Return of active suspension - 2017

[FW17]

What sort of features can we expect that the suspension will do?

Anti-dive during braking?
Lean into corners?
Always maintain a near perfect distance from the ground?

Probably none of that. Might not even be active in qualifying and race. A system with hydraulics which helps find a good setup with probably a few adjustments from the driver like the movable front wing.

I'm sorry, but you will have to explain your thoughts, WF1. How can an hydraulic system not be active. Most vehicles (even non-aero vehicles) require springs and dampers to control the sprung & unsprung masses. These can be passive (as per the Dernie system) or active. CW stated that "you can throw away all your springs, dampers and roll bars", so it follows that he thinks that springs, dampers and bars can be simulated actively. They certainly can, as was demonstrated by the Lotus system, but there are consequences some of which I referred to above. Over to you....

p.s. the MP4/8 was full of springs and dampers, apparently. It looked to me suspiciously like a "Mumford" system - roughly equivalent to a powered "FRIC"....

I meant a simple semi active type dampers common on bikes to change the damping strength.

With regard to the system locked in quali and race I meant that once the driver finds the best damping setup during practice the same is not changeable. Idea of this system is for quick set up of the car.

Was not aware that CW had mentioned the same to be a full hydraulic system

[Tim.Wright]

Thinking in the line of having a mechanical suspension trying to do so many job at once whereby with active suspension it could be dialed in and done. Would not quite have a third spring, anti-roll bar, FRIC, J-damper or a number of dampers localized or interconnected and the amount of work to tune that bit of interconnection maze is never ending. How long did it take Mercedes to get use to FRIC and all those time cost a good fortune. With active suspension, after all the math is dialed in, its just a matter of tuning it because it opens up a whole lot of variations rather than going through a never ending labyrinth

Active will be the biggest labyrinth imaginable. The "dialling in of the maths" will likely be an extremely tricky and expensive process.

Consider also, that since cars have been running passive spring damper systems for decades, there would be very few people who would know what a suspension should do if you remove the limitations of a passive system.

[DaveW]

... I meant a simple semi active type dampers common on bikes to change the damping strength.

With regard to the system locked in quali and race I meant that once the driver finds the best damping setup during practice the same is not changeable. Idea of this system is for quick set up of the car.

Was not aware that CW had mentioned the same to be a full hydraulic system

You are correct, in part. I am quite sure that the FIA's plan would be to lock down the system in some way, but the system would include (simulated) springs, bars, travel limiters (bump rubbers & packer gaps), ride height settings, geometry as well as dampers.

The issue for the teams is that, with the current designs, it is difficult to make suspension set-up changes in the time between free practice sessions. The issue for the FIA is that it nearly impossible to assess different suspension implementations independently (I think they have to ask, & believe what they are told, unless someone objects - remember backward looking cameras, and the lengths one team took to try & prevent it?).

Hence a "full" active suspension solution makes sense to both parties. It would also be a way of ensuring that all teams have similar functionality (though, no doubt, some will use the facility better than others).

What the FIA must do is to acquire the expertise to interpret executable code and the effect it has on suspension performance (I'm not thinking of offering)...

[FW17]

I hope if it is a full active system, it is not standardized like the ECU to one company.

Would prefer all teams allowed to develop their own system (with one software system provided by the FIA) but made available to any competitor at a set price.

[henry]

What the FIA must do is to acquire the expertise to interpret executable code and the effect it has on suspension performance (I'm not thinking of offering)...

Do I take it from this that you expect the active system to emulate a standardised physical system? I assume in this case there would only be a very limited set of sensors allowed. Perhaps only those associated with wheel movement and actuator positions and forces? This would explicitly bar things like road speed, air speed, vehicle attitude, tyre temperatures and probably others. I guess problems then occur when people try to infer things like road speed or attitude from what they are allowed to use. Hence the need to interpret the code?

[Tommy Cookers]

active has the potential as a driver aid of car control in cornering by dynamic control of 'corner weighting'
eg reducing or eliminating those opposite-lock moments and others
and similarly potential covertly to emulate the benefits of ABS and traction control

so there could be a lot of code to be studied and understood and controlled by rules and ad-hoc rulings
yes, intelligent control is helped greatly if it has inputs of real-world data from sensors, and these can be limited

also, banning hydraulic actuation is easy and could be seen as a useful precaution limiting covert driver-aid behaviours
though even 30 years ago Moog was recommending electro-mechanical actuation over hydraulic around this sort of force level

[DaveW]

Do I take it from this that you expect the active system to emulate a standardised physical system? I assume in this case there would only be a very limited set of sensors allowed. Perhaps only those associated with wheel movement and actuator positions and forces? This would explicitly bar things like road speed, air speed, vehicle attitude, tyre temperatures and probably others. I guess problems then occur when people try to infer things like road speed or attitude from what they are allowed to use. Hence the need to interpret the code?

Mmm. I think that the only transducers we used that are not generally available now were hub vertical accelerometers. We had available pressure inputs for estimating D/F, but these were not used (we couldn't find a representative location for the pressure taps) and we used scaled pitot pressure for D/F with two probes linked to "average out" wake effects. We also used a yaw rate sensor (unusual at the time, but commonly available now). We had ride height sensors fitted, but they had variable transport delays, so they were used off-line to compute tyre stiffness, and loads & ground speed were used to estimate tyre deflection. Apart from that we used four "strut" loads, four strut positions (LVDT's), airspeed, ground speed, RPM, lateral & longitudinal accelerations, and hydraulic supply pressure. The driver had a single control that we could patch to any variable.

