greg Locock
"Is it right to say that the low velocity in racing cars is due to the shorter travel, higher spring rate, and less rough of a surface (no major big bumps)? "

Yes, but I haven't seen data for a circuit kerbing event, but that's the sort of thing we see all the time on durability. It's amazing how much you can limit the high shock velocities by making sure the shock is right at low velocities but then the ride guys whinge.

"In the chicken/egg situation you mention with shock velocities, how do you know you're close with baseline settings? I've always wondered this ever since seeing some ~100K USD shock dynos that are able to playback shock travel from datalogs. How do the engineers know whether the track profile they're recording and trying to optimize for, is right? "

Dunno. I've seen various proposals for setting approximate shock calibrations analytically, before the car is built. I don't think any of them has really shown a convinciing methodology. Whatever worked last time seems as a good a baseline as any!

I've seen engineers have real fits over high shock velocities. The comments I've heard is 'It's moving too much and we need to settle the car by slowing the shock velocities." They then crank in the knobs or throw a few errant fender washers in the valve stack and PRESTO! shock velocities have been reduced.

The important question is, though, has it made the car faster? On street courses or places with curbs that you have to jump it generally doesn't.

The reason is simple. Shock velocities are important to engineers because we can accurately measure how much force the dampers make at certain velocities. We can obsess about our curves and bleeds and oils and shims to determine what forces we actually want to make. The suspension system, however, doesn't really care. In terms of contact patch load, the velocity of the damper (on it's own) is essentially a non-player. The important variable is the frequency. When the input frequencies are low, you want more damping to control the system. When the input frequencies are high, you want less damping to control the system. Driver inputs are low (or at least _should_ be). Because of that, they tend to like a higher damped system. Track seams and curbs and pavement changes are high frequency, they tend to like no damping.

What you have to do is determine what input frequency is the player at important places around the track. Then correlate that input frequency (of some small distrubing event) to a shaft velocity (something we can actually tune on). What Greg said about logging the data and then making changes from there is exactly correct (although I use linear or rotary pots, not string pots). If there's a way of doing it with strictly mathmatics, then you're one up on me.

Some one said on the mass damper thread that they generally shoot for less damping as opposed to more. I tend to agree. There's probably a curve where you can go from X damping ratio to Y damping ratio and not really effect lap time that much. My tendency is to be on the lower side of the damping ratio, but that's just me.

[Greg Locock]

"Driver inputs are low (or at least _should_ be). "

BMW 5 series have, or at least had, an isolator in the steering column, that progressively attenuated all driver inputs above 3 Hz

I can't remember what the roll-off was, but whatever it was, by 10 Hz the steering wheel is essentially decoupled from the road wheels.

The other thing I like to think about is - how fast can the car respond? That is a function of the polar moment of inertia, and the yaw stiffness of the tires (and a bit of suspension maybe).

If you want to play about with this the best manouevre is a pulse input, ie 0-1-0. The upper frequency of the input is the inverse of the time taken for the pulse input.

[Fat Boy]

Originally posted by Greg Locock
"Driver inputs are low (or at least _should_ be). "

BMW 5 series have, or at least had, an isolator in the steering column, that progressively attenuated all driver inputs above 3 Hz

I can't remember what the roll-off was, but whatever it was, by 10 Hz the steering wheel is essentially decoupled from the road wheels.

A few years ago I was working with a young driver who was constantly yanking on the wheel, even in the middle of corners. It was really weird what he was doing, but he would saw at the wheel for no apparent reason (He is the heir to a certain rather important US racetrack, and has the basic driving talent of his better known uncle). Anyway, I started dialing in caster to the car specifically to make the steering heavier. I figured if I made the steering heavy enough, he might actually keep his hands still.

While I can't say that it fixed his habit, after adding about 2 1/2 degrees it certainly limited it. It also slowed his steering inputs to the point where the car actually had a chance to respond. It didn't work miracles, but it did help.

I wouldn't want to decouple steering on a racecar, but I can understand what BMW were going for.

[RDV]

Fat Boy-'It's moving too much and we need to settle the car by slowing the shock velocities."

Mmmm... they havent separated damper velocity from damper displacement.... there are a lot of people out there that thing the low speed and high speed adjusters on shocks are for low speed corners and fast... (most drivers!)

Ultimately an overdamped suspension wont move at all... the car still will move, but it will be all in the tyres.

Very high damper speeds means big displavements, as you say on a kerb... ripple bumps cant achieve high velocities because very low displacement... speed also comes into it... a kerb run over at 1 kph can be smooth, at 200kph will throw car into air... or rip the wheel off.

The nub of the question is that it all runs through the tyre, and there is always the pesky aero effects coming into it... run a rigid tyre and things would be a lot simpler... anyone have damping curves for railway stock? What does it look like?

There was a vogue for dampers with shuttle control down the central shaft, and these were very good over aforementioned ripple bumps at low speed corners, main reason being that the cost factor was quite low, due to the shuttle hysteresis alowing fluid to flow freely in the reversals.

A conventional shim-stack and orifice-only damper was better at longer strokes and/or higher velocities, but consistently slower to the tune 0f 0.1~0.2 secs a lap... and you could not see the difference on a damper dyno, just track logged data.

McPherson suspended touring cars I have run benefited immensely from a very crude device= simple blow-offs in the stack, for those kerb crossing moments, and in off road racing we go a stage further , by using inertia valves for droop travel, if the car is traveling in a straigh line ( z axis..) and wheel drops into pot hole, the damper has no flow restriction, and suspension just extends. If chassis is moving upwards, the inertia valve allows the flow to be restricted.

If there's a way of doing it with strictly mathematics, then you're one up on me.

Ditto....

There's probably a curve where you can go from X damping ratio to Y damping ratio and not really effect lap time that much.

... just driver feel.