To anyone interested:
The following is a response to Doug’s
recent request for updates on flying rotaries:
In early June, having recovered sufficient
courage to try another cross country trip after feeding my prop to my tow bar (yes
Ed, I still have the tow bar) in Minnesota, we finally set out on another one
in our 13B powered RV-6A, this time to the west. After squeezing between
the top of the controlled airspace and a cloud layer over Great Salt Lake, negotiating
a narrow corridor which happened to be populated with scattered rain
showers between restricted airspaces, and a fuel stop in Wendover, UT, we
crossed Mono Lake, the Sierra Mountains, and Yosemite Park at 14500 ft with 50
knot headwinds. Thumbing my nose at the concept of shock cooling, we descended
over 11,000 ft in 25 miles and landed at our destination of Mariposa-Yosemite
airport. Other than a little initial roughness on bringing the power back
up in the pattern after that descent, probably due to some lead fowling of the
plugs since 100 LL was the only suitable fuel available at Wendover, the engine
installation behaved flawlessly. This is in contrast to the performance
of the pilot with the different sight picture at the slower ground speeds at
lower density altitude than I ever see at home in Laramie, WY, and the inclined
runway surrounded by hills at MPI. Thankfully, those details are not
rotary related however, and need not be expanded upon further.
The friends we stayed with near Mariposa
built an award-winning RV-6 with an O-360 and CS prop. While we were
there, we took a couple of side trips with the planes which gave the
opportunity for some side by side comparisons. Their take off and climb
performance was much more impressive than ours which was not surprising since
our Performance Propellers fixed pitch prop is set up primarily for cruise
rather than climb. On one trip, after flying a mile out over the Pacific
at Monterrey Bay (just to say we did it) I set up the plane at max power
and made the flight back to MPI under these conditions. Altitude was 5500
ft, fuel burn was 14.8-15 gal/hr, MAP was 23.8”, RPM was 5900, and IAS
was 148 knots. I don’t know what the OAT was, but it wasn’t
out of the ordinary. Our friends reported burning close to 9 gal/hr while
flying alongside us during this time. I was somewhat surprised by this
and one thing in particular came to mind: parasite drag. The attached
picture was taken during this flight and two sources of drag can be readily seen:
the large and sharp angled cooling air outlet, and the “bomb” which
is a muffler.
The trip home was relatively uneventful
and involved a detour over Mammoth Lakes, CA to allow more time to climb and a
lower altitude for crossing the Sierras. At the fuel stop at Wendover, I
discovered a crack in one of the braces between the rear of the engine and the
exhaust header, but this wasn’t a particular cause for worry since I had
flown for more than 40 hrs before even installing the two of those. Since
we did not have headwinds on the return trip, the total flight time home was an
hour less than the flight time out. The landing in Laramie was in
conditions not uncommon to this area, landing on runway 21 with the ASOS
reporting winds from 250 degrees at 22 gusting to 30. Thankfully, I had
completed a flight review using our plane just a few days before starting the
trip and crosswind technique was one of the things developed. At least
this was the last stop because the seat was definitely at risk of being soiled.
After returning home, a drag reduction
attempt was made by cutting away much of the bottom of the lower cowling and
reshaping the cooling air outlet to be much more similar to the original shape
as supplied by Van’s. This cut the outlet area approximately in
half to 77 sq in which is still larger than that in the unmodified original
cowling. This configuration is shown in another attached photo.
Tuft testing showed no turbulent areas except for one tuft right behind the
exhaust header outlet. Test flights at max power were conducted with and
without the “bomb”.
Without the
muffler: 163 knot TAS at 11500 ft DA, 12.4 gal/hr fuel flow, 5800 rpm,
19.9” MAP
With the muffler: 160 knot TAS at 11500 ft DA, 12.2 gal/hr
fuel flow, 5750 rpm. 20.2” MAP
The 3 knot penalty is a small price to pay
since the noise level without the muffler is unbearable for any length of time.
Cooling is still satisfactory but under
cowl temperatures are higher (~160 deg F) than before as would be expected with
decreased cooling air mass flow. An enclosure was constructed around the
stock Mazda ignition coils and a blast tube installed from the oil cooler air
inlet to this enclosure which keeps the coils at less than 110 deg F.
