Magnetic suspension, as referred to in that video, is currently available on (at least) the Audi TT as an option and on Cadillac STS models, although possibly not currently in the UK.
|
Didn't 2CV rear suspension have some link to the front so the rears follow the fronts ?
Or did I just dream that one up ?
|
Lotus had active suspension fitted in prototype form to an Elan+2(I think it was a +2) back in the early 70's.
The Xantia Activa has been around with the next best thing since the mid 80's
|
At some time in the dim and distant past I remember seeing a demonstration video (possibly on Tomorrow's World?) of a brand new Volvo 740 with suspension that would lean into corners, rather like a motorbike. It achieved this by means of the passenger operating a control panel similar to those attached to those American cars with hydraulic "bouncing" suspension.
The push of the item was that the means were there to adjust the suspension and once road-reading technology caught up, a fully predictive system would be feasible.
|
Yes it's very possible. Would be it costly? Yes, at first. Would it be accurate and reliable? As with most inventions, not so great at first, but eventually probably.
1: You could use a sensor based system on the front of the car.
This approach would probably use a camera system in the car, exploiting a system that's there for roadsign recognition (isn't there one available on the Insignia?), but that might get confused at night, or by potholes that have been filled with different colour tarmac.
A radar system that might be on the car for adaptive cruise control might also pick it up with a degree of confidence
Finally you might get somewhere with a lidar (laser equivalent of radar).
But it's difficult to see a system based on single sensors working well. More likely is to use integrate the data from several sensors of different types. It wouldn't give a yes/no answer, but is more likely to produce a result along the lines of "I'm 90% sure that there's a pothole in front of me, and 98% sure that it won't damage me at the speed I'm going".
Of course these sensors are big, expensive and at the moment need a better point of view than drivers eye - check out Stanford Uni's Passat in this picture - the turret on the top of the car is a Velodyne lidar, and there's more lidar under the headlights from Ibeo. The thing is also dripping with radar and cameras - we are starting to see individual bits in real cars (e.g. cameras, radars etc...)
en.wikipedia.org/wiki/File:UrbanChallenge_Standfor...G
2: You could also do it with Satnav, a phone, and a Nintendo. You drive through a big pot-hole, the kind of sensor that's in a Wiimote senses the jolt, and finds out where you are (satnav) and texts it to a central database, which collates reports from millions of people like you, then downloads that information back into the satnavs, and voila, a pot-hole map of the country. It might be a bit inaccurate at first, but as more reports flood in from a single hole, it could take an average of the positions to improve accuracy, and also spot when the reports stopped happening (in the unlikely event of council or HA getting it sorted out).
The flippant point is that all the basic technologies exist to do it, and are cheap, just needs somebody to integrate it all.
Satnav info providers are increasing the accuracy of the information they record for mapping and how accurately your location is known, so your satnav could stiffen your suspension just before a bend, or even apply brake or steering assist if it thought there was a high chance that the alternative was a trip across the fields.
There were a huge load of road/driving safety R&D projects funded by the EU in a programme called Framework 6. They involved many of the big OEMs, Tier1s and specialists/unis, and sets a direction of travel where much more technology will be there in an advisory (not supervisory - nobody wants the liability) role and to get you out of trouble when you make a mistake - 90% of road accidents have human error as a contributing factor.
|
so your satnav could stiffen your suspension just before a bend
Without liability, I have heard that it may already be possible to use the satnav view to plot a course on the motorway through thick fog in the middle of the night...
Edited by Dave_TD {P} on 14/10/2009 at 23:57
|
I can see that predictively stiffening or softening the suspension might be relatively straightforward, but it would still not actually move until the wheel had already hit the bump.
To even out the ride height surely it would have actually to depress the ride height at the instant of hitting, to compensate for the upwards wheel movement, and conversely raise the height just before a pothole. It would need to be powered, not purely reactive?
You can experiment yourself by applying the brakes gently at the instant of hitting a speed hump. If you get it right, the front of the car dips (from the braking) at the same moment as it rises for the bump - result, a level ride.
I've never tried it, but I have heard people who claimed to be able to negotiate bumps at speed in comfort by use of the brakes at the right moment.
|
You can experiment yourself by applying the brakes gently at the instant of hitting a speed hump. If you get it right the front of the car dips (from the braking) at the same moment as it rises for the bump - result a level ride.
I have seen front suspension struts put through a bonnet doing that. (It was a rusty Escort).
Edited by Old Navy on 15/10/2009 at 15:49
|
I have a tortuous set of speed humps at work, and it's interesting to think about the physics of all that. What I do is stop braking about 1/2 second before the bump, so the car gets back level (i.e. the front suspension gets compressed & therefore stiffened under braking, but then unloads back to level when you stop).
Of course the other experiment is to floor it just before the speed hump, making the car sit back, extending the front suspension and maybe that will make it more compliant over the bump? It would also defeat the purpose of speed humps in a hilarious way. One to try in a hire car, perhaps :-)
|
I can't think of a way of designing predictive suspension, suspension that knows what's coming next and prepares for it. What does exist is active suspension, one that works by fast-acting rams and valves without any spring medium. It works well in some applications, but it won't be cheap to buy, might get crotchety when worn and will be expensive and difficult to repair.
The way to flatten out speed bumps is not to brake as you hit them but to accelerate violently after the front wheels have climbed the thing. That will at least prevent the chin of the car from hitting the ground on the other side of the bump. But it won't be kind to the suspension or transmission. You wouldn't do it regularly in your own car if you had any sense.
|
I can't think of a way of designing predictive suspension suspension that knows what's coming next and prepares for it.
