How does a shock absorber work or should work?

[lightarrow]

An hydraulic shock absorber of a vehicle operates, usually, by viscous friction: the braking force is proportional to the relative speed of the mechanical parts (piston sliding in cylinder) through passage of the fluid (oil); in some more sophisticated cases there are valves that reduces the flow of fluid when the relative speed varies sudden (shocks).

Having complete freedom of design, and need not be limited to hydraulic functioning, what would be the “best” relationship between the relative speed of the mechanical parts and braking force?

I’m aware that here “best” is not defined, but at least I mean with the better stability of the veichle and/or the better continuous contact of the wheels with the road.

Of course, much depends on the type of stress to which the shock absorber would be subjected, but in general, would you design the working so that the braking force is simply a proportional function of the speed or with a different law (and maybe not depending on the speed only)?

I understand it’s not a simple question but at least I would appreciate a simplistic answer.

Thank you.

--
lightarrow

 

[Naty1]

all a shock absorber does is to work with springs to make a ride smoother. With springs, it dampens oscillations...
you can eliminate both, but you get a bumpy ride.

HAve you read the wikipedia article..shock absorber, and struts...

 

[lightarrow]

all a shock absorber does is to work with springs to make a ride smoother. With springs, it dampens oscillations...
you can eliminate both, but you get a bumpy ride.

Certainly But maybe I used the incorrect term. How would you design the damper element, with the braking force proportional to the speed, or to a certain power (greater than 1), exponentially with the speed,...?

HAve you read the wikipedia article..shock absorber, and struts...

Yes.

 

[AlephZero]

I would take out the spring and damper completely, and replace them with an active system to create a force between the wheels and the vehicle body.

If you keep the force on the body constant, you will get a perfectly smooth ride. You don't want a constant force on the wheel, because it has to accelerate vertically to go over bumps in the road. Ideally you would want to to predict the bumps and changes of gradient etc of the road, by scanning the road ahead of where the wheels are. Even better, you would also want to predict what the driver was going to do next, but that might be too hard a problem to solve

 

[CWatters]

I suspect it's a lot more complicated than people think. Without a shock you have a mass on a spring which equals an oscillator. The shock controls how that oscillator responds to an impulse. It's been years since I understood how damping works exactly but there is plenty of info on the web..

http://en.wikipedia.org/wiki/Damping

I imagine the shock might be tuned to provide something close to critical damping but you would have to ask a car designer. If I bounce on the fender of my car it appears it might be slightly under damped? eg I can make the car oscillate just a little rather than returning straight to the rest position.

 

[CWatters]

Some research has been done into recovering the energy dissipated in a shock absorber to improve fuel economy...

http://www.designnews.com/document.asp?doc_id=228409&dfpPParams=industry_auto,aid_228409&dfpLayout=article

Levant is also targeting Class-8 trucks, where it estimates the GenShock could boost fuel economy by 3 to 5 percent

http://web.mit.edu/newsoffice/2009/shock-absorbers-0209.html

Senior Shakeel Avadhany and his teammates say they can produce up to a 10 percent improvement in overall vehicle fuel efficiency by using the regenerative shock absorbers.

 

[Naty1]

would you design the working so that the braking force is simply a proportional function of the speed

take a drive off the speedometer, use it to control an active strut mechanism...say pump pressure related to speed of rotation...something is likely available already...

 

[lightarrow]

I would take out the spring and damper completely, and replace them with an active system to create a force between the wheels and the vehicle body.
If you keep the force on the body constant, you will get a perfectly smooth ride.

I imagine you have an interesting budget to achieve such a result

You don't want a constant force on the wheel, because it has to accelerate vertically to go over bumps in the road. Ideally you would want to to predict the bumps and changes of gradient etc of the road, by scanning the road ahead of where the wheels are. Even better, you would also want to predict what the driver was going to do next, but that might be too hard a problem to solve

Ok, let's say you can assume the road is smooth excepting for "bumps" bell-shaped (you can use a gaussian curve or another of your choice), 5 cm high, 1 m long and car's speed is 80 km/h. What would you want as F(v), that is the relationship between damping force F and relative speed v between the wheel and the car?

 

[lightarrow]

I suspect it's a lot more complicated than people think. Without a shock you have a mass on a spring which equals an oscillator. The shock controls how that oscillator responds to an impulse. It's been years since I understood how damping works exactly but there is plenty of info on the web..

http://en.wikipedia.org/wiki/Damping

I imagine the shock might be tuned to provide something close to critical damping but you would have to ask a car designer. If I bounce on the fender of my car it appears it might be slightly under damped? eg I can make the car oscillate just a little rather than returning straight to the rest position.

That page on wiki describe dampening when F(v) is simply proportional to v, that is "viscous" dampening, clearly because it's the simplest case. But I was asking myself if there are better ways than the viscous one, for a real case.