Other parameters were measured, but were not used by the active system.

Loads were the principle suspension input, and therefore received much attention. We designed an build our own load cells and high quality SCU's, but inevitably they responded to parameters we didn't want to see, and the didn't measure things we did. The loads were corrected for all deterministic effects.

Offsets (both of our making & actuator offsets) were removed by averaging.

Other than that, as it happens, we chose to emulate a "modal" suspension (i.e. heave, pitch, roll and warp springs and dampers) - but a conventional suspension emulation was always an option, requiring no hardware changes.

Failures were detected by controlling "model" actuators & using the outputs of these to drive the physical actuators. Differences were used to identify failures - which were notified to the driver & hydraulic supply was then "shorted out", and the car continued supported by physical springs, with the actuators acting as (not very good) dampers. A interesting failure case was puncture detection, the system was able to detect a slow puncture a good lap before the driver.

What the FIA will allow in 2017 is not for me to guess.....

[DaveW]

....athough even 30 years ago Moog was recommending electro-mechanical actuation over hydraulic around this sort of force level

I don't think so, Tommy. Force = pressure* CSA. We worked very closely with Moog at the time, & they never once mentioned the option. Bill Moog set up a facility in Florida specifically to research hydraulic actuation of vehicle suspensions....

[Powerslide]

Active will be the biggest labyrinth imaginable. The "dialling in of the maths" will likely be an extremely tricky and expensive process.

Consider also, that since cars have been running passive spring damper systems for decades, there would be very few people who would know what a suspension should do if you remove the limitations of a passive system.

Like every software/hardware, the beginning and development process would have the most effort but once everything is in place and with the amount of freedom active suspension offer mechanical devices post challenges, in the long run I would still think active might be cheaper to maintain. A lot of software already have advance physics dialed into them anyway and to transfer that to a analog mechanical system requires far more effort rather than potentially having an active system and simply hooking up to a computer. Look how far fuel injection have gone compared to carburetor.

[Tommy Cookers]

@ Dave W
DP Rose of Moog in the UK was writing this in 1985, though the word 'around' was my considered choice
I had already found in my work then electromechanical actuation that replaced servohydraulic equivalents

I have always held servo-valved hydraulic actuation necessary for F1 active to realise the fullest potential of active
however it seems that F1 intends to limit the potential

active does not mean springless except maybe to the FIA
IMO a good, intelligent active 'road-relevant' system could result from using electromechanical actuation with colocated springs
and this would serve all the FIAs currently apparent purposes and so be fine in F1
it seems bizarre to demand servohdraulic actuation (struts) and then limit their operation to less than their potential

@ Powerslide
advanced physics-based software functionality is a poor substitute for input from real sensors ?
(if the FIA severely constrains sensor numbers and type)

[henry]

.

Other parameters were measured, but were not used by the active system.

........


What the FIA will allow in 2017 is not for me to guess.....

Thanks for a rather more comprehensive reply than I was expecting

Do you think that you would use any other parameters if you were implementing now? I'm thinking tyre temperature or track position for instance. Or any of the others you measured but didn't use?

On the topic of the FIA Regulations I can think we can make the assumption that they will want to tie down whatever they are to prevent any team gaining a significant advantage over the others. They must be smarting over what has happened with power units this year.

In my naivety I imagine that it might be possible to define a fixed set of inputs, and the associated sensors, and then stipulate allowable responses to combinations of input values. But then I think that, whatever, there are lots of cats and only one mouse.

[DaveW]

... IMO a good, intelligent active 'road-relevant' system could result from using electromechanical actuation with colocated springs and this would serve all the FIAs currently apparent purposes and so be fine in F1
it seems bizarre to demand servohdraulic actuation (struts) and then limit their operation to less than their potential...

You mean like the Aura Systems suspension (probably now known as the "Bose" system), I guess. That has been around for 20 years or so, with only a prototype installation to show for it. I don't know how much it weighs, but it would probably be too heavy for F1.

I'm unsure why you have concluded that the performance of an active system would be limited to less than its potential, but you may be right....

Do you think that you would use any other parameters if you were implementing now? I'm thinking tyre temperature or track position for instance. Or any of the others you measured but didn't use?

GPS was not available for civilian use in 1987. However, one of my aerodynamicist friends (sorry John) wanted me to use "dead reckoning" to compute a track map and then to vary ride height with distance to achieve the lowest possible ride height around the circuit. That would be easier now with GPS, and I am sure the idea would appeal to many (but not to me).

An issue we discovered part way though the season (at Mexico I think) was that we were not working the tyres hard enough. A slightly crazy idea was to excite the tyres rapidly in warp to heat them whenever the tyres were cool and the vehicle wasn't maneuvering.

The last was vetoed for fear of reprisals, but both do indicate how closely the software would have to be monitored.

[Cold Fussion]

From a team perspective, what sort of performance benefits would a fully active system bring compared to the former interconnected systems?

[Hovepeter]

From a team perspective, what sort of performance benefits would a fully active system bring compared to the former interconnected systems?

I've heard it's worth seconds to have a full active suspension on

[Pierce89]

From a team perspective, what sort of performance benefits would a fully active system bring compared to the former interconnected systems?

I've heard it's worth seconds to have a full active suspension on

That would only be once the aero is optimized for the narrower ride height window they would be able to maintain.