light twins?

Thanks Corky,

My assumptions about the rotary were not based on experience, so it's good
to learn something new.

I would have expected the rotary to have less spiky torque output than a
piston, but perhaps the wobble of the rotor along its ellipsoidal path
creates enough inertia to cause some sizable spikes.

I may post a question about this to the Mazda newsletter -- once Paul gets
back from Osh. I'm sure there will be some people there with good insight on
this.

I still have to think though that maximum torque is the limiting factor in
both gear and belt design. Even if torsional vibration is an issue (with the
rotary or any engine), the way to address that is to dampen the spikes and
prevent harmonics from causing destructive resonance. Just using bigger
stronger gears is one approach, but not really the most elegant -- or
lightweight.

I notice that Tracy uses rubber doughnuts between the flywheel and the
gearbox coupling, just for that reason I would assume.

Your story is just another reminder that gearboxes are one of the big
bugaboos of any auto engine conversion -- and torsional vibration (or
resonance) is always the culprit. I know that in the Subaru community there
is not really a box that I would consider completely trustworthy.

I was hoping the rotary was less of a problem in this area. Darn.

Regards,

Gordon.




"Corky Scott" wrote in message
...
On Thu, 28 Jul 2005 16:08:46 -0400, "Gordon Arnaut"
wrote:

Still, compared to a piston engine, the rotary is a pussy cat when it
comes
to torsional issues, because it does not have the lever-arm effect of the
crankshaft throws to worry about -- which creates the bulk of the
torsional
flex in a piston engine.

This may not be accurate Gordon. There was a company based up in
Washington State that produced a planetary gear box as a PSRU for the
Mazda rotory and they had a HORRENDOUS developmental period with many,
repeat many broken boxes.

They finally got something that was extremely professional looking as
machined aluminum can be, and robust and long lived. They had it on
the front of an RX4 and flew it to various air shows. One of the
developmental partners was killed in an airplane crash and for a while
the psru was still available but I don't know if it still is.

The big issue, the one that was busting props and tearing their boxes
apart was torsional vibration. I remember reading that they claimed
there was something about the rotory engine that gave it a really
powerful torque spike.

I think they eventually solved the problem with some kind of cushion
drive. But for a while it was busting one attempt after another on
the test stand, and a bunch of dead stick landings.

Corky Scott

I notice that Tracy uses rubber doughnuts between the flywheel and the
gearbox coupling, just for that reason I would assume.

As I understand it, there are two ways to avoid the resonance issue. One is
to make the drive coupling tighter, and the other is to make it looser.
Powersports chose tighter I believe, and the precision they need could be
why their drive costs $6000. Tracy chose the looser path. Both work fine.
The biggest unknown I face with the single rotor is the resonance frequency,
and how it works with the rubber dampeners that Tracy chose for the two
rotor engine. I may very well have to change the durometer of the rubber
dampeners, but I won't know until I try it.

Cheers,
Rusty

On Thu, 28 Jul 2005 22:08:10 -0400, "Gordon Arnaut"
wrote:

Your story is just another reminder that gearboxes are one of the big
bugaboos of any auto engine conversion -- and torsional vibration (or
resonance) is always the culprit. I know that in the Subaru community there
is not really a box that I would consider completely trustworthy.

Really? Not even Eggenfellner's? I haven't heard of any failures of
his design yet, but I haven't been actively following Subaru
conversions.

Corky Scott

Yes, the Egg redrive has no failures yet, from what I know.

However, he seems to have taken the "build it strong as hell" approach and
doesn't use any kind of damping, such as elastomers, sprag clutch, etc. He
is also using a heavy flywheel that helps to smooth out the torque spikes.

The result is quite a heavy unit. Still his FWF package is competitive with
Lyc on a power-to-weight basis. Not bad at all.


Regards,

Gordon.


"Corky Scott" wrote in message
...
On Thu, 28 Jul 2005 22:08:10 -0400, "Gordon Arnaut"
wrote:

Your story is just another reminder that gearboxes are one of the big
bugaboos of any auto engine conversion -- and torsional vibration (or
resonance) is always the culprit. I know that in the Subaru community
there
is not really a box that I would consider completely trustworthy.

Really? Not even Eggenfellner's? I haven't heard of any failures of
his design yet, but I haven't been actively following Subaru
conversions.

Corky Scott

Gordon Arnaut wrote:
Yes, the Egg redrive has no failures yet, from what I know.

However, he seems to have taken the "build it strong as hell" approach and
doesn't use any kind of damping, such as elastomers, sprag clutch, etc. He
is also using a heavy flywheel that helps to smooth out the torque spikes.

The result is quite a heavy unit. Still his FWF package is competitive with
Lyc on a power-to-weight basis. Not bad at all.


Regards,

Gordon.


"Corky Scott" wrote in message
...

On Thu, 28 Jul 2005 22:08:10 -0400, "Gordon Arnaut"
wrote:


Your story is just another reminder that gearboxes are one of the big
bugaboos of any auto engine conversion -- and torsional vibration (or
resonance) is always the culprit. I know that in the Subaru community
there
is not really a box that I would consider completely trustworthy.

Really? Not even Eggenfellner's? I haven't heard of any failures of
his design yet, but I haven't been actively following Subaru
conversions.

Corky Scott



Sorry for coming in a little late on this; I usually frequent the
Flyrotary list & Rusty mentioned that this thread was alive over here.
The non-existent email address is there because I got tired of a steady
diet of spam.

