I picked up a couple of new brochures at the Mistral booth at OSH.

This is the best data we have so far on lean of peak
operation on the Wankel rotary. We have had lots of evidence that
the BSFC was much better when the rotary was running very lean
but no actual dyno data up until now.

The Wankel rotary engine is inherently a lean burn stratified
charge engine because the combustion chamber rotates. That throws
the heavier fuel molecules out towards the spark plugs. It will
run at mixture ratios the gasoline piston engine can only dream
about. The diesel engine will do it of course but at higher cost
and weight not to mention the particulate emissions (black smoke)
problem. Also, I might add, diesel fuel is only 15% of a barrel
of crude while gasoline is 30%. That is the fundamental reason
diesel fuel is more expensive than auto fuel. Too many cars
burning diesel fuel in Europe. It should be saved for trucks,
trains and ships.


The trouble with this lean burn from a road car point of view is
the NOX increases drastically as the engine is leaned. Mazda
would do well to develop a NOX filter for the RX8 and run the
engine super lean while cruising on the freeway. This would be a
major breakthrough in the fuel consumption of the Mazda rotary
powered cars. OBD II data from my own RX8 engine confirms the
mixture never goes above 14.7:1 air fuel ratio or stoichiometric.
Another possibility would be to have NOX eliminated from
the controlled emissions list. It takes both NOX and Hc to form
smog. Eliminate one or the other and you eliminate smog.

Mistral already has incorporated auto lean on their FADEC to take
advantage of this characteristic of the rotary while cruising.

Thanks Mistral for all your rotary engine R&D and publishing this
data.

http://www.mistral-engines.com/

Paul Lamar

Paul--your observation of the lean-burn capability of the rotary fits with
our flight test experience over a ten year period.  We flew x-country tests
lean of peak on all  the various configuration engines we built.  These
peripheral port engines in the RV-4 and RV-3 used relatively small intake
runners and ports and consistantly burned less fuel in side-by-side flight
tests in x-country, side-by-side tests with other  Lycoming powered RV-4's.
Had same result in the RV-3 compared to the prior years experience Alan had
in a Lycoming powered RV-3. All consistently produced lower mission fuel
burn at same speed/altitudes.  With a fixed pitch propeller, at 3.5 hour
legs we would run WOT at altitutde and leaning to max rpm drop until
misfire, tracking EGT temps all the way...then fatten it up just to barely
eliminate the misfire.  At a given airspeed, it would tolerate leaning to
EGT drops far, far below the Lycoming engines we built.  The rotary engines'
lower comparative mission fuel burns resulted even when we had worse
computed BSFC than the recips.  This appeared to result from a combination
of factors--the ability to run  leaner mixtures at a given airspeed; cooling
drag reductions made possible with the liquid engine cooling; and
efficiencies assocated with bigger diameter,slower turning propellers. The
rotary engines consistantly would tolerate leaning at cruise to a much
further extent than the recips.  With these old school, mechanical fuel
injection systems, we tracked with rpm and exhaust gas temperature.  From a
quick glance at it, I don't see EGT's at given power setting/rpm on
Francois' data.  Is it available?

Related background--  All engines had relatively small diameter intake
runners and peripheral ports with throttle plates located in the rotor
housings as previously discussed.  Several different injector locations
tested--some exclusively with direct-injection in the rotor housing (both
single and double injectors in each rotor housing).  Also placed injectors
in the peripheral port and later in the induction tubes. Direct injection
enabled the most radical leaning without a misfire.  But we had some
problems with misfires at certain throttle settings with direct injection.
These misfires were eliminated throughout the operating range when nozzles
were place in the intake tubes...at the apparent loss of some degree of
stratification and ability to lean which was possible to a greater degree
with direct injection.  We made this corrective nozzle location to the
induction tubes near the end of testing and have no definitive comparative
in flight data.  We do have some early dyno data where we tested and
compared different direct injection nozzle locations.  The first prototype
fuel system was constant-flow and built in-house. Later we used a mechanical
system designed working with Don Rivera at Airflow Performance.  It was
basically one of Don's  systems in modified form, using our remote venturi
for the signal.  That refined system worked flawlessly and that is what we
are reproducing to use on the rotary being installed in the new RV-3.

Steve Beckham