Posi Replies

Greg Hermann bearbvd at sni.net
Thu Jun 3 20:45:32 GMT 1999


>All clutch type limited slip differentials work the same way.  A series of
>clutches are placed between the axle and the carrier so that when one axle
>turns relative to the other, the clutches must slip.  These clutches can be
>either multi disc or cup and cone.  The mulit-disc are rebuildable, the cone
>type is not but they work the same.  When the spider gears apply torque to
>the axles, they tend to spread apart, as do all gears.  The spreading force
>from the spider gears applies the clutches.  The higher the torque being
>applied, the greater the locking force.  There is also some sort of pre-load
>spring to achieve a breakaway torque.  The stiffer the preload, the harsher
>the engagement.  For road racing, handling can be tuned with various amounts
>of preload as well as number of discs, etc.  Too much anti-slip creates
>power understeer, not enough lets the inside tire spin.
>
>Locker types send all the power to the slowest turning axle by using one-way
>clutches.  This is OK for drags but too harsh for the street or road racing.
>
>Torsens are really cool.  They use the pressure angle of worm gears to apply
>the torque.  They apply torque evenly until one axle tries to overspeed the
>other, then torque is shifted to the slower axle.  If one axle brakes, the
>diff runs free.
>
>There are also viscous clutches that don't apply the clutches with springs,
>instead, a rheopectic fluid, like silly putty which increases viscosity as
>it shears, is in between the clutch plates.  When they begin to slip, the
>fluid gets stiff and the clutches lock up.

Two other REALLY neat types--

The gerotor--an oil pump is driven by speed differential between the two
axles. The pressure output from the pump applies the clutches, and either
the greater the speed difference, or the longer it exists, the greater the
force applied to the clutches.

The limited overrun type--was used in the Jensen FF transfer case center
differ. There are two separate gear trains from one output, each with a one
way sprag clutch. If slip exceeds a pre-determined % in either direction,
the one way clutch prevents any more slip. The Jensen TC also used a
planetary differ--to the torque split was unequal, front to rear, about 40%
front, 60% rear. AND, the lockup ratios were different too--5 or 6% rear
slip would divert torque to the front via one one way clutch, but it took
10 or 12% front slip before the other one way clutch/gear train diverted
torque to the rear output. EXPENSIVE, and heavy, but pretty damn cool deal!

Regards, Greg

>
>Gary Derian <gderian at oh.verio.com>





More information about the Diy_efi mailing list