Re: Locking differential

[Frank Deutschmann <fhd@interport.net>]

Jerry FIELDS sez:
 [Lots of good and true stuff deleted]
>      Wheels must have 
>      dissimilar surfaces for a locking differential to be of significant 
>      help.

Or, more technically, wheels must be in a situation where they are each capable
of generating different accelerations in the forward direction.

I just want to note that sometimes a locked diff on dry surfaces does help --
Porsche used Ti spool "diffs" in the 930 and 935 racing cars for many, many
years, in all kinds of weather -- particularly dry -- with great success.

This can be explained by the following:

1) There are three approaches available to drive two wheels from one
drive shaft: open diff, fully locked diff (spool), or locking diff.

2) An open diff provides torque to the wheels in proportion to the rate
at which they are turning, with the faster wheel getting more torque.

3) As Jerry pointed out, a locked diff provides the torque to the wheel
which can handle it.

4) A locking diff uses a mechanical mechanism to switch from open to locked,
OR uses a "torque sensing" arangement to achieve the benefits of a locked
diff when a potential acceleration difference exists.

A locking diff seems the best approach, then -- especially the Torsen style,
which "locks" so smoothly you can't even feel it.  HOWEVER, a locking diff
must switch a tremendous amount of power, and that switching action causes
wear and heat.  (Actually, the heat situation is not so bad: an open diff
has almost equavallent heat problems.)

So why the spool?  Simply, the need for a locking function, with very tough
wear and reliability constraints.  With the rear weight bias of the 911,
particularly on accel out of a corner, a great accel potential difference 
exists between the two rear wheels.  With an open diff, max power to the
ground is limited by the inside rear wheel slip -- hence the motivation for
a locking diff.

Will a locked diff run hotter than an open diff?  That depends.

We have to look at where the heat comes from; if the diff is always open,
heat sources are 1) the pinion to ring (driveshaft to diff carrier), 2) the
internal diff pinions.

If the diff is locking or Torsen, the heat sources are: 1) the pinion
to ring (driveshaft to diff carrier), 2) the internal diff pinions.

If the diff is fully locked (welded, or a spool), the heat sources
are: 1) 1) the pinion to ring (driveshaft to diff carrier).

So a fully locked diff should run cooler than an open or
locking/Torsen diff, right?  Not so fast, buckaroo!  First, if the
locked diff in consideration is a driver-lockable diff, such as is
found in Adui Quattros (ob-Quattro ref!), then we have some additional
heat generated by the shift fork which manipulates the locking dog
drive.  Second, if locking the diff improves performance (gets more
power to the ground), we have to consider that more power is flowing
through the driveshaft-diff coupling (the ring and pinion), and there
will be more heat generated at that interface.  Typically, this will
cause the diff to run much hotter.

So will a locked diff improve performance?  Again, that depends.  But
its not out of the question, even on a dry surface.  (Of course, it
won't improve quarter mile times in the dry, though!)

-frank
-- 
fhd@interport.net  | [M]athematics is not the study of intangible Platonic
1 212 559 5534     | worlds, but of tangible formal systems which have arisen
1 917 992 2248     | from real human activities.
1 718 746 7061     | 	--  Saunders MacLane