RE: Torsen 201

[QSHIPQ@aol.com]

Dave E writes:
>well scott, as i understand it, you're saying that:-
>1) no locked-diff quattro can oversteer.  this is certainly not my
>experience.  very interested that it is your claim.  i agree that understeer
>is more likely.  just as it is with the torsen.  by design with both
>technologies.  as i've explained.2) that torque doesn't follow weight shift.
>really, we are still at the basics here scott.  you are failing to
>understand/comprehend the following points:
>1) the torsen operates as a locked diff does only *until* the bias ratio of
>torque difference between the output shafts.  there is no magic here scott.
>until the torque bias ratio is reached, the output shafts of the torsen are
>effectively locked together.  just like the locked diff.  the effect in
>torque distribution dynamics is the same as is directly proportional to
>weight distribution and wheelbase and roll characteristics.

OK, so dave, if the above is true, then in 885140, then please explain the 
statement "however, the front wheels follow a wider radius than the rear 
wheels on a circular course, so that .2% of forced slip occurs between the 
two axles, which reduces the slip under traction at the front wheels, and 
increases the slip at the rear wheels.  This results in the tractive forces 
being redistributed towards the rear wheels, so that the tractive force 
distribution is 38/62."  Where is the weight distribution variable here?  How 
does weight distribution change more rear with a tighter radius turn giving 
75r/25f?  With the same acceleration rate, wouldn't a higher given side load 
actually reduce the rear weight distribution during acceleration?

>2) the only mechanism which the torsen has to proportion torque is the
>*same* as the mechanism the locker uses.  it's called front/rear slip. based
>only on this factor both the locked cantre diff and the torsen proportion
>torque to the rear in a turn.  the only difference is that the torsen starts
>the turn with a 50:50 torque ratio while the locker has a ratio which
>mirrors the weight distribution of the vehicle at that point.  

See above.  I think you might be missing something while turning.  The above 
conclusion would mean the locker should have quicker turn-in, not your claim

>fwiw, the issue of torque shift in a locked centre quattro is well covered
>in jeff daniels book "car suspension at work" in the chapter on 4wd.  also
>j.s. freeman has written a paper "awd powertrain models for real-time
>simulation" which empirically models just this behaviour on a humvee with an
>open and a locked centre differential.  i quote:
>"as expected the open differential case model exhibits equal torque
>proportioning between the front and rear axles.  while the locked
>differential model displays a greater proportion of torque being transmitted
>to the front axle at low to moderate tractive loads, while at a higher
>tractive loading the torque proportioning becomes greater at the rear axle.
>these results are similar to those of koga ("types of 4wd vehicles and their
>basic characteristics", 1988), however that study showed much more dramatic
>changes in torque proportioning of the locked centre differential case.  the
>change in front/rear torque proportioning is due to changes in wheel loading
>... without the effects of weight transfer, the tractive torque
>proportioning follows the static wheel load proportions.  in koga's study,
>the static front wheel loading was proportionally much higher on the front
>wheels, than it is for the humvee model in this study."
>until you understand these fundamentals scott, further discussion is simply
>a waste of time.

I see a basic straight line traction argument and experiments, with which I 
have no disagreement.  Does the above give a matrix for turning, or are those 
your assumptions/extensions?  Because what that wouldn't explain, is the 
885140 turning radius differences accounting for torsen appropriating torque 
rear.  If we hold acceleration a constant, but decrease radius, how did the 
torque reach the rear Bias Ratio in 885140?

Please help me dave, your presentation is now conflicting the 885140 paper.  
I still request we stick to torsens, locker arguments are adding confusion 
right now.  

Scott Justusson