Touching on Scott's comments, the factory manifold is built primarily
for emissions purposes particular catalyser light-off on cold startup.
Mazda's design is not far off from Paul's 3D tangential misload
except using a small diameter collector.  They bring in short 1.75" dia.
tubes aligned close to the top of the 2.75" collector for the center and
rear port with the front port elbowing directly into front collector
end.  Other internal emission features are the flange openings being
larger that the port opening and then round tubes butted flatly over
the rectangular flange opening to create turbulent shear vortices that
promote mixing the secondary air with the combustion gases.  Even the
short tangential tubes end abruptly with no flairing or radiused edges.

I would think that a stripped custom version of the Mazda manifold build
for our purposes with individual plate flanges of minimal dimension and
extra light gauge inconel tubing could be closer to 4 or 5 lbs, but the
implications of a crack failure are much greater for an aircraft than a
race car.  If you wanted to get fancy I suppose elbowing the front port
into a megaphone and then bringing the center and end ports into the
megaphone on a simple tangential might be a consideration if you are
willing to throw the primary length tuning out the window, as well as
the muffler chamber concept.

As for welding the tangential design utilizing long primaries and a
small diameter collector it might be more feasible to fabricate the
primaries using the flat plate design.  Cutting and wrapping ring
segments with shared walls and a top plate fit flush and the wrapped
through the spiral curvature would likely only require seal welding the
exterior joints.  The shared inner walls would likely only need stitch
welding to the collector surface.  In my case it would also allow using
a rectangular 1.75" wide x 1.5" high primary for a lower profile space
consideration should that design concept be pursued.

There are more traditional tube header designs coming from Japan and
they are quite expensive and barely lighter than the Mazda manifold.  I
picked a used one a while back, a bit of a rare find here in the
states.  The power gains claimed haven't been seen here in the US, but
tuning options short of complete stand alone engine management have been
limited until recently.  Some have attempted to keep the shared center
port split into separate smaller diameter primary tubes, but they have
fared no better on the dyno.  1.75" diameter seems to be the common tube
size and the lengths seem to vary between 16" - 24", some close to equal
length and some not.  Of course they are somewhat limited to mating with
the existing exhaust design as opposed to a completely custom system.


Unfortunately long primaries carry the cost of complexity and weight.
If the tuning gains do prove to be minimal at best, then the next best
option would appear to be the simplest, lightest design possible.  So on
the surface it appears to be a race between the simple and tuned
tangential designs.

Mark Sipe


The Rotary Engine NewsLetter. Powered by Linux.
ACRE NL web site. http://www.rotaryeng.net
Copyright 1998-2006 All world wide rights reserved.