> >I have to confess that I need someone to give me a short, one paragraph
> >lesson on the mechanical factors involved in turbo-compounding.
I'm not
> >an engine guy but I can learn quickly. Here's what I understand so
far.
> >
> >If the turbo is always turning at, say, 75k rpm cruise power, then it
> >would be easy enough to direct drive thru gear reduction to assist the
> >final drive at the prop hub. OK. Reduce 75k rpm down to 2500 rpm
> >(30:1). But, and this is the part that I don't understand completely,
> >things are not always turning at these speeds. As the prop slows down
> >for cruise flight (say down to 2,000 rpm) that would also drag the
> >governor down to 60k. Not good because that's not its efficient speed
> >and will build up backpressure. It wants to keep turning at 75k.
Also,
> >if the engine is throttled back to idle and the aircraft dives, the
prop
> >may be back up to 2,500 rpm but the turbo only wants to turn at half
> >that speed. I think I understand that much. What's needed is a
> >variable connection between the 2 with high efficiency (i.e. direct
> >gears, not hydraulic or electric motors). Also, I believe that some
> >system needs to identify what the correct ratio should be and
> >automatically set the gearing accordinly. From here, I'm lost. What
> >are we trying to accomplish? Are we trying to allow the turbo to turn
> >at whatever speed it naturally wants, but also connect it somehow to a
> >prop shaft that can vary in rpm from 500 to 2500? Or am I missing
> >another aspect completely.
> >
> >Anyone got a "turbo-compound for Dummies" book they can loan me??
> >
> >Greg
> >
> >Since the aircraft engine spends 95% of the time at constant cruise RPM
> >it is easy to optimizes the gearing for that cruise RPM. Also back
>pressure has
> >just a little effect on the engine power.
> >
> >Here are some numbers.
> >
> >Paul Lamar
> >
> >Greq did you watch these youtube videos?
> >
> >http://www.youtube.com/watch?v=iy_Jd5kewbM I want my MPG!
> >
> >http://www.youtube.com/watch?v=luoUHrX6DeQ I want my MPG Diesel
> >
> >http://www.youtube.com/watch?v=emLKhfm6MWE I want my MPG Q and A
> >
> >Any questions? Where did I go wrong?
> >
> >This chart pretty well sums it up. 30 inches of Hg exhaust pressure is
> >sea level pressure. The big difference between a piston engine
> >and a rotary is a piston engine has only half the kinetic energy
available
> >for blow down turbines due to the exhaust valve. If it is this good
>for a piston
> >engine think what it would be for a rotary.
> >
> >Paul Lamar
> >
>
>Paul,
>
> I wonder if this chart data is actual, predicted or both? The first
>two lines of data seem believable the second questionable.
>gear efficiency at 95%.
>
>Doug in Japan
>
>I assume you got the entire NACA-822 pdf.
>
>Here is a planetary drive where the gear eff. is claimed to be 99%.
>Granted it is not 10 to one or 60 years ago.
>
>Paul Lamar
>
>Greg,
>
>Here is the other piece of the puzzle. Kinetic energy loss across the
>exhaust
>valve. This is the fundamental reason the rotary is twice as good as the
>poppet
>valve piston engine such as this CW R3350 TC.
>
>Paul Lamar
>
>Paul:
>
>I'm still not sure what would happen to the turbo if it's output were
>geared down and direct connected thru an RGB to the prop shaft. I know
>the rpm range of the engine, from idle all the way up to max 2,500 rpm.
>What I'm missing is the relationship with the prop speeds to the turbo.
>Yes, the turbo varies its speed depending on the throttle settings, but
>is it just too simple to say that an idle setting (i.e. 800 rpm) would
>produce a low turbo speed (low exhaust gas velocity) and a high throttle
>setting (i.e. 2400 rpm) would correspondingly speed up the turbo (high
>exhaust gas velocity)?? Yes, there would be mismatched gear ratios
>along the way from idle up to full and back to idle, but 95% of the life
>of the engine would be at 75-80% max cruise power. Why could the turbo
>not be geared for this flight condition and forget about the
>inefficiencies during throttle up and down? Why did Doug in Japan say
>that we all agree there would be some broken gear teeth if we did this?
>If the turbo is trying to turn faster than the engine, it will slow
>down. And if it's trying to turn slower, it's because the aircraft is
>in a descent with the prop turning faster than the engine wants to push
>it (or spin the turbo). I still don't understand why a torque converter
>or fluid coupling is needed other than more efficient at all speeds.
>
>Greg
>
>Yes that is it exactly. However what I have implemented is an over
running
>clutch. If the e-shaft is running faster than the turbine no torque is
>transmitted to the the turbine. Only when the turbine is running
faster than the
>engine does torque transfer from the turbine to the e-shaft.
>
>See the attached jpgs.
