[torsen] RE: differential/torque sensing

Andres Kroonmaa andre at online.ee
Thu Mar 27 12:11:01 EST 2003


> *  Read C498/30/14, specifically section 3.1

 for that I specifically wrote an explanation. Please reread it.
 C498 assumes that you understand that without further details.

> *  Agree or disagree with the statement:  
> All tire slip (longitudinal or lateral) forces axle slip.
 If you want to be precise, then no, disagree.
 4-tyre lockup due to braking for eg. 4wheel drifting may and may not
 have interaxle slip. lateral slide doesn't necessarily mean axle slip.

> *  Try to figure out in Chocholek's paper how exactly you would change the 
> TBR in the drive and coast modes, since clearly that can be done.

 Look at thrust washers. There are 3 of them: two on outside of sidegears
 and one between them. On forward torque, both sidegears are pushed against
 same side of diff housing. In reverse mode, both are pushed to other side.
 difference is that outside washers are different. By picking them with
 different coeff of friction, you get different TBR in drive vs coast mode.

 Then, teeth of gears. They have finite contact surface, depending on
 application angle of lathe tool. That angle can be different on opposite
 sides of teeth, meaning different friction for opposing directions.

 Opposite element gear faces are pushed against housing. Same mu tricks.

> *  Your dissertation on tire slip confuses me.  First and foremost, let's get 
> agreement that ANY force that causes the 2 axles to slip is what the torsen 
> reacts to.  That is tire slip based on either turning radius (lateral slip) 
> or actual traction (longitudinal slip - accel or decel).  The torsen doesn't 
> care which it is, it only reacts to driveshaft speed differences caused by 
> either phenom.

 Oh, You are full of that speed difference thing. Torsen doesn't know if
 it is tyre slip or turn radius, they cause same kind of *positive* speed
 difference. In coasting mode, there is external force causing axles to
 have *negative* speed difference. And faster negative axle is equivalent
 to being slower to torsen. You don't get it because you are locked on
 that speed sensing, and from its perspective there is no difference.
 Its not speed sensing, its property of speed sensitivity is flaw, not
 primary function. It is sensitive to axle torques.

> *  Reverse torque is reverse torque.  How does the torsen know the difference 
> between engine braking and engine applied torque in reverse?  I don't think 
> it's that smart, so I claim they are exactly the same thing.  You claim they 
> are similar.  That implies to me there are differences between them.  What 
> are they?

 For christ sake, please reread it with thought. Don't just glimpse over
 it nodding with confident btdt. I don't say its smart, I say that the
 conditions are not exactly same.

 Lets try analogy. You have bike. You turn it upside down and stabilise
 it. Wheels are free in air. You take your hand and give one wheel a kick,
 it starts rotating. You give it more. At some point you won't be able to
 give it more torque. ok. Now go to other side, and give it a kick *opposite*
 of its rotation. ouch. hopefully you won't hurt your hand.

 Another. Start rotating bike wheel with constant torque. Let your friend
 resist that rotation. Now ask your friend to suddenly accelerate the wheel
 in direction of rotation. What you feel is loss of torque. Now ask friend
 to accelerate the wheel in opposite direction with helluva lot of force.
 What are you going to feel? Given that other wheel have been at rest for
 all of the experiment, your other wheel have been in 2 different torque

 Third one. Suppose your car is accelerating and rear axle has 8rpm and
 front has 10rpm. Torque in rear. Now, suppose god moves whole pavement
 under the car with quite some force forward, so that car inertia can't
 allow car movement at same pace. Now ground moves under the car forward
 so that axles are forced to turn backwards. Its a shift -20rpm: relative
 rpm difference is same +2rpm, but actual values are -12/-10. Torsen is
 applying torque in same direction, trying to oppose that external force.
 Now note that -12 is faster than -10rpm. But torsen still applies more
 torque to rear, because for it its still +2rpm difference between rear
 and front. This will continue until wheels pass through 0 point, and
 start rotating forward again. Then you'll see 0/2, 4/6, etc and you
 can say that torsen sends torque to slower axle. But it hasn't changed
 its behaviour the slighest. In coasting mode, torsen has this scenario,
 it sees negative axle speeds. In reverse gear, car starts off from stop,
 the 0-point, and there are no negative speeds to consider.

 If you are stuck with speed sensing, this confuses you. Torsen behaves
 that way because it is sensing opposing torque, not speed. Speed has
 secondary implications to torsen, but reasons are in torque mechanics.

> *  Change the rotation of the input at the basket, then look at the 
> frictional force arrows applied to the helix gears.  Any change?

 Yes, arrows change direction. Those applied at teeth will also change
 side of the teeth and opposite surfaces of teeth will contact.
 Completely mirror image. Same behaviour.

> The rest of your post appears unclear to me, I'm not here to insult, I'm here 
> to learn.  You claim 4hours of study.  Damn dude, several of us have put that 
> time in exponentially over the years.  

 Man, I didn't mention amount of study I've put into it. I said that I
 wasted 4 hours to write that single email, trying to put words right.

> I'd appreciate a baseline here.  IMO, we are beyond the operation, and into 
> how to develop some sort of matrix as to it's effects on chassis dynamics (my 
> personal interest) of U and O at the limit of adhesion.  Given a university 
> special, how does cof, cf ratio, turning radius, wheelbase, cog, and Trg 
> affect the actual chassis it's installed in.  I say that's a huge project 
> given the device.  Once we get to the basket and the shafts, what the device 
> does is pretty clear IMO.

 Clear - really? We haven't reached agreement here. As to chassis dynamics,
 I would like to start completely other thread on other matter, where your
 input and real racing experience would be really appreciated. But I feel
 awkward until we find agreement here.

> Thanks for your posts.

 thanks for encouraging. I was already feeling unwelcome.

 Andres Kroonmaa <andre at online.ee>
 CTO, Microlink Data AS
 Tel: 6501 731, Fax: 6501 725
 Pärnu mnt. 158, Tallinn
 11317 Estonia

More information about the Torsen mailing list