Paul and Jeff, With the involute outlet scroll within the 2" billet and the
inlet centered on the back side of the plate, the pump flange will be pretty
much surface mounted. That will require a dished drive sheave to line up
with the mazda e-shaft sheave. this will put the drive forces between the
shaft bearings and close to the flange end bearing where bending loads will
be small due to the short lever arm and there will be no overhung loads from
the fan and fan clutch which won't be used. Not having the fan will
substantially reduce the bending and torque loads on the shaft as well as
the gyroscopic loads of the fan. The original installation of these pumps
had a drive sheeve slightly smaller than the crankshaft sheave so that a
2100RPM on the engine the pump would be about 2500RPM.We will be driving the
pump probably 3700RPM with an e-shaft speed of  6000RPM. At full flow this
would be a higher shaft torque, however the pump horsepower relates to
pressure and flow rate. We will have a smaller flow rate due to the
restriction of the coolant passages in the rotary block and a higher
pressure so the net shaft torque might be about the same as the original
installation. The pump speed will be dictated by fitting the drive sheave
around the pump and the distance between shaft centers which should not
exceed 2-1/2 times the diameter of the smaller sheave. I am leaning towards
a 9" sheave but will have to test the actual pump for flow and pressure at
some point.
      IMHO this housing should be more designed to provide superior water
flow rates through the block as an aid to installations with marginal
cooling problems rather than to avoid the replacement cost of the stock
water pump. Having the housing machined for some gearotor dry sump pumps
would be a plus. If the pumps were not installed the metal removed would
make the part lighter. Also oil routing might be looked at. If the only oil
that goes through the cooler as that which is vented out the relief valve
feeding the e-shaft, the hot uncooled oil is sent to the bearings and only a
small portion is cooled. Since a major part of engine cooling is from the
oil it should be routed through the cooler before the relief valve.This
housing  will likely look as spectacular as the PSRU/ motor mount plate
which alone would make it a commercial success but performance is a real
selling point also.

george grimes

Just a thought. Should we not finalize the engine control unit design prior to redesigning the front cover? That way, if the preferred method of engine control would benefit from front cover modifications, it could be incorporated before any metal is cut?
C. Smith