Four wheel steering (4WS)

[JuanJo]

When implementing a 4WS in a suspension, What shall be taken in consideration? Which parameters? What would change?
I think is not an easy adaptation so I wish someone could help me to understand better this measure.

Extra information: the current suspension is a pushrod/pushrod suspension type, with 4 outboard (in-wheel) engines, thus 4WD. The system that will generate the 4WS is based on an actuator system that pushes the tie rod in the rear during cornering.
Thanks for your time.

[Pierce89]

If you're going for laptime, the best thing you can donly with 4ws is eliminate it to save weight, because you will not make the car faster by adding 4ws. It's been proven over the years that noone who's tried it has gained any legitimate lap time from it.

[Greg Locock]

Exactly. If the purpose of circuit racing is to exploit the friction circle at all times usefully then it is hard to see what 4ws adds to the mix. You could, I suppose, use it to compensate for massive deficiencies in design elsewhere, but I'd have thunk the obvious solution would be to sort the design out, not overlay another system which has proved to be of little value except for fork lift trucks.

[strad]

It's been proven over the years that noone who's tried it has gained any legitimate lap time from it.

Now Now,, I was just watching a show with Tommy Kendall and some others talking about how cool the new 4WS is on the new Porsche and how they wish they could have it on their race car. Took some cats around Barber motorsports park.
They say it's really cool.

[DiogoBrand]

Why would Porsche use it and praise it so much on their 918 if it has no technical advantage?

[bill shoe]

Extra information: the current suspension is a pushrod/pushrod suspension type, with 4 outboard (in-wheel) engines, thus 4WD.

You have the means for a pretty good 4WS system right there. Make the steering wheel an electrical device that dials up side-to-side differential torque proportional to the steer wheel angle.

The steering by itself would always add net zero forward torque. So if you are stationary with no accel or brake application, you can turn the steer wheel to the right. This gets you get positive torque on left and negative torque on right. Your car slowly starts rotating to the right (clockwise in plan view), although with much tire scrub. Accel or brake simple adds pos or neg torque to the steer values that were dialed up by the steer wheel.

Has anyone tried this? Probably someone has. It's just tracked-vehicle steering on a 4-wheel car. But interested to know about any results.

[Greg Locock]

Why would Porsche use it and praise it so much on their 918 if it has no technical advantage?

Because it sounds interesting and techie. And no journo is going to take the piss out of it, as it gives them something interesting and techie to write about. And they all say Porsche is great, publicly, otherwise they don't get Porsches to play with for a week, and invitations to launch events.

[DiogoBrand]

But why would they actually want to add weight into something with no performance benefits?

[Per]

Extra information: the current suspension is a pushrod/pushrod suspension type, with 4 outboard (in-wheel) engines, thus 4WD.

You have the means for a pretty good 4WS system right there. Make the steering wheel an electrical device that dials up side-to-side differential torque proportional to the steer wheel angle.

The steering by itself would always add net zero forward torque. So if you are stationary with no accel or brake application, you can turn the steer wheel to the right. This gets you get positive torque on left and negative torque on right. Your car slowly starts rotating to the right (clockwise in plan view), although with much tire scrub. Accel or brake simple adds pos or neg torque to the steer values that were dialed up by the steer wheel.

Has anyone tried this? Probably someone has. It's just tracked-vehicle steering on a 4-wheel car. But interested to know about any results.

https://www.youtube.com/watch?v=EaKmdz_CV6w

This is a car without 4WS. But like you say, with 4 electric motors the big gain is in controlling the torque for each wheel individually based on longitudinal and lateral acceleration. Theoretically this allows you to keep each tyre on the limit of friction al throughout the lap (as long as you're not power limited of course).

[Andres125sx]

Can´t help in any regard, but I´ve always found interesting the system BMW use on the GT series wich moves the rear wheels to add steering torque when in a slow corner, and opposite on fast ones. I´ve always thought it would be usefull to compensate the usual unstability at high speed, or lack of maneouverability at low speed, or even both

Does it really achieve some of these or it´s just my imagination?

