Here is a pictogram of the thrust assembly.
The problem crops up when the front pulley bolt is removed in order to
remove the pulley.
When the pressure is lost on this stack of parts the oil film will in
rare cases, pull the spacer forward
and that will allow the rear thrust bearing to fall down onto the crank
nose. Then when the bolt is torqued
on assembly, the spacer sleeve crushes the bearing.
A simple way to avoid this is if the engine is out of the chassis or
airframe, sit the engine gently on its flywheel. This will trap the rear bearing
against the rear face of the thrust plate, so no problem developes.
So-long as the engine rests on the flywheel the whole time.
If the engine must remain in the chassis or airframe during this procedure,
attach several bungee straps to the flywheel/flex plate, and pull the flywheel
flex plate forward firmly and maintain this tension during the entire procedure.
It is also possible for the front bearing to fall off of the sleeve, or for
the hard washer on the rear face of the counter weight to fall off as well. The
difference here is that this problem can be repaired by removing the pan and
inspecting the stack.
A dental pick with a 90 degree end works for manipulating the pieces from
below.
But suppose all has failed and the bearing is off the sleeve. The front
cover is off but the thrust plate is still installed. Slide the sleeve out of
the way. Take a big goober of axle grease and press it into the opening on
one side of the crank. Use a thin wire to center up the rear bearing around the
crank. Slide the sleeve back into place. Slip the front bearing on and note the
sleeve sticks through the front bearing no more than .003"
or, nil to the naked eye. If there is no protrusion then the rear bearing
is properly located. Now tension the bungees on the flywheel to trap the
bearing. Check the protrusion again, and if all is well, reassemble the stack,
case and pulleys.
With the assembly torqued, end play should be .0025" to .0035". Move the
shaft by hand to check end play. Not with a giant screw driver. The thrust plate
is flexible and will seem to yield an end play measurement even when none
exists.
Lynn E. Hanover