# Planetary centre diffs with uneven torque distribution

15 replies to this topic

### #1 Malmedy

Malmedy
• Member

• 33 posts
• Joined: May 10

Posted 23 December 2011 - 17:36

Does anyone know how a constant uneven torque distribution is achieved in this type of differential? I understand that for the torque distribution to be constant the torque flow must be independent of relative speeds between output shafts, but in my mind this contradicts the fundamental laws of gearing (i.e. a larger torque increase is coupled to a larger reduction in circumferential speed). (Calculations or good texts very welcome)
Thanks

### #2 Powersteer

Powersteer
• Member

• 2,460 posts
• Joined: September 00

Posted 23 December 2011 - 20:30

If the outer inverted gear is where input is coming from, planet gears hub torque goes to the right wheel while an inner sun gear goes to the left then the right wheel would deliver more torque with its lower ratio. If the planet gears is twice smaller than the sun gear then distribution is back to even, 50-50. Thats the direction to change the torque bias, there or there abouts. Guessing here, I'm certainly not qualified to confirm this. If I am not mistaken the planet gears output is 2:1 ratio if all gears are evenly sized while the sun gear is 1:1 to its output shaft being solid.

Edited by Powersteer, 23 December 2011 - 20:36.

### #3 kikiturbo2

kikiturbo2
• Member

• 724 posts
• Joined: December 04

Posted 23 December 2011 - 21:04

here is a very nice explanation from the Lancia Delata integrale workshop manual..

Edited by kikiturbo2, 23 December 2011 - 21:05.

### #4 Malmedy

Malmedy
• Member

• 33 posts
• Joined: May 10

Posted 24 December 2011 - 22:30

here is a very nice explanation from the Lancia Delata integrale workshop manual..

Thanks a lot for the answers. That pdf is excellent! I have one more question though: the pdf states that the two planets only rotate relative to their own axis when no slip occurs which is a very strong simplification of course, as every vehicle acceleration is caused by tyre slip.

Imagine the following case:
- a vehicle with 50 - 50 weight distribution
- linear tyre characteristics
- a 30 - 70 % torque distribution at the wheels

This means the rear axle must rotate about 3 to 4 % faster than the front axle.
Does this difference originate in the diff itself so the satellites are not stationary in normal operation, resulting in a small torque loss?
OR is the speed difference generated by two different final drives at the front and rear axle?

Edited by Malmedy, 24 December 2011 - 22:31.

### #5 24gerrard

24gerrard
• Member

• 2,008 posts
• Joined: March 11

Posted 24 December 2011 - 23:27

If the tyre sizes are the same at each end and a different torque is required at one end of the car to the other, then the differential ration for the front two wheels must be different to the differential ratio for the rear two wheels.

The Lancia system is just a well packaged method of placing the central differential at one end (the front) and tightly packaging a pre-loaded plated slipper disc pack between the front and rear drive shafts.

For full limited slip on all wheels, it would need the addition of a disc pack between each road wheel drive shaft.

Edited by 24gerrard, 25 December 2011 - 11:01.

### #6 rory57

rory57
• Member

• 97 posts
• Joined: November 10

Posted 25 December 2011 - 13:58

Thanks a lot for the answers. That pdf is excellent! I have one more question though: the pdf states that the two planets only rotate relative to their own axis when no slip occurs which is a very strong simplification of course, as every vehicle acceleration is caused by tyre slip.

Imagine the following case:
- a vehicle with 50 - 50 weight distribution
- linear tyre characteristics
- a 30 - 70 % torque distribution at the wheels

This means the rear axle must rotate about 3 to 4 % faster than the front axle.
Does this difference originate in the diff itself so the satellites are not stationary in normal operation, resulting in a small torque loss?
OR is the speed difference generated by two different final drives at the front and rear axle?

Think of the static condition and therefore forget the tyre characteristics; assume that the tyres are all of the same diameter: the shaft speeds are equal (zero). The torque can be different either "side" of the differential if the differential (as in the Lancia example) is configured with pinions of different diameter and/or pinion centres at different diameters.

This is analogous to standing on a plank supported by two blocks: if I stand on the plank with a force of 100kg half way between the two blocks, each block will bear 50 kg (plus half the weight of the plank). If I stand 1/4 the way along the plank, one block will bear 25 kg, the other will bear 75 kg.

### #7 kikiturbo2

kikiturbo2
• Member

• 724 posts
• Joined: December 04

Posted 25 December 2011 - 14:07

Thanks a lot for the answers. That pdf is excellent! I have one more question though: the pdf states that the two planets only rotate relative to their own axis when no slip occurs which is a very strong simplification of course, as every vehicle acceleration is caused by tyre slip.

Imagine the following case:
- a vehicle with 50 - 50 weight distribution
- linear tyre characteristics
- a 30 - 70 % torque distribution at the wheels

This means the rear axle must rotate about 3 to 4 % faster than the front axle.
Does this difference originate in the diff itself so the satellites are not stationary in normal operation, resulting in a small torque loss?
OR is the speed difference generated by two different final drives at the front and rear axle?

any speed difference between the front and rear axles originates in the planetary diff itself, satelites are never stationary.. final drives are identical.. You also have a separate method of locking the front and rear drives..
in lancias case, it is a ferguson multi plate clutch... Other cars like Mitsubishi evo have a 50/50 torque split and a hydraulically operate metal multi plate clutch that server only for locking the front and rear drives togeather under controlled circumstances (i.e. varying degrees of lock/fully open, etc.)

if you need pics of the actual satelites, I can do that too..

