1) The most challenging part of my installation was
the packaging design of the blower it'self as
everything else is pretty much is dictated after that.  
My challenge was to find a location which resulted in
the most optimum flow path and also allowed the
best blower volumetric and thermal efficiency within
the constraints of the NSX engine compartment
using an Eaton M90 style blower.  The S/Cer intake
and exit tract geometry were of primary importance.  
I wanted these to be specific lengths and volumes as
well as being as straight as possible to keep air
speeds optimum and take advantage of 2nd and 3rd
order wave dynamics.  The only way to accomplish
all my design goals was to mount the blower
sideways, exit towards the rear.  Shown above is
one of my early blower support templates used for fit
checks.
2)  From templates to initial layout for the S/Cer.  Horz plate is
bolted to intake at two front manifold bolts and six EGR cover
screws at this point.  Strong and stiff enough for further
mockup with repeatability.  The EGR screws reacted both
bending and shear loads while the large maifold bolts are
mainly to resist bending.  Additional supports for torsion and
bending were added later.  3/8 6061 T6 plate stock was used
for both Horz and Vert supports.  The S/Cer was
pre-positioned and pre-drilled to the Vert support plate.  
Locating the Vert plate to the Horz plate was done on the
engine for precise alignment.  Note the snap line over the
S/Cer pulley.  It was used to locate the X axis (pulley) to the
crank and idler pulleys.  The Y axis was chosen so that the
S/Cer would be mounted as far rear as possible while still
allowing room for the IC and manifold ducting to be sized
correctly and clear other components.  Mock-ups of these
were made to simulate the actual articles.  The Z axis was
dictated by access to the oil filler cap.  Tilt about all three axis
was aligned relative to the crank axis using various
coordinates on the block.  Note two "V blocks" with kant twist
clamps securing the two plates as they are aligned.  Once the
final alignment was obtained, the plates were removed as an
assembly (clamped) and tack welded in place followed by
final welding.
3)  Two Spearco air-to-water charge coolers were
selected, each rated at 250hp min.  One key design
consideration was not only HP/flow rating but size
and configuration.  Large frontal cross section area
and a relatively short core thickness results in the
best compromise between pressure drop and
thermal efficiency.  Since the configuration I chose
required the two cores to be positioned side-by-side
I machined the mating surfaces flat (the blue tape
covers the actual bars and plates to prevent debri
from entering during fly cutting.  
4)  Here's the two cores after they were welded
together.  Note the studs in the blower.  The blower is
installed to the "farside " of the support plate and
secured using nuts torqued on the nearside of the
plate
5)  Once the S/Cer support assembly was welded and the IC
mounting flange machined and bolted in place, the IC to blower
plenum was fabricated using 1/8 6061 T6.   Besides enhancing
airflow using plenum size, volume and configuration, there are
several other more subtle advantages to this design.  For example,  
the S/C support assembly is relatively thermally isolated from
engine heat which backloads into the S/Cer (an issue with the
Comptech system).  This keeps blower housing temperatures and
thermal gradients lower and results in less thermal distortion of the
housing which causes rotor to housing tolerances to grow and lower
overall blower volumetric and thermal efficiency,  In addition, the
integral design of the IC helps absorb internal heat generated by the
blower to lower temperatures even further.   A remote IC cannot do
this.  As the air flow is very turbulent upon blower exit  the integral IC
help create a more laminar flow very quickly thus reducing flow
losses through the majority of the intake path.  Finally, an air-to
water IC is inherently more efficient air flow wise (compared to
air-toair) since the water passages can be much thinner due to the
superior heat energy absorption properties of water.  This results in
a much more aerodynamically friendly package resulting in very low
flow loss through the IC.  
6)  Here's the view from the input side of the IC into the
blower prior to plenum integration to the core.  The "V"
shaped S/Cer outlet is a function of the rotors helix
shape and is one of the areas for both volumetric and
thermal efficiency improvements shown later.  The two
smaller ports help keep blower "whine" reasonable by
reducing dynamic pressure variation during the
backflow compression stage.  The hole in the left hand
side of the plenum is for the bypass valve.
