RE: Torsen stuff

[QSHIPQ@aol.com]

Dave E writes:
>scott, a fundamental understanding of the torsen is required before you can
>understand how it works in a vehicle.  it's finally starting to become clear
>where your misunderstandings are:

Ok, so when I walk thru this, you answer my overdue 885140 quiz?

>1) bias ratio.  you completely misunderstand this. the bias ratio is not a
>variable, it's a given for a particular differential design (set of gears).
>it is the maximum ratio of torque bias the diff will support between shafts.
>hence your statement "bias ratio [sic] rear varies with every turn" is plain
>wrong (and nonsensical).  bias ratio is a *constant*, not a variable as you
>state.  in a torsen, bias ratio is determined by the respective worm gear
">interfaces" which also govern the degree and rapidity of locking behaviour.
>much like the setup of a vc where different silicones change the speed of
>locking.  anyway, i digress...

Please, one concept at a time.  Problem:  can we take a turn *at* a Bias 
Ratio other than 75r/25f without *any* traction intervention.  You say BR is 
a constant, technically correct for a traction input.  However, it's not a 
constant with a slip angle input.  In fact, we could do a test per your 
paper, that would indicate indicate that BR is *not* a constant.  In fact, we 
could do a test on slip angles that would never get BR to the front, and 
never get Maximum rear.  In an 80q (per your ref) that would be at a given 
accel, high cf, >/= 45m turn.  Correct?

Given this concept confuses some, I'm happy to use BR as a constant, and 
further ref turning with torque bias.  Does that help?

>2) you state that "bias ratio [sic] rear varies with every turn" with the
>torsen.  incorrect as stated.  replacing "bias ratio" with "torque", then
>this is correct.  up to the bias ratio of torque split.

I'm happy to use either, if it helps this discussion.  Hardly a major point.  
You got it, now let's go to chassis dynamics while turning at a torque bias 
(we could say ratio, but that might confuse folks more).

>3) the torsen distributes torque on the *basis* of front/rear slip
>differences.  until the bias ratio limits torque proportioning.  ditto the
>locker, except that the locker will continue to proportion torque to 100%
>either way (i.e. no bias ratio).  you should understand this point. 

Ok, see questions below.

 >each
>differential uses the *same* inputs.  in other words, the *only* thing which
>causes the torsen (and the locker) to proportion torque is the *differences*
>between front and rear slip.  ok?
>these points are really torsen 101 scott.  btw, if you actually had the audi
>sae torsen paper you reference so often, you see this in the tests.
>once you have agreed to these points, we can move on to torsen behaviour at
>the bias ratio.

Whoa up Dave.  The torsen uses *two* inputs to proportion torque.  Relative 
slip angles and/or traction.  The locker doesn't use *any* inputs to 
proportion torque, cuz it has no bias ratio.  A locker *can* support 100% 
torque at either axle.  That's only a traction argument.  Explain how that 
can be a slip angle argument?

Adding to my 885140 quiz:  What happens to the 80q that got front wheel spin 
on the .4 cf?  Did 100% of the torque go to the rear axle?  Maybe a more 
specific question would be, is 100% of the engine torque at the ground?  What 
would you expect the chassis handling characteristic be with that scenario?  
It conflicts your O moment statement made earlier doesn't it?  If front 
wheels are spining during a turn, and Trg reduces, what chassis handling do 
you expect.  Care to correct your earlier post?

I promise to use "torque bias" from now on, though Dave.  About that 885140 
quiz?

Scott Justusson