15)  Finally, the end tanks were fabricated and welded
onto the cores with the four #8 water hose bunges and
fill cap.  In this view I'm lapping the mating surface
after post weld machining using a flat table.
16)  After quite a bit of analysis I decided the stock
intake manifold left quite a bit to be desired and put a
major restraint on my design goals.  The stock
manifold is a fine piece of engineering, combining
the flow and accustic properties of both a single
plane and dual plane manifold into one with it's
variable volume system.  However, it had two
drawbacks in my mind.  One, it required the air flow
negotiate a significant turn (for packaging
considerations I suspect).  Two, the casting has very
thin walls to keep weight down and therefore the
intake opening could not be increased to my design
spec of 3".  Therefore, to incorporate a larger more
direct entry for my 3" dia IC to main intake manifold
tube size it had to be heavily modified.  This mod
began with the "beheading" of the stock intake
snorkel from it's upper casing as shown above.  
17)  With the TB portion of the intake essintially chopped off, I
needed to create an interface for a new intake entry manifold
into the main plenum.  I sized an opening and machined an
interface plate for this purpose.  The problem is that the stock
manifold is multi-piece and split at my new interface.  Therefore,
I needed to split my interface plate and weld one each to both
top and bottom castings.  The top casing and welded on
interface plate is shown above.  Also note that the EGR
distribution network has been machined away revealing the
actual individual EGR ports in each intake runner.  These were
welding closed.
18)  Here's a view of the same upper casing from the
bottom.  The center divider is removed and the edges
radiused.  After welding, the engine mating surfaces
were fly-cut flat.  
19)  The stock variable volume butterfly housing
became the building block for my lower manifold
housing.  The butterflys were removed and all
associated material machined away.  The upper and
lower housings were bolted together and the lower
section of my split interface flange was located and
tacked in place.  In addition, individual pieces of
6061 T6 were cut to size were tacked in place to
complete the  "belly pan" part of the main plenum.  
Plenum size was selected to complement mid to
high speed performance.  Final welding and
machining was done on individual parts and as an
assembly to square things up and ensure good
sealing at all interfaces.
20)  To create an optimum transition from IC feed
tubing to main manifold yet another fabricated
housing is required.  This one is made using 3/8"
6061 T6 for the bolting flange and two hand shaped
sections of 3" dia 0.075 wall 6061 tube for the main
body.  Two bunges were added, one for the power
brake vacuum take-off and the other for the IAT
sensor.  Additional vacuum ports can be seen on the
lower righthand side.  These are used for the fuel
pressure regulator, MAP sensor, bypass valve and
dash mounted boost gage.  The housing was
clamped in position on the main manifold and all
holes piloted and matchdrilled.  Holes in the main
manifold flange were then opened and tapped for the
8 attach bolts.
21)  The entire Main manifold assembly (upper manifold, Belly
pan and transition manifold) were pressure tested to look for
leaks.  Note the bar plates covering the runner openings.  The
only surprise found during this test was that those fancy $10
each "T" bolt hose clamps are worth a SH--!  At least when
were talking medium boost levels.  No matter how tight I
torqued them, lubed the hose (silicon grease), or rotated them
they still leaked at all the clamped interfaces (particularly if the
interface wasn't PERFECTLY round).  I suspect it has
something to do with the rather low local clamping pressure on
these wide band type clamps.  I replaced them with good
quality stainless steel worm type clamps (as shown) and the
problem immediately disappeared.  Now I'm not saying you'll
loose much boost as the leak rates would be relatively small
compared to the total flow, but it was still very interesting.  
That's about 12 psi on the gauge.  
22)  A view showing both manifolds bolted to the engine.  
Note: Optimum air-flow entry has some side affects.  In this
case, interference with the thermostat housing assembly
(removed in this shot).  Several modifications to the
thermostat housing needed to be performed to remove the
interference.  In addition, the IAC (idle air controller) and fast
idle valves were discarded.  The IAC is not a big deal for my
purposes as I can adjust my idle speed the old fashion way,
at the throttle body.  Adding a few hundred rpm to the basic
idle speed adequately compensates for additional engine
loads and varying environmental conditions.  The loss of the
hot idle circuit simply means I need to keep engine speed up
using my right foot while warming the engine .
23)  One of the major changes to the thermostat housing is
the complete removal (cut off) of this particular part of the
casing and subsequent sealing of same.  It houses water
circuits for the hot start valve (small tap), coolant tank feed (
med tap), and passenger compartment heater return (large
tap).  All but the hot start tap were re-established at other
non-interfering locations.  A better perspective of this mod
can be seen in the "Thermal Control" section
24)  Before the Supercharger assembly could be mounted some
additional modifications to several support systems needed to be done.  
For example, the location of the pressure side fuel line to the fuel rail
need to be moved due to interference with the S/Cer support plate.  I
machined off the stock fuel rail to line interface boss and welded it
closed.  I then machined the forward end of the rail to accept a #8 AN
bunge and welded it in place.  A new fuel line from the stock filter was
fabricated, routed and installed as shown.  550cc RC injectors were
also installed.  Note the fuel pressure sensors incorporated into the
banjo bolt on the rear bank fuel rail.  Detail on this mod can be seen in
the engine management section.  Also shown is a lengthened alternator
lead, re-routed harnesses and the front S/C support which also secures
the belt tensioner pulley and Idler pulley.  Note, the belt routing shown in
this picture was the revision A design, not the initial design.  Another key
item is the elimination of the PCV valve and incorporation of a second
valve cover vent on the front bank.  Both vent and PCV hole are attached
to a the same breather hose and routed to the catch can.  Oil in the air
stream in any form via the PVC valve, bad valve seals, blow-by hose or
poor oil control rings leads to detonation and must be avoided. Hence
the removal of the PVC and blow-by recirculation systems.  
25)  Here's the S/Cer blower assembly ready to bolt
on the engine.  Once the S/Cer assembly is mounted
the IC can be bolted in place.  I machined a small
adapter so that the stock throttle positioner could be
used on the Accufab Ford throttle body.  Support strut
is made of graphite.  The funny looking tube jogging
out from the TB manifold is the bypass tube.
26)  This shot shows quite a few things.  Significant
are: The re-location of the coolant tank air separator
spigot on the thermostat housing (see directly under
new manifold on lefthand side), addition of the ICs
upper and lower water tanks along with it's pressure
and return water lines (note; flow is bottom to top),
tacked together IC to main intake feed tubing.  
27)  Another view of the overall air flow system.  Take
note that the air flow from throttle body to intake
runners is as short as possible with large radius bends
(as few as possible) and smooth transitions.  Note the
dual IC water hoses.  Due to the need to keep the
overall IC profile as low as possible to allow adequate
dynamic clearance to the hatch glass, the end tanks
needed to be half height.  This dictated that two #8
size fittings be used as opposed to one #12.
MSC Performance NSX S/C