>
>The torque coupling is designed to isolate the engine torgue pulses from
>the turbine gear box. In my design I don't need it as I am using a
friction
>belt drive from the e-shaft to the gear box that can slip when it sees
a torque
>pulse.
>
>Paul Lamar
>
Paul,
I am using a multi-v friction belt as well. Do you really think the belt
will slip a bit to absorb torque impulses? Slipping would create heat.
I would think it would flex instead. An idler puller that tensions the
belt with a spring or better yet, a hydraulic damper might be just the
ticket. Greg's question was actually addressing the differences in
turbine and e-shaft rpm through the engine power band. Is the fluid
coupling that the Connie and new diesel use, just to isolate torque
pulses or is it actually a torque converter? In other words does it
adjusts for different rpms inputs from the turbine side over the power
range.
Greg.
I'm not the sharpest tool in the drawer, but at this stage in design as
I see it, turbo-compounding is something we have to research in steps
and follow prior art until we understand what is happening. In Paul's
example, the turbine power is isolated from the engine via a belt.
This means it is not 100% gear to gear. I have followed that design as
well.
You brought up a good point about the possible changing ratios between
engine speed and power turbine speed. From my understanding, turbo
compounding shows good benefits when the engine is at high power
settings and that narrows (excuse the pun) the challenge down a bit. At
low power settings like idle, the T.C. would automatically disengage
because of the over running clutch. Used as a point design device only
with, for example, a 6,000 to 6,500 engine cruise rpm, perhaps we
don't need a torque converter in the drive train. If engaged on max.
take-off rpms as well, perhaps we do. That is at least what the first
experiments are striving to prove or disprove. I personally wouldn't
trust my first design attempts enough to depend on it at T.O. but don't
mind engaging it when the aircraft is at the desired flight level, the
plane trimmed out and I have the time and state of mind to pay
attention to it.
Now as to where to bring the power into the system, prior art has shown
to bring it in as directly as possible into the power shaft. So instead
of bringing power into the prop hub at say, 2000 rpm, which would
entail redundant reduction systems, we would want to bring it to the e-
shaft before the planetary PSRU, or on other end of the engine in
Paul's case. If for example the PSRU has a gear ratio of 3:1, the
engine would be turning at 6000 rpm. The biggest challenge yet, to sort
out, depends a lot on the actual turbine selected and how it performs
under load. Some radial turbines, like on common turbo chargers, are
happy to work over a wide range of exhaust flows at varying efficiencies
of 68 to 75%, while others devices, like axial turbines, have higher
efficiencies but operate at a narrower band of rpm in which they can be
used. With axial turbine major design engineering has to be done to
figure out gas flows and plumbing. With radial turbines the gas flows
are easier to route with basic engineering and we have a big
off-the-shelf selection of turbines. In addition with radial turbines
we already have some performance data on how they work under load,
because the load they move is the air compressor directly attached to
them. I am of course speaking about turbines coupled to compressors or
turbo-chargers. The data shows how much the turbine slows down depending
on how much work the compressor is demanding from it. Higher compressor
output pressures cause the attached exhaust turbine to work harder, and
to some extent, slower within its design parameters. This explanation is
simplified as there are several variables that affect how a turbine
works and it will differ on different engines.
There are a few candidates from Garrett that may work, if we can get a
enough power out of them at a rpm, say, 80,000. For example lets say we
use a larger Rotrex traction drive and run it backwards (no problems
says the makers except it needs the oil routing reversed) so instead of
stepping up the speed to run an radial compressor we used it as a step
down device (7:1 ratio) and thus can safely transfer up to 80 hp.
OK, so from 80K rpms this traction drive would lower the rpm to 11,428
at the output.
Say we want the rpms to arrive at the e- shaft at 7,000 rpms. Note
after testing we may find that this number is wrong and it should be
6500 or even 8,000. If we decide on 7K rpm then the pulley to pulley
ratio is 1.63:1 My engine design has a combination torque absorber,
and load cell within a multi-V pulley mounted on the counter weight
inside the bell housing, and it is 142mm (six inches) in diameter. The
pulley on the traction drive output would have to be about 86mm in
diameter to maintain the desired input speed. That is 140mm / 1.63 =
86mm dia. But what happens now? As you indicated, in operation, the
exhaust turbine will slow down because it is doing work and as such it
is exchanging rpms for torque. This torque is transmited to the engine
e-shaft via the turbo compound system and causes the engine to want to
speed up.
In a perfect world the engine speed and output of the turbine cum
traction drive rpms would both equalize at some point between the two of
them. The increased torque is helping the engine out, lightening the
load so to speak so it will want to run faster. And now for the key
point of which you know very well. The world is not a perfect static
place. Without the constant speed prop, the pilot would be on a wild
goose chase trying to fly the plane exactly level, and adjusting
throttle settings, and perhaps turbo exhaust by-pass valve to try to
stabilize the engine rpms. Not my idea of fun:) Using a constant speed
prop the governor (mechanical or electrical) senses the engine rpm
change and increases the pitch to keep the engine at the selected
setting.
