Driver-adjusted differential locking

[autogyro]

Normal, yes. But what about dual clutch electronically controlled? And I think as long as this is torque distribution summing up to the amount of input torque, it is not regarded as traction control.

There would then be no need for a conventional diff.
It would be a 'spool' axle with a controlled slipping clutch pack on each wheel shaft.
A type of traction control.
It is much easier to control the torque to each wheel with hydraulic pumps in place of the friction plates (wet or dry).
You can even make use of the heat extracted.
Still a form of traction control though.

Got a much better way now using the magnetics in an energy harvesting powertrain.
Gives perfect control over rear torque steer as well.
Its all pointless in F1. The diff control is only needed to balance the huge downforce under braking and cornering.
Anymore tech would be masked by the aero.

[928S]

Tim the drawings I have seen that are about 18 months out of date and recents comments from Mark Webber at the Hungarian GP suggest otherwise. Mark said that one side on the diff failed which effected the cornering when going around the relevant corners that you would need those clutches working, i.e the opposite side. I have seen the drawings of the Red Bull and they have clutches either side of the differential. I just can't remember about the other planetary gears, however I am not saying they are not there either. So even though what I saw was out of date now, Mark's reference confirms they are still using the same system.

Brake steer is no myth and I was sort of on the right path, in fact the diff is torque vectoring as seen here in this Audi demonstration. This is the reason the diff helps turn in and general corning stability.

A torque vectoring diff is something different to a normal LSD. This is what we would like to know, if this is allowed in an F1 differential, but I have the impresison that its not.

To get torque steer from one of these, there is a gear stage which steps the rotational speed of the clutch so its faster than the carrier speed. Then you can send torque to the left or right by operating one of two independant clutch packs.

Yes there are advantages but also massively increased complexity. Diffs could be a bottomless pit for development money if its allowed, and thats the main reason why they are restricted now.

Id be surprised if such a torque vectoring diff is allowed in F1.

Tim

[Tim.Wright]

Tim the drawings I have seen that are about 18 months out of date and recents comments from Mark Webber at the Hungarian GP suggest otherwise. Mark said that one side on the diff failed which effected the cornering when going around the relevant corners that you would need those clutches working, i.e the opposite side. I have seen the drawings of the Red Bull and they have clutches either side of the differential. I just can't remember about the other planetary gears, however I am not saying they are not there either. So even though what I saw was out of date now, Mark's reference confirms they are still using the same system.

Normal Salisbury type diffs have can have clutches on both sides. And yes, if one clutch pack stops working for some reason it will affect the torque bias, but still, such a diff will only ever bias the torque to the slower wheel.

If anyone has the Ferrari book by Peter Wright, there is a good section view of the diff. Though I don't remember seeing any step up phase before the clutches. My book is in another country at the moment. Greg, do you have your copy handy?

Tim

[autogyro]

The books written by Alec Stokes (ERA BRM Lotus) are more useful to you.
Peter Wright worked with Alec in his early career.
Alec pioneered F1 gearbox technology.
His ideas were used on such cars as the BRM H16, in the days when powertrain development meant something more than aero.

[928S]

Spoke yesterday to a F1 engineer who worked in this area 6 years ago, he confirmed it is a torque vectoring differential designed to give maximum corner speed and excellent drive out of the corners. He said it was a closed loop system back then, it monitors the wheel speeds with strain gauges on the driveshafts so they know exactly how much torque is going to each wheel.

[autogyro]

Spoke yesterday to a F1 engineer who worked in this area 6 years ago, he confirmed it is a torque vectoring differential designed to give maximum corner speed and excellent drive out of the corners. He said it was a closed loop system back then, it monitors the wheel speeds with strain gauges on the driveshafts so they know exactly how much torque is going to each wheel.

Monitoring the wheel speeds will not tell you the torque distribution.
It will tell you when to reduce torque to a wheel if the wheelspeed on one side rapidly increases due to wheelspin.
That is traction control.
If you are measuring with a strain guage (usualy a magnetic strip that bends with a shaft), it will give you the torque being transmitted through the shaft. A sudden reduction in the torque read will also show wheelspin and a control action taken from this, forms the basis of another form of traction control.

You cannot measure wheel speed with a strain guage.

Torque vectoring can be achieved in many ways, the F1 criteria is not to achieve it by the best means but to achieve close to it within the regulations and without it being defined as traction control.

[928S]

Spoke yesterday to a F1 engineer who worked in this area 6 years ago, he confirmed it is a torque vectoring differential designed to give maximum corner speed and excellent drive out of the corners. He said it was a closed loop system back then, it monitors the wheel speeds with strain gauges on the driveshafts so they know exactly how much torque is going to each wheel.

Monitoring the wheel speeds will not tell you the torque distribution.
It will tell you when to reduce torque to a wheel if the wheelspeed on one side rapidly increases due to wheelspin.
That is traction control.
If you are measuring with a strain guage (usualy a magnetic strip that bends with a shaft), it will give you the torque being transmitted through the shaft. A sudden reduction in the torque read will also show wheelspin and a control action taken from this, forms the basis of another form of traction control.

You cannot measure wheel speed with a strain guage.

Torque vectoring can be achieved in many ways, the F1 criteria is not to achieve it by the best means but to achieve close to it within the regulations and without it being defined as traction control.

Excuse my poor use of english, it was just meant to be they have wheel speed sensors and also strain gauges on the driveshafts for their respective applications. Not that strain guages are monitoring the wheel speeds.

[autogyro]

Excuse my poor use of english, it was just meant to be they have wheel speed sensors and also strain gauges on the driveshafts for their respective applications. Not that strain guages are monitoring the wheel speeds.

No my apology, your use of English was fine and your description was also.
I am sorry my explanation read as a contradiction it was not meant to be.

[strad]

Magic Diff???
AS you see it doesn't exactly work as traction control but more closely resembles McLarens third brake pedal
click photo

[Tim.Wright]

Spoke yesterday to a F1 engineer who worked in this area 6 years ago, he confirmed it is a torque vectoring differential designed to give maximum corner speed and excellent drive out of the corners. He said it was a closed loop system back then, it monitors the wheel speeds with strain gauges on the driveshafts so they know exactly how much torque is going to each wheel.

This is what I wanted to hear. So that means these differentials are capable (theorectially) of biasing torque in either direction. I'm interested to know how the rules restrict the use of such a system, because torques vectoring opens up a massive amount of possibilities.

Tim

[autogyro]

We could have used full torque vectoring in 1976.
The only difference today is the better electronic control.

[hardingfv32]

So that means these differentials are capable (theorectially) of biasing torque in either direction. I'm interested to know how the rules restrict the use of such a system, because torques vectoring opens up a massive amount of possibilities.

Tim

Wouldn't 9.9.1 preclude this? "9.9.1 Any system or device the design of which is capable of transferring or diverting torque from a slower to a faster rotating wheel is not permitted."

You also have an issue with the ECU programing being monitored by the officials.

Brian

[Jersey Tom]

To be fair, the quote said the guy worked in the field 6 years ago. Possible the regs have changed since then.

Also possible for there to be a slight misinterpretation in nomenclature or misquote... as to what "torque vectoring" (or any similar term) implies.

[Tim.Wright]

Ok those rules effectively ban torque vectoring. So it seems most likely they are a clutch type LSD with a (hydraulically or electrically) controlled plate pressure.

To me, torque vectoring is the possibilty to split the torque left and right however you want.

Tim

[strad]