[torsen] RE: Haldex differential

[QSHIPQ at aol.com]

--
[ Picked text/plain from multipart/alternative ]
In a message dated 11/20/02 3:55:55 AM Central Standard Time,
Dave.Eaton at clear.net.nz writes:


>errr no.  100% torque is 100% torque.  when the haldex clutch is locked, the
>rear can and will support up to 100% of the driveline torque.  i have
>absolutely no idea how you come by your limit of 45%.  it is certainly not
>supported by any data from haldex.

Capacity vs torque shift. Torque Shift is expressed in terms of Torque Bias
Ratio.   If you look at SAE paper, it's formula for torque bias ratio is:

T1/T2 = Bias Ratio.  Solve for audi locked diff
T1/T2 = Bias ratio.  Solve for audi haldex VC

How do we solve these two to demonstrate they are equal, or are they?  Aren't
we then really looking at capacity, not Bias Ratio?  Specifically, isn't a
TBR reflecting the ability to SHIFT torque in a given device, not support
torque within it.  What happens then, it we put the Haldex releif valve at
Trg of 150hp out of 200?  I'm liking this Haldex more and more.

>you might also want to figure that the haldex lsc can only support 2,000nm
>of torque prior to overload - somewhat short of 50% of that supported by the
>torsen.  btw, it is not a "hadlex vc" either, it's got no viscous component
>to it - i assume that is a mistake?

Dave, isn't it possible to increase that torque support by adding more
clutches?  Haldex claims so.  Torsen does too, you can add more helix gears.
Demand of current applications?  1400lb/ft would support the heresay that the
current haldex G1 can be run in full time mode on a 225hp TT (and just about
any other vehicle for the time being)?  Could it be with the relief valve
allowing slip (haldex) vs. shock load (torsen) dictates that a lower peak is
necessary?   The Haldex IS a VC by definition.  Wet multiplate clutches
however activated are "viscous couplings".  A magnetic lockup wet multiplate
clutch (ala scoobie) is a VC as well.

>anyway, wrt the 45%, put up the data, or lets drop it because the
>signal/noise ratio is getting pretty high here.

Solve for T1/T2.

>btw, the statement on the 3 different audi torsens i posted yesterday was
>directly from the zexel vp of engineering 3 years ago via email.

AND it doesn't support either the family album OR the Zexel published
applications.  Audis mysteries continue.  The most blatent issue appears to
be with the "revised bias ratio" on the S4tt (audi claims), when the family
album shows S4tt/RS4/Urs4/6 all carry the same torsen part number (5 OR
6speed O1E).  Stasis Engineering offers a 500USD shim service for the S4tt to
"put it back to the RS4" TBR.  How can that be so?!?!

Thanks for the posts Dave.

SJ