Limited slip differentials vs electronic control of brakes?

[riff_raff]

An LSD is torque limiting but it also vectors torque.
It vectors the lost torque from the slipping wheel to the gripping wheel, if it did not it would be an open diff and the torque would be lost resulting in no drive and powertrain over reving.

There is not torque loss from the clutch slippage in an LSD, there is power loss. Thus there would be no possibility to "recover torque". With a simple geared diff, the differential allows a relative difference in rotational speed/direction between two outputs from a common input. And the torque transfer is limited by the resisting force produced at each output point. By definition, a simple geared diff system must operate in equilibrium. Ignoring system frictions, if one tire slips at a given level of torque, then the other tire will be limited to the same amount of torque input.

LSD systems that use clutches limit the torque applied to the wheel turning at a higher rate via a mechanical device that proportionally unloads the friction clutch in response to the relative rate of rotational speed difference. The only system I can think of that would produce true "torque vectoring" would be a TC system that uses active braking of each wheel. By actively applying a braking force to the unloaded wheel, up to 100% of the available driveline torque could be "vectored" to the wheel with traction.

[autogyro]

There is not torque loss from the clutch slippage in an LSD, there is power loss. Thus there would be no possibility to "recover torque". With a simple geared diff, the differential allows a relative difference in rotational speed/direction between two outputs from a common input. And the torque transfer is limited by the resisting force produced at each output point. By definition, a simple geared diff system must operate in equilibrium. Ignoring system frictions, if one tire slips at a given level of torque, then the other tire will be limited to the same amount of torque input.

LSD systems that use clutches limit the torque applied to the wheel turning at a higher rate via a mechanical device that proportionally unloads the friction clutch in response to the relative rate of rotational speed difference. The only system I can think of that would produce true "torque vectoring" would be a TC system that uses active braking of each wheel. By actively applying a braking force to the unloaded wheel, up to 100% of the available driveline torque could be "vectored" to the wheel with traction.

Not that torque power thing again.
A simple open diff does not distribute 'power' evenly when the vehicle is turning.
One wheel turns faster than the other.
The 'torque' available' is evenly applied to the gearset but the output 'power' is different.

Most platted LSDs lock the clutch pack as the slipping wheel increases in rpm, locking the diff as a spool diff.

Modern designs of mechanical torque vectoring diffs can use brakes, clutches or both in various combinations.
It is the counter shaft in the design that allows controlled torque vectoring.
If you like it converts up to 'power' and back down to a different torque figure than the input via its own gearset.

[JoaoB]

Has anyone read the Mitsubishi technical paper that Autogyro pointed?

I understood the explanations but not the graphs. Anyone did?

[jamsbong]

I thought the F1 diff is a electronic diff which adjust the car's rear diff so that it will feel loose or tight.
It is not smart like the old traction control days where the car literally corrects your action by managing the throttle based on your steering feedback and cornering G. Similar systems are found in GTRs, EVOs and F430 onwards.

The new Pseudo limited slip diff system that uses brakes to stop the inside wheel from spinning was first introduced by VW. This was quickly adopted by Renault, Ford and McLaren (road car). To me, this is just a budget LSD which does not work that well.
All the system does is to apply moderate brake forces on the inner wheels so that the wheel will not rotate faster than it should.
This effectively waste energy which should be put to the road and wears a bit of your brake pads. Moreover, you can't accurately adjust the amount of torque distributed between the 2 wheels.
A proper LSD does its job on the diff distributing torque to the tire with more grip. This type of diff is more predictable especially on a RWD.

[thisisatest]

In my mind, the ideal would be a diff that contains a type of cvt between the driveshaft and the two halfshafts. I see a spherical traction drive system, if it could handle the torque. When the inside wheel starts slipping, some OTHER mechanical device (or electronic control) would tilt the drive system to slow down the inside wheel and speed up the outside wheel, without the losses incurred from just hitting the brakes on the inside wheel. How the unit senses slipping, I don't know. maybe the drive speeds would be split based on steering wheel input, or a combination of that and where the wheels are in their suspension travel. I would personally want to have the fewest variables influencing it, so it stays simple and predictable...

[autogyro]

In my mind, the ideal would be a diff that contains a type of cvt between the driveshaft and the two halfshafts. I see a spherical traction drive system, if it could handle the torque. When the inside wheel starts slipping, some OTHER mechanical device (or electronic control) would tilt the drive system to slow down the inside wheel and speed up the outside wheel, without the losses incurred from just hitting the brakes on the inside wheel. How the unit senses slipping, I don't know. maybe the drive speeds would be split based on steering wheel input, or a combination of that and where the wheels are in their suspension travel. I would personally want to have the fewest variables influencing it, so it stays simple and predictable...

