RE: What's the diff? torsen tech...

[David Eaton <david.eaton@computerland.co.nz>]

re: driving styles.

i've had another think about this issue.  this is not a manhood thing,
but i've certainly overcooked my ur-quattro and learned the hard way
about how to treat it.  i imagine that phil is the same.  we both
haven't been bitten.  not trying hard enough?  fair enough, but i don't
think so.  one difference that occurs to me is the car.  my and phils
ur-quattro's are quite different to the generation 1 cars (for example,
i notice that the wishbones (front and rear) are different part numbers
on my 20v than on the generation 1 (and 2) cars.  does this mean that
the geometry changed with the cars?  different bushes with altered
compliance?  what changed (my front wishbones are cast on the 'rr',
while the rears are still pressed (like the gen 1 and 2 cars) for
example.  does the 'mb' and 'rr' run a lsd rear.  phil does, and i'm
waiting to check my diff to find out...

also no-one in the us has yet had experience of a torsen ur-quattro to
directly compare.  hey you know where i live ;-)

2nd data point is that most people accept that the incidence of
understeer went from high: 'wr', to med:'mb', to low:'rr' (reference any
number of independent road tests).  this is also my experience with the
'rr' versus the 'mb'.  this will affect the handling cocktail.

3rd data point is the use of the circuit versus the road.  my 'rr' has
had little circuit time (thank you walter) and didn't 'bite', but scott
has spent much more circuit time, with more highly 'tweaked' cars than i
(or phil) have.

4th data point is that there is a lack of understanding about the
details of the torsen operation (relative slip to distribute torque, the
speed of the diff to go from one extreme to another etc), which is
clouding the issue.

comments?

re: active technology

scott, i think you are mis-understanding the function of "active" diffs.
 in my understanding that the use of "active" diffs varies fundamentally
according to the application.

it is quite achievable and has been done, to program an active diff to
"learn" a circuit and feed off the engine map to provide a changing
level of assitance depending upon the position on the circuit and the
type of corner.  in these applications the driver has a "reset" function
which allows the active to "re-program" itself to a particular circuit
point.  most of this technology was banned in formula 1 in 1993, but
some has made it's way back to the mechanical diffs (by definition
according to the f1 rules) and the engine maps (essentially
torque-programming maps) which the cars ran last year.  this has also
been banned this year.  the fia class 1 cars were using this technology
in conjunction with their (mostly) awd transmissions until the class
disappeared in '96...

a rally car is fundamentally different.  there is no attempt to "learn"
a specal stage because of the length and complexity of this and the fact
that the cars are denied service between most stages according to the
rules.  [while possible the diff would need to be reprogrammed to each
stage and this would require service intervention].  so state-of-the-art
in the game is the ability of the diff to call on various maps depending
upon the input parameters (speed, slip angles, lateral acceleration,
braking effort etc).  the diff *will* change torque setting through a
corner, in line with the changes in these parameters.  most of the top
wrc runners (mitsi, subaru, toyota and ford has active centres and
rears, and are starting to play with active fronts)..

as stated, the latest race-car engineering has a reasonable coverage of
the state-ot-the-art in last years wrc cars...

hth,
dave
'95 rs2
'90 ur-q

>-----Original Message-----
>
>Date: Wed, 25 Feb 1998 08:23:33 EST
>From: QSHIPQ@aol.com
>Subject: What's the diff?
>
>>It's not sufficient to say that Audi don't use a TORSEN centre differential 
>>when racing.  Sure - they don't - but they don't use any other kind, either.
>>If I were to find a viscous coupled differential in a motorsport Audi, I
>might 
>>think they have a case to answer.
>
>Not exactly.  I think if you take a look at the audi racing efforts thru the
>80's you will find reports that show that audi used either the 50/50 split
>locked Gen I in the early efforts.  Then in the later '80's fixed f/r biased
>locked between 55/45 and 70/30 based on venue.  Only very recently your no
>center diff racing came about.  Audi also used active, which is similar to
>the
>f/r locked with a small adj for each corner.  Dave L indicated that some more
>detail of active diffs can be found in the latest Racecar Engineering.  Volvo
>was collecting data per track per conditions for programming active for a
>specific track, then feeding into the computer program of that track.  I
>hardly think Audi wasn't doing the same thing.  You will find the same
>"deviant" intent to program behavior in F1 as well in regards to active
>diffs.
>This follows exactly what happened with ABS in Indy cars.  Start programming
>tracks, officials don't look kindly.
>
>