[bill shoe]

Has anyone tried this? Probably someone has. It's just tracked-vehicle steering on a 4-wheel car. But interested to know about any results.

https://www.youtube.com/watch?v=EaKmdz_CV6w

This is a car without 4WS. But like you say, with 4 electric motors the big gain is in controlling the torque for each wheel individually based on longitudinal and lateral acceleration. Theoretically this allows you to keep each tyre on the limit of friction al throughout the lap (as long as you're not power limited of course).

Great car, but did they use all that great potential to just do donuts?

There is little in fundamental road car vehicle dynamics that hasn't been explored and pretty well understood. I think it would be quite interesting to start fleshing out the fundamentals of torque steering for high performance applications. I've always thought that really high performance driving was an exercise in yaw management, and differential torque steering seems like it may offer a more direct way to control this at the limit.

I did a 5-minute google search and found a couple papers about using differential torque steering (via braking torque only) as part of a stability control system on a road car with otherwise ordinary steering. Also an endless stream of papers about how to make robots track around corners smoothly. Nothing that seemed like a high-performance perspective.

Please pipe up if anyone finds something or knows something. I suspect some work has been done here and I just haven't found it.

[Greg Locock]

Torque vectoring is a pretty standard part of a RWD production car's electronic stability control these days, and offers most of the benefits of a rack on the rear axle without any extra weight, and better reliability.

If we can agree for the time being that it is possible to set a car up for either high speed or low speed regime acceptably, passively, and the trick is to say improve low speed yaw gain (ie tight corners) whilst not compromising high speed understeer (say) then the passive options are limited to :

aero
variable ratio steering gear (Bishop)
ackerman
tire construction


The active options include

active variable ratio steering gear (BMW)
epas tuning - you can add damping to the high speed response,
epas - change the SWT required for a given assist on a speed dependent table, which subjectively alters the feel of the car albeit a demand driver won't notice.
torque vectoring
esc, which mostly uses single wheel braking
I think magnaride shocks would probably be useful, as would active suspension.
and of course 4 ws

I'd have thought the wiki site would have some handy links https://en.wikipedia.org/wiki/Electroni ... ty_control

[Per]

Great car, but did they use all that great potential to just do donuts?

Of course not, this was just a fun demo after the racing season. That car won the electric class at Formula Student Germany and set a (now obsolete) Guiness world record for fastest accelerating electric car (0-100kph in 2.13s).

It will indeed be tough to find proper literature on the vehicle dynamics of such cars. But this one had more than decent torque vectoring where, like you say, yaw rate control was what it was all about. Plus slip ratio and slip angle control for each individual wheel, I believe. Sadly I'm not exactly a vehicle dynamics specialist myself.

In 2012 this car was a trendsetter in Formula Student but this stuff is pretty standard now among the top electric teams. The team from ETH Zürich showcase the implementation of an inertial measurement unit here: http://www.lp-research.com/lpms-cu-in-e ... -fluelela/

[Greg Locock]

I'm a bit puzzled by this reported dearth of information on ESC, it seems to me that years ago every man and his dog was writing papers about it (which I certainly never bothered to read). A google search on sae ESC papers gets pages of hits, and other acronyms are used instead of ESC.

[Thunder18]

As with a lot of innovations that Honda seem to introduce but get no credit for, the Prelude back in 1987 was the first production road car to offer 4ws to the masses.

http://www.drive.com.au/it-pro/mileston ... 916-10hpaj

http://world.honda.com/history/challeng ... index.html

They also now have the Acura NSX and RLX/Legend Hybrid which although doesn't offer 4ws, but does off torque vectoring through electric motors in the front (NSX) and rear (RLX/Legend)

The first mechanical iteration now praised by journalists since the Ford Focus RS was recently launched with something similar
http://world.honda.com/automobile-technology/SH-AWD/

Now hybrid-ized with electric motors.
http://world.honda.com/automobile-techn ... D2/topic1/

Plenty of topics in outside forums to aid you in your quest too.

http://www.preludeonline.com/f61/2ws-4ws-191356/

Though whether or not the system is faster on track is still questionable, but the reviews seem to think that it does offer a more engaging drive.