Any CVT system requires energy to operate the cones or discs.
The losses would be nearly as high if not higher than using brakes
It also introduces the need for special oils and more cooling.

Hydraulic designs are better but the best would be electromagnetic.
Both these types are almost impossible to police to prevent the use of traction control.

Powertrain development continues to struggle under the mask of high downforce and highly restrictive regulation.

[riff_raff]

In my mind, the ideal would be a diff that contains a type of cvt between the driveshaft and the two halfshafts. I see a spherical traction drive system, if it could handle the torque. When the inside wheel starts slipping, some OTHER mechanical device (or electronic control) would tilt the drive system to slow down the inside wheel and speed up the outside wheel, without the losses incurred from just hitting the brakes on the inside wheel. How the unit senses slipping, I don't know. maybe the drive speeds would be split based on steering wheel input, or a combination of that and where the wheels are in their suspension travel. I would personally want to have the fewest variables influencing it, so it stays simple and predictable...

In theory, the ideal drivetrain would be one that drives each wheel directly and independently, such as an electric or hydraulic system with a drive motor at each wheel. As for F1, there have been many types of diffs used over the years. There were the old cam/clutch systems used by Hewland and others, which could be tuned by changing the pressure angle of the ramps. More recently, there were the hydrostatic systems that used hydraulic pressure to transfer torque from one axle to the other, and they could be tuned using valving of the hydraulic flow.

[autogyro]

In theory, the ideal drivetrain would be one that drives each wheel directly and independently, such as an electric or hydraulic system with a drive motor at each wheel. As for F1, there have been many types of diffs used over the years. There were the old cam/clutch systems used by Hewland and others, which could be tuned by changing the pressure angle of the ramps. More recently, there were the hydrostatic systems that used hydraulic pressure to transfer torque from one axle to the other, and they could be tuned using valving of the hydraulic flow.

A drive motor at each wheel would drastically increase the unsprung weight.
Electric or hydraulic motors would need to have sufficient capability to supply full engine power.
They would be very heavy and would place far to much mass at one end of the car.

I have designed pure hydraulic diffs and they work well depending on the control system used.
They do need sufficient cooling however and are heavier than mechanical types.
(1980s not new, and in plant and heavy vehicles of course)
The best method is electro magnetic, which can be lighter than pure mechanical and with a huge range of control.
Of course not in F1 where the regulations keep powertrains firmly in ancient history.

There is a huge potential development area in all types of powertrain and differential drives using electro magnetics.
Manufacturers have yet to invest properly in the technology mainly because much of it will be sourced from EV and Hybrid developments as these progress with the limited budgets allowed.
It all cuts against the manufacturers mainstream established production and doesnt sell oil.
Much of this could have been achieved decades ago.

[thisisatest]

I just want a diff system that does not "transfer" torque by attempting to equalise wheel speed left-to-right. it seems like a cop-out to me.

[autogyro]

I just want a diff system that does not "transfer" torque by attempting to equalise wheel speed left-to-right. it seems like a cop-out to me.

The current torque vectoring diffs do not work by attempting to equalise wheel speed.
They transfer torque from one side to the other using outside control input.

[riff_raff]

I just want a diff system that does not "transfer" torque by attempting to equalise wheel speed left-to-right. it seems like a cop-out to me.

A basic gear differential functions by allowing independent rotational speeds at two separate outputs from a common input. While this device also distributes torque between the two outputs (wheels), by definition the torsional moment at each wheel must be in equilibrium even though the rotational speeds may not be. In other words, with a simple gear diff, you cannot apply more torque to any one wheel than the amount of torque which can be resisted by the opposing wheel's traction. In this regard, LSD's are no different. LSD's don't increase (or "vector) more torque to the wheel without slippage, LSD's only limit the torque transferred to the wheel with slippage.

On the other hand, an active independent braking system would prevent both loss of traction at each driven wheel. And it would also allow the diff to apply more torque at the wheel without slippage, since the resisting torque at the unloaded wheel would be the combined capabilities of the brake and tire.

[autogyro]

On the other hand, an active independent braking system would prevent both loss of traction at each driven wheel.
And it would also allow the diff to apply more torque at the wheel without slippage, since the resisting torque at the unloaded wheel would be the combined capabilities of the brake and tire.

The system you describe is torque transfer brake steer, as developed by Gorden Murray for McLaren F1.
It absorbs to much torque in the brakes to be used simply as a passive mechanical differential.

A torque vectoring diff can use brakes and or clutches to transfer torque via a torque transfer counter shaft.
An electronic control system operates the brakes and or clutches.