[torsen] Re: [urq] RE: Quattro handling

QSHIPQ at aol.com QSHIPQ at aol.com
Sun Apr 14 09:51:15 EDT 2002

In a message dated 4/14/02 5:50:01 AM Central Daylight Time, 
Dave.Eaton at clear.net.nz writes:

>mmmmm, - lets just say that it's not my experience, nor my understanding.

>one other bit of hard data here is that torque transfer happens on an ur-q
>with both the locked centre, or with the torsen. 

Bad terminology.  Either the front axle or the rear axle in a locker *can* 
support 100% of the torque.   Again, I would point out that one really needs 
to assume the same cf.  The difference between the  locker and the torsen on 
the same cf is that the torsen tranfers torque which can *cause* breakaway of 
EITHER the front or the rear axle, and/or in an alternating progression.  The 
locker doesn't allow any speed differences between the axles, so the fronts 
break away always before the rears.

> the torque transfer in
>either case is *entirely* dictated by traction under the wheels.  

Not exactly.  The torsen will allocate torque based on turning radius only, 
which isn't a "traction" argument.  Specifically, it's an "attempted" speed 
differential argument, which the torsen reacts to as a traction argument.

>in fact
>the torsen is technically locked while the differential is operating within
>the torque bias it was designed for (for audi's 25:75%).  when it is asked
>to operate outside of this (usually in low-traction conditions), it unlocks
>(while maintaining the torque split).

IT CAN.  But I think you have to remember, that on snow or ice (I've watched 
it in the wet, and even in the dry with enough power), the traction is so low 
that it begins to oscillate, front wheels spin, rear wheels spin as it tries 
to bring itself back to the 50/50 torque split.  IN theory what you describe 
above *can* happen, but it doesn't at the limit of adhesion.    What you end 
up seeing is fronts spinning, then the rears, BTST many times.  In the 
locker, both spin at the same rate, cuz they have to.  And many a race driver 
finds this predictability to be advantageous,  rally or race or street or 

>here are only 2 differences between the locked centre and the torsen:
>1) in the locker up to 100% of torque can get shunted from front to back,
>while in the torsen, this is limited to 50%.

This argument doesn't hold true UNLESS you have a difference in cf, 
specifically, fronts on one cf, rears on another.  Then the above is true, 
but in reality that lasts for the length of the wheelbase.  OR look at it in 
the real world, track, street, rally, or steamboat.  The above just isn't the 
conditions one finds real world driving.  In which case, AT the limit of 
adhesion, ON the same cf surface, the locker will always break away the front 
wheels first (U) because at the limit of adhesion with equal axle torque 
load, the wheels with the most slip (the ones steering) will overload first.  

IN the torsen, it "can" be the opposite.  Turning radius causes the rear tor
que allocation, causing the rears to overload first (back end begins to step 
out).  As soon as they do, the torsen will shift 56% of engine torque to the 
front axle, which will result in the front end to step out.  As soon as that 
wheel speed exceeds the rears, the torque shift will go to the back again. 
OR, you can attempt a corrective action:  steer more - overload the fronts,  
steer less - U, decrease throttle - U, increase throttle - U.

>2) the torsen reacts differently to the locker in wheel lift conditions,
>much like an open differential.

Exactly true

>torque transfer does not happen on a ur-q with an unlocked centre
>differential.  but lots of understeer does.

A lot of understeer happens in ALL urq's.  In a torsen "some" oversteer can 
happen, but as you reach the limit of adhesion it *can* become a fore aft 
torque allocated overload, which results in a lot more U than either the 
locker or the open.

>as this thread is going to fast deteriorate methinks into a torsen re-hash,
>i'd be happy for those wanting further reading to correspond off-line.

I'd be happy to just move it to the proper list.  The video footage I have, 
I'm happy to forward to anyone.  It clearly shows the phenom of torsens 
oscillations causing a way outside line vs the locker.


-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.audifans.com/pipermail/torsen/attachments/20020414/38d8b49e/attachment.htm

More information about the Torsen mailing list