 

[lightarrow]

take a drive off the speedometer, use it to control an active strut mechanism...say pump pressure related to speed of rotation...something is likely available already...

Haven't understood exactly what you propose here, but if you mean to measure the speed v of the wheel relative to the car, you assume that my car's shock absorbers work perfectly, which is not
And even if they worked perfectly, how do we know that they couldn't work better? After all, hydraulic dampers are limited by the viscous properties of a fluid, and I don't think is possible to make chemical compounds or mixes of them which can vary their viscosity with the speed v (or I'm wrong?)

 

[Pkruse]

There are many types of shock absorbers in many types of mechanical designs. All have one thing in common, which is that they absorb energy. The viscous damping you site absorbs energy by heating oil. Other EA systems absorb by friction, or by deforming material, or by tearing it. A common EA used in fall protection systems simply tears stitches to minimize arresting loads for a person who falls. As already noted, the purpose of the energy absorber you talk about here is to damp out the natural harmonics of the suspension system.

 

[lightarrow]

There are many types of shock absorbers in many types of mechanical designs. All have one thing in common, which is that they absorb energy. The viscous damping you site absorbs energy by heating oil. Other EA systems absorb by friction, or by deforming material, or by tearing it. A common EA used in fall protection systems simply tears stitches to minimize arresting loads for a person who falls. As already noted, the purpose of the energy absorber you talk about here is to damp out the natural harmonics of the suspension system.

In the 30', the damping element was a system of some friction disks (made of wood or asbestos):
http://en.wikipedia.org/wiki/Friction_disk_shock_absorber
Later, they were replaced by hydraulic dampers, because these have low braking power at low speeds.
Maybe the next level of damping is a braking force which varies even more with the speed v, that is not simply proportional to v but to vⁿ or to e^v ?

 

[Pkruse]

The really nice thing about viscous damping is that it is linear. Friction damping is called Coulomb damping. An example is found commonly in a washing machine to damp out the drum during the spin cycle. Another is the anti sway mechanism in a trailer hitch to prevent a trailer from fish tailing in traffic. It is not linear, so calculation are a bit of a pain. I suspect that is why we don't see much of it in real designs, except in trivial designs where calculations are unimportant, or in turbine blades where no expense is spared to optimize the design.

 

[lightarrow]

Thanks for the answer.

--
lightarrow

 

[Enthalpy]

Shock absorbers are NOT dampers. They are designed so that the force is proportional to the square of the speed, and at constant speed, the force increases as the braking way is used.

In this way, shock absorber provide a braking force constant over the braking way, and they adapt the braking speed optimally to the initial speed. But they don't adapt to the mass! A smaller mass will just brake stronger at the beginning, but a mass too big won't have stopped before the end of the way.

The force varying as speed squared results just from oil flow through holes. This is much easier than in a damper, where flow must be kept laminar.

The force constant over the braking way, as the speed decreases, is obtained by the piston that covers more and more holes, so the available section decreases as the remaining speed, that is as the square root of the remaining way.

 

[Pkruse]

I find it interesting to get engineers and physicists involved in the same discussion. One is concerned with the pure science, while the other with practical application. As an engineer I use the science to the extent that it is useful, as often it is. But other times I do stuff simply because it works, regardless of the opinions of scientists.

I've been an engineer a long time, and can easily say that if you see a viscous shock absorber in an engineered design, it is probably there to damp out the harmonics. We sometimes use them for other reasons, but not often.

 

[lightarrow]

Shock absorbers are NOT dampers. They are designed so that the force is proportional to the square of the speed, and at constant speed, the force increases as the braking way is used.

In this way, shock absorber provide a braking force constant over the braking way, and they adapt the braking speed optimally to the initial speed. But they don't adapt to the mass! A smaller mass will just brake stronger at the beginning, but a mass too big won't have stopped before the end of the way.

The force varying as speed squared results just from oil flow through holes. This is much easier than in a damper, where flow must be kept laminar.

The force constant over the braking way, as the speed decreases, is obtained by the piston that covers more and more holes, so the available section decreases as the remaining speed, that is as the square root of the remaining way.

Very interesting, I didn't now that. Where I could find more informations about it?

 

[sophiecentaur]

I have read through this thread but I can't find (or have missed!) any reference to a vital part of most vehicle shock absorbers. The damping system is non linear and works differently in the two directions. The fluid is forced through a 'large number' of holes when the can hits a bump, this allows for a smooth ride and avoids an upwards 'shock' for the passengers. However, when the wheel returns to the ground, a flap valve blocks fills many of the holes, causing more energy loss but increasing the time constant. This system has a better damping effect for a given subjective smoothness of the ride. It is easy to test this by leaning hard on a (total stranger's haha) car and then releasing it. It takes longer to come up than it took to go down.
This is only true for passive systems, of course and is easier to observe in light cars with soft suspension (Citroen 2CV is an extreme example).