[torsen] Re: [V8] differential/torque sensing

QSHIPQ at aol.com QSHIPQ at aol.com
Mon Mar 24 11:50:52 EST 2003

Comments inserted.

In a message dated 3/24/2003 2:39:01 PM Eastern Standard Time, andre at online.ee writes:
> But this quotation applies to testrun with 40m turnradius where non-blocked
> wheels will already have forced axle speed difference. Thus if you translate
> that into engine torque in reverse gear, then its case for fronts already
> slipping, thus also torque shift to rear.

Slip is slip, doesn't matter if you are turning or a straight line.  What this quote says specifically is that under engine overrun torque, the open diff will ALWAYS allocate torque (any torque) 50/50.  The torsen won't, in fact it allocates reverse torque to the faster spinning axle.  Read Chocholek 4.3 paragraph 2.  The "opportunity" to vary the coast torque in the audi university special wasn't ever practically realized.

> Look, there are many different ways to describe the device, I offered
> just another one, based on purely mechanical interactions of not so many
> parts. Hope you agree there is difference between how its behaviour is
> perceived and how it actually behaves?

Oh man...  Read the archives :)

> Already this alone would be easily
> cause for different interpretations. That sae paper focuses on perceived
> behaviour. In my view, my description doesn't confict with it anywhere.

Sure it does.  Your explanation doesn't explain how a torsen works better in one direction than the other.  Reverse torque causes the faster spinning axle to get more torque in a center torsen.  This is EXACTLY stated in the Zexel quote I gave you earlier, and IMO it makes perfect sense.  You explain things as "unlocking"  I say a torsen can never "unlock".  It can send reverse torque to the faster spinning axle at exactly the opposing TBR as a forward torque allocation.  

> Look at torsen picture. Can you find anything that can possibly behave
> differently in different direction of torque? Its completely symetrical.

All that explains is that the static state is 50/50.  When you apply forward torque to the device, the thrust loads cause the helix gears to allocate torque to the slower spinning axle.  When REVERSE torque is applied to the device, the thrust loads are on the other side of the helix gear, which causes the torque to be allocated to the FASTER spinning axle.  Andre, the reason the pix show rotation and forces with arrows, is that when the rotation is opposite, the forces change as well.

> It can't apply more torque to faster spinning wheel, that 
> would defy
> mechanics and energy conservation laws.

Sure it can, remember torque goes to a differential where it is allocated.  In an LSD it's allocated T1 and T2.  In a torsen travelling forward direction, more torque is allocated to the slower spinning axle.  IN a reverse direction, the toque is allocated to the faster spinning axle.  T1 + T2 = Tring gear (torsen input).  The components of T1 and T2 are the sum of traction + slip (acceleration) minus FD gearing.

The Chochelek article clearly shows that the forward and reverse (forward and coast) modes of the differential put the helix gears in exact opposite locations within the basket (read in detail section 4.3).   You CAN change the TBR from forward to coast mode, audi didn't and for good reason, since the higher the TBR the better the ABS component of the braking equation.  

No question in my mind that reverse torque causes the device to send more torque to the faster spinning axle in an audi application (25/75/25).  It's clearly stated as an advantage in "braking" applications.  It's unstated downside is that under reverse "power" applications (driving in reverse), it's doing exactly opposite what it does in forward drive mode.  Including overloading the faster spinning axle.



More information about the Torsen mailing list