### #8 Malmedy

Malmedy
• Member

• 33 posts
• Joined: May 10

Posted 27 December 2011 - 13:55

Think of the static condition and therefore forget the tyre characteristics; assume that the tyres are all of the same diameter: the shaft speeds are equal (zero). The torque can be different either "side" of the differential if the differential (as in the Lancia example) is configured with pinions of different diameter and/or pinion centres at different diameters.

This is analogous to standing on a plank supported by two blocks: if I stand on the plank with a force of 100kg half way between the two blocks, each block will bear 50 kg (plus half the weight of the plank). If I stand 1/4 the way along the plank, one block will bear 25 kg, the other will bear 75 kg.

Take the following static condition: The vehicle is driving 100 km/h and requires 200N to overcome driving resistances. This means the front wheels will each deliver 30N while the rear tyres deliver 70N. Now a linear tyre characteristic where at 0% slip 0N is created and at 10% slip 100N is created for the given loads (this is below the peak of the curve). This means the front tyres will constantly operate at 3% slip ie 103 km/h while the rear tyres operate at 7 % slip => rotational speed for rdyn = 107 kmh. With same tyre radius this means that the half shafts will not rotate at the same speed.

any speed difference between the front and rear axles originates in the planetary diff itself, satelites are never stationary.. final drives are identical.. You also have a separate method of locking the front and rear drives..
in lancias case, it is a ferguson multi plate clutch... Other cars like Mitsubishi evo have a 50/50 torque split and a hydraulically operate metal multi plate clutch that server only for locking the front and rear drives togeather under controlled circumstances (i.e. varying degrees of lock/fully open, etc.)

if you need pics of the actual satelites, I can do that too..

If I understand correctly in the example above the satellites will rotate, but the clutch should be inactive (otherwise this would result in very high wear, as the clutch would be permanently required and slipping). The side with the higher torque will rotate slightly faster causing the lower torque output to rotate slightly faster. Would love some pictures!

Thanks again for the responses and patience

Edited by Malmedy, 27 December 2011 - 13:57.

### #9 kikiturbo2

kikiturbo2
• Member

• 724 posts
• Joined: December 04

Posted 27 December 2011 - 15:11

ok,

this is what I have... I can make some more specifically for you..

centre diffs, complete... mitsubishi on the left, lancia on the right.. Mitsubishi one isn't a planetary one, but a normal open diff with 50/50 split

in both cases you have two output shafts, one within the other... rear one goes trough a 90deg and to the back, the other one feeds the "power/torque" to the front diff..

mitsubishi transfer box... incorporating 90deg drive to the rear, clutch basket (dissasembled in pic), hydraulic piston is still connected to the housing (on top of the housing..)
The round plate with a bearing on top connects the front diff to the clutch basket that is connected to one half of the plates. Other half is connected to the sprocket in the middle connected to the gear outputting drive to the rear axle...

in case of the mitsubishi, the split is 50/50, and centre diff is closed most of the time... interesting part happens when you get to the corner, and than you get a need for variable speeds between front and rear axles... the clutch pack then opens in a controlled fashion..

In lancias case, the front and rear axles are interconnected with a viscous drive... that means you can tolerate some slip all the time.. but not a lot, as when you get slip, the viscous drive tends to close...

90 deg drive to the rear.

the front diff sits in this housing that is bolted to the ring gear..

the purpose of the housing is to connect the rear drive to the clutch basket on the other side of the transfer box..

Edited by kikiturbo2, 27 December 2011 - 15:12.

### #10 24gerrard

24gerrard
• Member

• 2,008 posts
• Joined: March 11

Posted 27 December 2011 - 16:45

After 40 years of gearbox work, I can tell you that all 4x4 mechanical/plated or viscous systems for allocating torque to the wheels sap power.

The only sensible way is with an electric car and powertrain system.

### #11 kikiturbo2

kikiturbo2
• Member

• 724 posts
• Joined: December 04

Posted 27 December 2011 - 16:51

After 40 years of gearbox work, I can tell you that all 4x4 mechanical/plated or viscous systems for allocating torque to the wheels sap power.

The only sensible way is with an electric car and powertrain system.

and I will not be dragged into another electric vs petrol debate...

### #12 24gerrard

24gerrard
• Member

• 2,008 posts
• Joined: March 11

Posted 27 December 2011 - 18:17

and I will not be dragged into another electric vs petrol debate...

OK

### #13 kikiturbo2

kikiturbo2
• Member

• 724 posts
• Joined: December 04

Posted 27 December 2011 - 18:55

unless, of course, you can show me an electric with 4x4 and 2.0 turbo group N rally car pace (ok, lets say 1 sec per km slower than that), 100 mile range at that pace, heating, aircon and all that in sub zero temp.. ?

### #14 cheapracer

cheapracer
• Member

• 10,388 posts
• Joined: May 07

Posted 27 December 2011 - 19:07

After 40 years of gearbox work, I can tell you that all 4x4 mechanical/plated or viscous systems for allocating torque to the wheels sap power.

The only sensible way is with an electric car and powertrain system.

Yes Ferrari knew this that's why they came up with their simple and clever dual gearbox setup.

### #15 24gerrard

24gerrard
• Member

• 2,008 posts
• Joined: March 11

Posted 27 December 2011 - 21:17

Yes Ferrari knew this that's why they came up with their simple and clever dual gearbox setup.

Hahahaha
I am not going to discus mobile powerstations like the Ferrari block of flats thank you cheapy.
The technology is way to old.

### #16 24gerrard

24gerrard
• Member

• 2,008 posts
• Joined: March 11

Posted 27 December 2011 - 21:18

unless, of course, you can show me an electric with 4x4 and 2.0 turbo group N rally car pace (ok, lets say 1 sec per km slower than that), 100 mile range at that pace, heating, aircon and all that in sub zero temp.. ?

Simple job, give me a budget and I will design one for you.