We speak about these events as if they happen consquetively but they are
near simultaneous.
As a mind exercise, chose an altitude, as high as possible, taking into
consideration your airplane loading and wing specifications. For
instance assuming you have the oxgyen on board can your plane with its'
rotary engine make it up to 15, 20, or 25K feet? The supercharger or
turbo charger must be sized so that it can increase the engine power to
compenstate for lowering air pressure at altitude. If so then altitude
is your friend for decreasing drag, providing cooler intake air,
increase in TAS; and for turbo-compounding, providing more power to the
exhaust turbine because of the increases differences (delta P) between
the exhaust pressure on the inlet side, compared to the ambient pressure
on the outlet side. If possible it would highly advantgeous if the
engine was supplied boost at a level higher than that which it normally
experieces at sea level thus maximizing the delta P.
In closing, think of turbo compounding as having an engine producing,
250+ hp but only having to pay for the 200hp worth of fuel burn. The
other benefits of altitude are icing on the cake that add indirectly to
increased m.p.g. and less time in flight.
In addition there may be some tweaking of engine settings that can be
done to further increase fuel mileage but I'll let Kariban explain that
since I have no experience there. He is one of only two people I have
contacted that have actual flight experience with turbocompounding
adjustments.
Doug in Japan
Doug:
The Mistral has a PSRU with flywheel bolted to the front of the e-shaft
turning at engine speed (up to 6,000 rpm) and a reduction of 2.8:1 thru
planetary gearbox. The interesting thing is that this flywheel is
already dampened for torque pulses. Looks just like a clutch plate.
Notice that the starter connects to the outer gear ring of this
flywheel. I would connect the output of the turbo-compound shaft (after
appropriate reduction) directly to this flywheel (gear-to-gear) and see
how much the "torsional vibration absorber" smoothed out the torque
pulses. No extra torque would be introduced into the e-shaft (as in the
rear powered designs that send the turbo-compound output into the
e-shaft and expect it to carry the additional power). Also, this PSRU
flywheel is only turning about 10-12 times slower than the turbo so not
a big gear reduction is needed.
I could have a sprag (overrunning) clutch installed prior to the final
gear to allow for disengage during low turbo/high engine speed ratios.
Also, just prior to the final gear, the NuVinci CVT type gearbox could
be installed to allow for variations in the turbo/engine ratios. As
mentioned above, the C/S prop will keep the engine turning at, well,
constant speed while the engine exhaust gas pressures will
increase/decrease depending on how much M.P. has been selected by the
pilot (or the FADEC). So for a cruise setting of, say, 2,000 rpm prop
with 75% power at the manifold, the turbine would be spinning at, say,
75-80k rpm. The CVT would adjust to ensure the turbo is loaded up
(slowed down) enough to get some work out of it. If, however, the pilot
then selects a lower power setting (long range cruise, or starts a
descent), the prop would still be set to 2,000 but the manifold
pressures are now way down, with a slower exhaust pressure and a slower
turbo speed. The CVT would recognize the lower M.P. compared to the
prop rpm and readjust the ratio needed to keep the turbo under load.
This constant adjustment of the CVT would be done by simple computer
commands, not unlike a cruise control.
Direct gear-to-gear, fewer parts, torsional dampening already provided,
variable gear ratios to optimize turbo speeds, what am I missing here?
Oh yeah, all these components are already available. I would think I
could just bolt them on to a Mistral PSRU and try it out on a dyno.
What's the worst that can happen??
Greg in Canada
Greg,
Substitute the words "multi- V belt and torque absorber combi pulley"
for the words "flywheel and pinion gear" and we are speaking
essentially about the same approach.
One trouble with driving the starter gear is keeping it lubricated.
That area of the bell housing is full of holes and inspection plates.
In addition a starter gear's duty cycle is extremely light. I doubt
it sees more than a few rpms during each starting event. Grease is fine
in that application.
You may not need a CVT, especially if you use the new variable nozzle
turbines Honeywell Garret is bringing out. Varying the angle the
exhaust gases contact the blades, effects the torque output and speed of
the turbine. Simpler than a CVT and it is already built into the
turbine.
That reminds me. Paul does Garrett make an appearance at Seven Stocks?
Doug in Japan
I don't think they do. They should. Some of their dealers do.
However I have yet to find a dealer that could answer my questions.
Example: What is the HP output of the turbine as a function of exhaust
gas velocity?
Paul Lamar
The Rotary Engine NewsLetter. Powered by Linux.
ACRE NL web site.
http://www.rotaryeng.net
Youtube key word UTUBPLEASE
Copyright 1998-2008 All world wide rights reserved.