snips ................


   Rumors at the time said it was a disconnect failure of the gear box
to the
prop.

The NTSB frequently gets things wrong so in this case they start talking
about the gearbox and then jump to the roller bearings in the engine
like
that was the cause of loss of power. We had long discussions on the
viability
of the rotor needle roller bearings with Eric Barger. It appears I was
proved
right as Midwest engineers warned me about that. The rotor needles were
lubed
by oil injected into the intake as this engine uses a charged cooled
rotor.
The attached picture is the roller bearing that Eric used.

The NTSB report did not say where the two ball bearings were located.
They could have been main bearings or they could have been in the
gear box.
Must likely they were in the New Zealand (as I recall) built gear box.
Kind of a strange failure as gear boxes usually give a bit of warning.

This is still a mystery as far as I am concerned because I am still
lacking
details on the RPM of the engine, the prop dimensions and the gear ratio
used in the gear box. Effective oiling also appears to be an issue

Can you help me out Eric?

Paul Lamar

===========================================================
Paul,



The engine was running 6000 RPM with a 2.47:1 ratio gearbox (built
in New
Zealand), and a 58" prop (I believe).

The report states "A total loss of engine power due to the failure
of the
reduction drive assembly as a result of excessive wear and high
temperatures"

We believe that the gearbox ran hot and that the oil failed in the
gearbox,
it heated up literally creating bluing on shafts etc. this caused
the engine
stoppage. After engine teardown we were "scratching our heads" as to
what
happened, we dug into the PSRU and as soon as we opened it we knew
where the
problem was as it appeared as the PSRU literally "blew up" inside.
The ball
bearings referred to were in the PSRU and there were balls floating
loosely
inside the box. We did not do the teardown on the PSRU from this initial
examination; the FAA/NTSB boxed it up and took it to their lab.


There was some heat and material marks on the rear needle bearing noted
during disassembly in front of the FAA/NTSB done by me personally, not a
cause for stoppage a noted blemish per say during the teardown.

Paul commented "The rotor needles were lubed by oil injected into
the intake
as this engine uses a charged cooled rotor." The RotaMax engine actually
lubricated the rotor bearing directly from the oil injection pump
thru the
eshaft, and then the oil would enter the air fuel mixture and lastly
to the
Trochoid for lubricating the seals.

As you know Paul the needle roller bearings in a charge cooled
engine have
always been a "Swiss watch" per say as to the required material,
lubrication, temperature and RPM. To this the charge cooled engine
with a
needle roller bearing works well in varied load applications. In
constant
load applications the ambient temperature along with other operating
conditions can cause an engine/bearing running on the edge of its
capacity.

Several ways now to solve this concern if someone considers building a
charge cooled engine, as the simplicity of the engine is sure a
value worth
pursuing. If the right design and testing were done to assure that the
engine performed well at all operating conditions with a factor of
comfort
this could be a simple and reliable approach.

This should clear things up

Eric



Paul and Mark S,

The oil system was an internal bath type system with a small fill/sight
reservoir. There was no oil pump or external components other than
the small
reservoir. We did not do much as to work on the PSRU we bolted them
up and
began testing on the ground, on a stand and then in the aircraft.
The lesson
we learnt here is that you can test engines and PSRUs on dynos,
stands etc
but until the real application testing is done it is still a potential
mis-match that causes failure. This usually plaques a lot of product
attempts as the continued testing and validation costs are usually not
supported especially in start up mode. To me this is the most
important time
to do it right. A costly lesson where in the RotaMax case the more money
spent on testing would have validated the problem in a less exposed
application.

BTW My son in law was flying the plane so we surely had a level of
comfort
and confidence in our previous testing efforts.

Eric Barger

--
Paul and Eric,
Could the gear box failure have been caused in part by the wicked torque
pulses of (any)
rotary design? I know of another failure using an Axro and Rotax C
gearbox.
My
Autoflight gearbox (on a Subaru) has a rubber biscuit dampner. Could a
harmonic vibration
have contributed to this failure with helical gear design? (as Tracy
found
out with his
design and fortunately solved) Is there an easy way to check for
harmonic
stresses?
Dan F


Not likely as the rotary torque pulse are more benign than a 4
cylinder four
stroke piston engine or a two cylinder two stroke engine. Note the 4
cylinder
4 stroke goes negative but the rotary does not. Severe rotary torsional
vibration is a myth. The only difference compared to an aircraft
engine is the RPM is twice as high.

It probably had more to do with the length and small diameter of the
drive shaft from the gear box to the prop. It would have been
much better off with a 3 inch diameter carbon fiber shaft.

However inadequate gear box oiling seems to be the main culprit.

What Tracy found out had nothing to do with torque pulses.
It had to do with the thrust developed by a helical gear.
Tracy first gear box destroyed the thrust bearing in the Mazda.
If anything helical gears are stronger than spur gears.
They also have little to do with harmonic vibration.

Tracy used a rubber torque isolator right from the start.
So did Ross but it was a clutch disk center with steel springs.

The rubber torque isolator in the GB used with the Rotomax was a
bit stiff so High Class Aviation bought one with lower
durometer and installed it in the New Zealand built gear box.
That lowered the resonant frequency from about 2000 RPM
idle to about 1850 RPM idle. That however was a direct
drive prop mounted on the gear box and not shaft driven
like the Vampire.

Paul Lamar



Paul,

I have had a lot of people comment to me on the shaft from the PSRU
to the
Prop. The input given to me is this is a very touchy engineering
task and
that there is severe flexing of this shaft during operation. We believe
this with the other items like oil and the torque Isolator all
contributed
to the PSRU being fatigued to a point of catastrophic failure.

Eric Barger


for what it's worth, there is a multi-year project in central Europe at
a University where they are using a Yamaha R-series 150 hp motorcycle
engine to drive a ducted fan.  They use a drive-shaft to the rotor
because of inlet considerations and a composite shaft.  Interestingly,
they appear to be using BMW type rubber isolators at both ends of the
shaft!  I've never seen this one before, but haven't heard of any
problems, either.

best,

George W