Several other changes have been made since
first flight over two years ago:
One is the elimination of the cowling combustion
air inlet (snorkel) and filtered air box as originally supplied by Van’s.
A NACA duct inlet and air filter was installed in the cowling side very near
the throttle body. This eliminated the over 3 foot long 3.25 inch diameter
skeet duct from the air box to the throttle body. This was done to avoid
the heating of the air while inside this duct which resulted in temperatures of
the air entering the throttle body in excess of 110 degrees F in flight
regardless of the OAT. Now, the air entering the throttle body remains
within 2 degrees F of the OAT. The result is a lower Density altitude
seen by the engine.
Another change was the elimination of the
vacuum system which included the DG, AI, Mazda smog pump used as the vacuum
pump, vacuum regulator, and associated tubes and filters. In place of the
smog pump, a second alternator was installed. The second alternator led
me to eliminate the second PC680 battery. The instrumentation was
replaced by a Dynon D10A EFIS. A net weight decrease of 19 lb was
realized by these changes.
One other change that I made before the CA
trip involved problems I had with tuning the system to transition the injector
staging point without hesitation or misfiring of the engine. This had
always been of only slight concern until during one flight when advancing the
throttle after a stall, the engine hung up and didn’t produce full power
for almost 10 seconds. As best as I can remember since some sense of
panic ensued, first backing off the throttle position significantly from full
did not clear the condition, but tweaking the mixture control did. The
behavior seemed somewhat similar to SAG as described by others on this list,
but I could not correlate the event to spark plug condition. Further
attempts at tuning the mixture table in the staging region were unsuccessful at
eliminating this behavior. As also described as having been tried by
others on this list, I tried to pneumatically filter the manifold pressure
input to the EC2 using fuel filters and various sized orifices in the lines
between the plenum and EC2. It was possible to induce surging with orifices
of too small opening, but the staging difficulty was not improved with various
combinations of reservoir and orifice size. During these tests, I found
that by proper (improper?) manipulation of the throttle, I could repeatedly induce
the extended lack of power increase on injector staging. Data recorded by
my monitoring system clearly showed the rpm plateau persisting while changing
manifold pressure with the throttle, average fuel flow consistent with the
associated MAP at any given instant, and a lean condition as shown by
both O2 sensors. Since the fuel flow was reasonable, the excess O2 could
result from ineffective ignition: fuel was there, but not being burned
and using up the O2. No faults with the ignition system could be found by
tests including among others checking the ignition timing and using new spark
plugs and wires. Since the condition could be induced by changing
manifold pressure, I postulated that the manifold pressure as seen by the EC2
may be unstable or unreliable. Searching the Motorola literature on the
sensors used in the EC2, I found a reference to additional components
recommended to be installed on the sensor output if the output is connected to an
A/D converter. In desperation, I installed these components into my EC2 and
immediately the misbehavior on staging disappeared. I have not been able
to induce it again no matter how hard I try. My theory is that under
certain conditions, the MAP sensor output goes into oscillation which may or
may not persist depending on the circumstances. I have not verified this
by bringing the output signals external to the EC2 to an oscilloscope since if
they are subject to instability, doing so may provide more questions than
answers. I have pretty much kept this under my hat to this point, but
enough time on this change has been accumulated that I am comfortable that the
change is not detrimental at least. I certainly do NOT recommend or
encourage changes to the EC2 without Tracy’s input without fully accepting
the risks. I have no proof that my interpretation of the events related
above is correct, simply that it helped in my individual case.
I had hoped to make it to the roundup this
year, having bought the charts, planned the flight and arranging time off.
The front that Bill managed to cross was too much to try to take on from this
far away. So here I sit watching the blowing snow outside wishing that I
had made it to the roundup and that the front had materialized after getting
there and thus preventing my return.
One additional note for those who made it
this far. I’ve been spending considerable time in our local ice
cream parlor in an attempt to maintain a constant system in-flight gross weight
after the elimination of the hardware as described above.
OOOOH the sacrifices one has to make in
the name of research….
Steve Boese
RV-6A, 1986 13B NA, EC2, RD1A …