[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Diff summary (warning some T*rsen content)

Dave E. writes:
>george, i wouldn't worry about making mistakes on the list.  we all do.  i
>certainly find it amusing to see you being lectured about "open diffs
>splitting torque 50:50" by someone who, not 3 weeks ago, was vehmently of
>the opposite opinion :-)  so, clearly, we're all learning here....

George and I had a quick and fine offline discussion regarding his story 
problems.  Your interchanging of terms in your theories confused your 
intentions.  Had George posted this 3 weeks ago, you might have found we 
actually agreed on power/torque at the ground.  You are welcome to correct my 
math as posted to George, if you can't, maybe we weren't as far apart as you 
think.  If we all get A's in the math...

>a locked diff will transmit 100% torque to either output shaft, as required
>by tractive conditions (wheel lift etc).  however, it will not allow speed
>difference between the shafts.  it functions as a torque distribution,
>speed-difference-limiting device.  it's inability to allow speed
>differentiation between output shafts means that, in the classic audi
>arrangement, major understeer reigns.  this is also due to the fact that the
>quattro with locked centre will take more torque to the front during
>*normal* operation due to the chassis kinematics (i.e. being front heavy).

No, this is not correct as stated.  Weight distribution has little to do with 
torque distribution of a locked center.  A locked center doesn't know the 
difference.  Think about what you are saying.  Statically a quattro is front 
heavy.  If the above is true, isn't the 'accelerating' weight vs torque 
distribution closer to 50/50?   I'm not buying into this "summary" as stated 
Dave.  What then, happens under locked center braking?  

>this means that the locked centre quattro will behave more like a fwd car,
>will understeer more under power, and exhibit a greater tendency to snap
>oversteer on throttle lift during cornering, than the open centre quattro.

Whoa up there Dave.  Snap oversteer on throttle lift?   No, the locked center 
will understeer predictably under power, and without power.  I think you 
might be confusing your diffs, LTO is a Torsen property, you posted the spin 
paper, it's in the archives.  Getting a locked center to LTO is not a 
tendency, in fact it's really tough to do, ask anyone that's been to 
Steamboat.  Many a rally driver wishes that this were a property of the 
locked center.  A quick read of Buffum's book doesn't speak of LTO, it speaks 
of U an more U, right off into the woods when too hot.

>interestingly, another side effect of this 100% torque distribution effect
>is that you must now size your output shafts to accept 50% of the input
>torque, whereas with an open diff, 25% is all that is required.  this has
>obvious implications in weight.

You are reaching way out here Dave, and you are confusing terminology again.  
 This conclusion you found is only a race car argument, and you didn't 
translate it correctly, IMO.   Take a look at the size of a 200HP Eagle Talon 
pencil thin output shafts.  When it's VC center locks, each output shaft 
accepts 50% of the torque (extending YOUR argument, but maybe you want to 
reread what you printed).  This has little to do with design on any street 
audi or other awd.   Audi did nothing regarding driveshaft design in terms of 
size and/or weight thru all generations of the awd concept.  Why not?  
Looking at the locker rear audis vs the open front on the same car, why isn't 
the locking rear output shaft not bigger than the non locking front?  Why 
isn't the turbo driveshaft not bigger than the N/A driveshaft?  Not buying 
this conclusion at all, and you really should reread what you printed.

>a torsen otoh, will re-distribute torque based on the traction requirements
>of the output shafts.  

No, the torsen (oh boy) will re-distribute torque based on the torque 
differences across the primary output shafts.  That is to say, it's Bias 
Ratio is not necessarily a traction requirement, it can be a slip angle one.  
Totally crank the wheel in first gear do the circle test.  Sounds like the 
car is falling apart?  No, it's the torsen exceeding it's Bias Ratio due to 
slip angle differences front and rear.

>but the important fact is that (1) because of it's
>internal gear friction, it will *limit* the maximum torque distributed to a
>shaft to a % (the bias ratio) of the input torque (the internal friction is
>providing a torque "drain"), and (2) it allows output shaft speed

Er, it "can" allow some output shaft speed differences.  At full lock in a 
circle it won't, and with slip angle differences across the front and rear it 
won't necessarily.  One of the main reasons audi is the only one to use the 
torsen center.  

>the actual bias ratio varies depending upon the type of torsen
>design (e.g. whether the diff is used in the centre or on an axle).  the
>torsen starts to shift when it detects a torque imbalance between the
>shafts, *not* a speed imbalance.  

Somewhat correct.  By the time there is a speed imbalance, the torsen has 
already shifted it's maximum BR.  

>the amount of torque shifted will be
>limited to the output shaft with the *most* traction.  

Let's be clear here Dave, Bias Ratio theory is based on the output shaft with 
the least traction.  The above is unclear and not correct as stated, you are 
confusing primary and secondary axles, IMO.  The secondary axle with the 
"traction" is not a torsen diff necessarily.  When you jump to the above 
conclusion, you ignore slip angle differences.  A big factor in the Bias 
Ratio of any LSD center diff.

>so, the torsen is a
>torque distribution device, but allows different shaft output speeds. 

All differentials that aren't 100% locked allow different shaft output 
speeds.  How much is the only question.

>will shift torque up to the bias ratio (as will any limited 
>the vc (again there are a great many types with fundamentally differently
>characteristics of torque transfer) uses axle speed differences to initiate
>torque transfer and will continue to differentiate up to the bias ratio
>(again dependent upon design), or until heat causes "hump".
>there is a better paper of the torsen in awd setups (more descriptive than
>the chocholek paper on audidudi's web site) from autotech '95 which i
>circulated last year.  when i get some time, i'll get around to publishing
>some of the other stuff i have on the torsen (yeah right)... 

Jeff can put those papers on his website.   Might help all of us to 
understand some of your drawn conclusions on all diffs.  Given the above 
post, you are again confusing primary and secondary axles vs torque and 

My .02

Scott Justusson