Main FI Selection Specifications
o  8000 rpm will need a minimum of 675 CFM
o  PD SCer is initial design choice due to low rpm torque needed
o  Just over 10 psi expected
o  3000 rpm will require 240 CFM
Design analysis
There are many Positive Displacement super chargers available today.  The most
common manufactures are:
The Eaton is a true "roots" type blower while the Whipple and AutoRotor are Screw
types.  The operational difference is that the roots type simply moves air from
intake to discharge and makes pressure very similar to a simple oil pump.  Screw
types do most of the air compression internally.  Both have advantages and
disadvantages (a detailed discussion can be found under technical section).

Like most folks, I looked to the current crop of aftermarket kits for a "off the shelf"
solution to my needs.  This searched boiled down to two choices, the Comptech and
the Gruppe M systems.  Out of the two I felt the Gruppe M had more potential as an
external intercooler could be custom designed and incorporated without disturbing
any of the main components.  An intercooler for the Comptech would need to be
integrated with there custom intake manifold and require relocation of the super
charger and associated intake manifold and drive belt issues.  A quick look at the
Eaton M90 blower specifications showed that the unit could NOT supply the necessary
volume of air to support my HP and torque goals.  

The main limitation on the flow fiqures is the blower rpm being limited to 12,000
rpm (550 CFM).  The plots below are for the M90S model.  This is a "ported" version
of the standard M90.  It allows the blower to make more flow by spinning it faster
without any major loss in thermal or volumetric efficiency.  The advertised limit on
the M90S is about 14,000 rpm and about 650+ CFM.  The Gruppe M uses the "S".

Still, the final number is still a bit shy of my target 675 CFM (about 46.5 lbs/minute)
at near std atmospheric conditions.  Several things to note on all these charts.  On
the flow charts, the flow at 10psi is LESS than 5psi.  Why? The way these blowers are
tested is pretty simple.  They apply a know rpm to the blower and restrict the output
until pressure (boost) is achieved.  At 5psi the actual flow through the blower is
measured along with the hp required to spin it and the air temperature. To test at
10psi the restriction is made even SMALLER to increase the pressure.  At higher
pressures the reversion or "leak back" through the blower is greater (makes sense)
and therefore the NET flow is slightly less than at 5psi.   Note however, that it takes
more power to pump roughly the same volume of air through a restriction 1/2 the
size (5psi vs 10psi).  The power is double at low speeds and increases at a
decreasing rate all the way to 14000 blower rpm.  At the top it's only about 30%
more.  That's a good thing, but ultimately I'll need to factor in some 50hp of loss in
my air flow calculations and that's not good.  Also note that the faster you spin an
Eaton S/Cer, the more volumetric efficient it gets, up to and above 90% at 14000+
rpm!  Remember the flow loss due to reversion, well the VE chart confirms this as
the pump is more efficient at lower pressures but not by much.  The temperature
graphs indicate that charge temp will be increased some 200 plug degrees F above
ambient.  That's about 270 degrees at standard conditions.  Hence the need for a
intercooler is confirmed.
So where are we?  I'm about 25 CFM short of my target and I need another 70 or so
CFM to make up for the 45-50 hp lost in driving the blower.  So my new adjusted
target is about 750 cfm.  Can the Eaton M90 do it?   

As the extrapolated graph shows the answer is YES!  But I'll need to spin at 16000
rpm to do it! (see above).  That's a two to one drive ratio!  Now, simply
extrapolating data as I did is typically a no-no without some backup.  Heck, the
thing could blow apart, stop making flow, go ballistic thermally, or suck a
disproportionate amount of power.  However, After some research I found several
sources for modified M90 housings (ported) that not only make 16000 rpm and 750
cfm possible but also make the basic M90 much more thermally efficient and take
less hp to drive!  Can't beat that!  (to learn how this is done, see technical section)
Design Specifications