Obviously tricky, but why not? If you or I could look at a road and work out at what points we need to lift the wheel (bumps) and where to drop it (potholes), why couldn't a computer with cameras, in theory?
|
a computer with cameras, in theory?
In theory no doubt, but how well would it work in practice? And how much would it cost? and what would happen if insects or road dirt polluted the vision of the forward sensors?
Who needs it, other of course than technophile consumers?
|
>>Who needs it
I imagine it might be very good for ambulances, to enable them to negotiate speed bump ridden badly maintained roads without harming the injured inside.
|
speed bump ridden badly maintained roads without harming the injured inside.
Great. Roads are deliberately made uncomfortable and bumpy for no conceivable good reason. Then the cost of ambulances has to be doubled to protect their passengers.
Edited by Pugugly on 15/10/2009 at 20:25
|
In theory no doubt but how well would it work in practice?
Right - I thought when you said you couldn't see how it could be designed you meant even in theory.
I think it could be designed today, but either the technology (software and/or hardware) isn't there yet, or as you say it would be too expensive. But I'll bet someone somewhere has designed it and is working on it.
Edited by Focus {P} on 15/10/2009 at 16:38
|
I'll try again.
Already exists in a Range Rover.
"The suspension now includes what Land Rover claims is the world's first (in a production car) adaptive damping system that measures the road ahead of the vehicle and adjusts the suspension accordingly."
|
I'll try again.
ahem thanks Dipstick
|
I'll try again. Already exists in a Range Rover. "The suspension now includes what Land Rover claims is the world's first (in a production car) adaptive damping system that measures the road ahead of the vehicle and adjusts the suspension accordingly."
Standby for pedantry! :-) I'm my own worst enemy sometimes ...
www.landrover.com/int/en/vehicles/range-rover-spor...m
The official blurb on the website suggests this is actually a reactive system (i.e. based on current road surface and vehicle behaviour, rather than a predictive one that looks down the road, and a future vehicle behaviour. That being said, if their system can genuinely respond in 1/500 second, then the suspension would be responding almost instantly, which undermines the value of having all the predictive garb that my earlier post outlined.
Interesting posts, and thanks for the info on LR.
|
That being said, if their system can genuinely respond in 1/500 second
I think that's a big if!
Doubtlessly the controller is sampling at 500 Hz, but, I struggle to see the heavy suspension of an off road vehicle reacting at anything like that speed. As an example, with conventional suspension, I would be surprised if the wheel hop frequency were above 15Hz.
Having said that, if the system monitors hub and body accelerations, and has a well specified actuator, the really clever work is in defining the algorithms in the controller to define an appropriate output at any given time. I remember an ex-colleague working on exactly this type of algorithm over a decade ago.
|
I don't know how it translates into algorithms, NC, but what is needed for comfort and decent roadholding is a long-travel system that won't react at all to bumps or holes smaller than the suspension travel, just let the wheels follow the road surface. Only if they are bigger than that is there any need for real - or virtual - springs and dampers. I don't think there's a perfect system, Citroen or other. Haven't driven enough cars to know really.
Followed a BMW M3 over some speed bumps in Islington tonight. It looked to me as if its dampers were completely shot. Should have banged over them like a child's trolley but did three bounces over those broad speed bumps. Or perhaps it was just a clapped-out 3 series with a badge on the boot? Got a glimpse down the side of it and it looked like the real thing. But that's the trick, eh?
Edited by Lud on 16/10/2009 at 01:17
|
I remember reading about an adaptive system years ago.
There was a prototype system fitted to an XJS (the argument being that if it could improve on that then it *was* a good product).
I seem to remember that it used hydraulics to stiffen up and keep the car level. The big point was that it made a huge difference in cornering without impacting comfort in general driving.
Some sort of derivative of the citroen systems?
|
The traditional citroen system uses the pressure of they hydraulic fluid, that is also used to jack the suspension up and down. I don't know about the more advanced active systems
You can do the same tricks with air I guess, and changing the air pressure in the system should adjust the spring rates (i.e. stiffness)
I think it might be Delphi that have developed a system that uses magneto-rheological fluid - i.e. a fluid that changes its viscosity when an electromagnetic field is applied. i.e. imagine a liquid that has the consistency of water but when you turn on an electromagnet suddenly has the consistency of your mother-in-laws gravy.
This replaces the conventional oil in the dampers, and basically what happens is when the car realises it's hurtling into a bend, it applies a magnetic field to the damper fluid and the suspension magically stiffens. I expect the response time of that kind of technology to be limited by the cars ability to work out what is required, as applying the magnetic field should give instant change to the viscosity and therefore the damping.
|
Is predictive suspension possible ?
>>............
>>>>>>>>>>>only on predictive tarmac............text ends..........
|
We perceive the bump and react to it when our brain has registered the fact, which must be a short while after the car actually hits it. So the system does not actually have to be predictive, merely have a reaction time faster than ours.
If I hit my thumb with a hammer, I don't feel the pain until a second or two later.
|
|
I can envisage that if predictive suspension is ever developed and applied toa car, it will work less well than conventional suspension - a bit like how the tilting mechanism on the experimental Advanced Passenger Train made passengers feel ill or spill their drinks - although the Pendolino trains seem to tilt OK (I haven't ridden on one).
|
|
