Early 901 LSD

[chris_seven]

Plate type LSD for Early 901 with Lobro Outputs.

Several ramp and plate configurations.

Available in 2 weeks

[Grady Clay]

Chris,

Will this take the standard (non-simplified) transmission axle flange with the aluminum plug or the ‘simplified’ version? Either?

What length ring-gear bolts?

What clutch configuration? How many discs? Size?

Can’t wait to see one in steel.
Looks great.

Best,
Grady

[chris_seven]

Grady,

I will publish a more detailed drawing showing number of plates and plate sizes but in general they will have 6 plates per side.

I have been thinking about the number of plates needed for some time and in absolute terms it should this feature affect the behaviour of the LSD?

If the axial force being applied to the plates is fixed, then the ability of the plates to slide should not be dependent on the number of plates.

I think the early ZF with 2 plates per side changed characteristics with 4 plates per side because the extra plates increased the preload.

Any thoughts welcome.

With regard to plates:

The prototype uses steel plates but the basic production unit will use moly coated friction plates.

Upgraded plates with higher levels of friction will be available.

The fisrt upgrade will be to a sintered plate material and should allow reduced preload whilst maintaining a high level of torque bias. I also believe they will have smoother operation and be less 'grabby'

I also hope to have 'Carbon'plates available in a short time. These plates will generate even more fricton than the sintered material and should eliminate all traces of low speed judder and give very smooth operation.

In the longer term we also hope to able to to produce a light weight casing with a 'wavy' flange and some extra machining.

Bolts will be standard length.

The initial design has been made to suit the 1968 flange 901.332.209.12 which is from the 68MY

I imagine we could make them to suit the 901.332.209.10 flange but I am not sure of demand Which is 67MY

It is our medium term plan to be re-manufacturing the output flange in something robust such as S155 or 300M.

We have 3 other diffs drawn and are planning production very soon.

Early 915 with coarse spline - (also fits the 911 box I believe)

Late 915 with fine spline

Later 915 with fine spline and fitting for mag sensor unit

We also plan more driveline parts later in the year.

[Flieger]

I was under the impression that having more plates increases the friction area much like a bigger brake pad and should therefore slow down wear on each individual plate and also distribute the heat generated throughout a greater mass, further reducing wear rate. This is assuming the plate thickness and material are held constant, with the only variable being number of plates. I thought, too, that more plates make for a more progressive break away from locked to differential wheel speeds.

Friction force = mu * normal force, and the pressure (F/A) will affect the mu. Each plate has the same preload on it when they are stacked in series, so each has the same pressure on the friction surface. So mu should be the same for any number of plates in series.

It seems that if the plates were in parallel somehow, they would wear slower due to lower pressure and more mass to dissipate the heat. They would also have higher breakaway torque even before factoring different polar moments of area due to higher mu because of the lower pressure.

With plates in series, the plate interface with the lowest static friction force (breakaway torque) will slip first, so there would only ever be 1 plate interface slipping. It would be like analyzing buckling modes- only the lowest order is necessary because that one is assumed first, unlike when examining higher order resonances like in crankshafts, etc.

[chris_seven]

Max,

It is this discussion about locking that has made me think more about the situation.

When both wheels travel at the same speed can we say the diff is locked? It doesn't operate as there is no forcing function so it behaves in precisely the same manner as an open diff.

When it needs to slip I agree that the plate with the least friction will slip first and I then can't see how adding plates with the same loading helps this model.

I feel sure that I have missed something blindingly obvious and in a minute will have to come to terms with making a basic error.

There are two situations that I think need to be modelled.

The first is when the diff has to allow differnetial wheel speed during cornering.

Clearly some slip needs to take place during this action. If tyre inflation is different side to side this would also cause the diff to operate all of the time.

2mm difference in diameter is about 5rpm at 120mph I think.

If one wheel loses traction the diff should now not slip it should transfer torque to the wheel with some grip but again only to the limit of its friction capability.

It would be great to have a transmission test rig with the ability to test an LSD and see exactly what was happening.

[savaden]

I looked at joint flange part numbers and I don't see the ones that you mentioned. Perhaps you could identify which ones are supported from this list.

-67 902 332 209 02 joint flange -Nadella-
'68 901 332 209 12 joint flange -Loebro-
'69, '70 901 332 209 15 joint flange -Loebro- for chilled cast and typ 905
'69, '70 901 332 209 21 joint flange -Loebro- for pressure casting
-68 904 332 209 01 joint flange for LSD
'69, '70 901 332 209 21 joint flange for LSD with pressure casting housing
-71 911 332 209 00 joint flange typ 915 and 925
'72- 915 332 209 01 joint flange typ 915 and 925
71- 915 332 208 21 joint flange for LSD typ 915
70- 914 332 209 01 joint flange for 914
70- 914 332 209 00 joint flange for 914 w LSD

SV

[chris_seven]

I looked at joint flange part numbers and I don't see the ones that you mentioned. Perhaps you could identify which ones are supported from this list.
SV

Excuse my crummy typing, I have corrected the numbers in my earlier post.

The 901 type diff has side gears to suit the following:

'68 901 332 209 12 joint flange -Loebro

The 915 diff will have the following:

915.332.209.01
923.332.209.02

I believe these are the basic coarse and fine spline parts.

I think cataloguing all of the detailed differences will be quite fun and trying to find interchangability will be quite tricky but we have to start somewhere.

[savaden]

OK, i'm waiting with bait on my breath...

SV

[chris_seven]

I have been thinking about the number of plates needed for some time and in absolute terms it should this feature affect the behaviour of the LSD?

If the axial force being applied to the plates is fixed, then the ability of the plates to slide should not be dependent on the number of plates.

I feel sure that I have missed something blindingly obvious and in a minute will have to come to terms with making a basic error.

I just knew that I had missed the obvious - have now worked it out and am suitably embarassed