 
7)  A better view of the bypass valve which is similar to a vacuum
actuated mini throttle valve.  The bypass valve connects the pre S/Cer
intake manifold to the post S/Cer IC plenum and is operated by the
spring loaded diaphram connected to manifold vacuum .  The valve is
"normally closed" as shown.  Under light loads the valve is opened by
high manifold vacuum.  This allows air to effectively "bypass" the
S/Cer.  The theory is that this reduces paraisitic S/Cer loads on the
engine during cruse for increased fuel mileage and better driveability.   
When the throttle is opened significantly the vacuum drops and the
spring overcomes the negating vacumm pressure and the valve closes
allowing boost to build.  This operation is different from a typical
"turbo" bypass valve which dumps pressure when the throttle is
suddenly closed.  
8)  A view of the input side of the IC from the blower's exit
standpoint.  As you can see, the bar/plates are very thin (these
are the water passages) and the vast majority of the air-to-water
IC is dedicated to airflow and heat transfer to the water.  While an
air-to-air IC might TEST out at a higher efficiency, the air-to-water
IC produces a MUCH more consistent air temperatures due to
the thermal mass of the water in the system.  This becomes a
very important factor when tuning and during actual track
conditions when air flow through the front heat exchanges is low
(slow speed corners, etc).  In these cases, IAT (intake air temps)  
will be lower then a comparable air-to-air.   Note the
counterbored holes.  The IC uses the blower studs for
attachment.  The blower is installed to the support plate as
described in #4 and  the "heads" of the blower attach nuts reside
in these counterbored areas during IC installation.  The IC then
uses a second set of attach nuts to secure it to the support plate.
 
9)  Initial fitcheck of the unfinished IC/blower
assembly onto the engine.  Various critical clearance
measurements are made for subsequent IC exhaust
manifold and end tank fabrication.
10)  Another view of the Initial fitcheck of the
unfinished IC/blower assembly onto the engine.  Note:
bypass valve is removed..
11)  Next the IC outlet plenum is fabricated.  The
manifold takes relatively slow moving air out of the IC
and begins to accelerate and turn it as gently as
possible.  I found out during testing that the dynamic
pressure amplitudes and frequencies produced by the
S/Cer and the large and flat surfaces of the IC exit
plenum caused some flexing of the sides and
subsequent stress at the weld joints under boost.
Additional gusseting was added to elimiate this
problem.  Again, 1/8 plate was used.  The "bar and
plate" core design is still visible as I have yet to
fabricate the water endtanks.  Welding the core to the
plenums proved to be tricky as very small pinholes
can develop.  These can only be found and corrected
by pressure testing the IC BEFORE the endtanks are
welded.  
12)  Anther view of the IC exit plenum.  Note the
"bypass valve plug" .  My design can accomodate
various types and sizes of S/Cers by adding adapter
plates to the support plate.  Some S/Cers have
internal bypass valves so I fabbed a plug to replace
my bypass valve just in case of a future swap.  Note
that I have already machined the throttle body intake
mounting plate and have fitchecked it onto the
blower's intake port.  
13) Next up was the fabrication of the throttle body's
main manifold.  It was designed to accommodate a
75mm TB and with a specific length and diverging
angles.  I used four sections of 3" dia 0.10 wall 6061
T6, each specifically cut and hand formed to produce
the desired shape.  At this point the quadrants are
tack welded in place for a fitcheck to the mounting
plate shown.  After fitcheck the manifold and mounting
plate were welded completely to form an assembly of
two of the seven components making this particular
part.
14)  The third component, the throttle body attach
housing was machined and welded to the end of the
manifold and it was ready to accept the 75mm throttle
body.   Not shown is the bypass tube and additional
gussets needed to re-enforce the upper mounting bolt
location for the addition of the support plate strut.
Note: The stock intake manifold is still mounted to the
heads.
MSC Performance NSX S/C