Several things come to mind about the previous few messages in this
thread, from the stuff I've read in about 10 years of following
Powersport, then Tracy Crook's development trials & tribulations. This
is from memory & I never claim to have a good memory. :-)

Gearbox strength for 1rotor vs 2rotor: The big deal about a 1rotor is
that the torque curve actually reverses (goes negative) with a 1 rotor,
like a 4cyl 4stroke piston engine. With a 2 rotor, the torque curve
never actually reverses so the gear box isn't stressed as much in the
torsional resonance dept. even though there's twice the power. If you
frequent Paul Lamar's list I'm sure he will be happy to show you the
torque curve for the 2rotor. IIRC, the torque curve for a 1rotor looks
like a 4cyl piston engine, going negative between each positive torque
peak. If the system resonates & you continue to excite it without
damping the resonance, no amount of strength will keep it from breaking.

The 1st incarnation of Powersport are the guys in the northwest with the
rotary powered RV-4 that had such horrendous torsional resonance
problems *on a dyno*. Current thinking is that they had a problem with
resonance on that particular dyno with that particular engine/dyno
coupling (it was built to test V-8's) They also had severe problems
getting their P-port engine to idle properly. Others have had no problem
at all getting them to idle smoothly. The developers had racing V-8
backgrounds & some of that stuff doesn't transfer well to the rotary.
Their internal tooth ring gear, designed to keep the gearbox 'tight',
like Rusty mentioned, is very heavy, very expensive, & if it isn't heavy
enough will actually loosen up as rpm comes up & the ring gear tries to
stretch. Kind of self-defeating. The 'tight' vs 'loose' issue is really
an issue of moving resonant frequency above the operating rpm range or
moving it below the operating range. 'Tight' moves it up; 'loose' moves
it down. Manual transmission cars are 'loose', moving resonance below
normal operating rpm. You've probably experienced the automotive version
of torsional resonance if you've put a manual trans car in 2nd or 3rd &
let the idling engine try to pull the car. If the engine continues to
run, the car will move forward in big surges. That's the resonant
frequency of the drive train. I don't remember Powersport ever having a
problem with broken props or gearboxes; my memory is that they went
straight from their dyno problems to the big internal spur gear. They
did have a gearbox failure when competing in time-to-climb at SNF
because they were using nitrous & over stressed a bearing in the
gearbox. I think they were producing somewhere between 350-400hp (13B
without turbo) when that happened.

Damping torque pulses with belt slippage: inefficient & produces a lot
of heat.

I think Corky mentioned the nightmare of an intake manifold on RX-7
13B's; fortunately a much simpler & lighter intake works fine for
aircraft since low rpm torque isn't needed.

Eggenfellner: I believe they've recently had the 1st failure of one of
their gearboxes.

Charlie
(Rusty's 'hangar away from home' for the next hurricane)

Charlie,

Thanks for the insightful post. What exactly is the source of the torsional
vibration in the rotary? And how does it compare in severity to a piston
engine?

Regards,

Gordon.


"Charlie" wrote in message
.. .
Gordon Arnaut wrote:
Yes, the Egg redrive has no failures yet, from what I know.

However, he seems to have taken the "build it strong as hell" approach
and doesn't use any kind of damping, such as elastomers, sprag clutch,
etc. He is also using a heavy flywheel that helps to smooth out the
torque spikes.

The result is quite a heavy unit. Still his FWF package is competitive
with Lyc on a power-to-weight basis. Not bad at all.


Regards,

Gordon.


"Corky Scott" wrote in message
...

On Thu, 28 Jul 2005 22:08:10 -0400, "Gordon Arnaut"
wrote:


Your story is just another reminder that gearboxes are one of the big
bugaboos of any auto engine conversion -- and torsional vibration (or
resonance) is always the culprit. I know that in the Subaru community
there
is not really a box that I would consider completely trustworthy.

Really? Not even Eggenfellner's? I haven't heard of any failures of
his design yet, but I haven't been actively following Subaru
conversions.

Corky Scott



Sorry for coming in a little late on this; I usually frequent the
Flyrotary list & Rusty mentioned that this thread was alive over here. The
non-existent email address is there because I got tired of a steady diet
of spam.

Several things come to mind about the previous few messages in this
thread, from the stuff I've read in about 10 years of following
Powersport, then Tracy Crook's development trials & tribulations. This is
from memory & I never claim to have a good memory. :-)

Gearbox strength for 1rotor vs 2rotor: The big deal about a 1rotor is that
the torque curve actually reverses (goes negative) with a 1 rotor, like a
4cyl 4stroke piston engine. With a 2 rotor, the torque curve never
actually reverses so the gear box isn't stressed as much in the torsional
resonance dept. even though there's twice the power. If you frequent Paul
Lamar's list I'm sure he will be happy to show you the torque curve for
the 2rotor. IIRC, the torque curve for a 1rotor looks like a 4cyl piston
engine, going negative between each positive torque peak. If the system
resonates & you continue to excite it without damping the resonance, no
amount of strength will keep it from breaking.

The 1st incarnation of Powersport are the guys in the northwest with the
rotary powered RV-4 that had such horrendous torsional resonance problems
*on a dyno*. Current thinking is that they had a problem with resonance on
that particular dyno with that particular engine/dyno coupling (it was
built to test V-8's) They also had severe problems getting their P-port
engine to idle properly. Others have had no problem at all getting them to
idle smoothly. The developers had racing V-8 backgrounds & some of that
stuff doesn't transfer well to the rotary. Their internal tooth ring gear,
designed to keep the gearbox 'tight', like Rusty mentioned, is very heavy,
very expensive, & if it isn't heavy enough will actually loosen up as rpm
comes up & the ring gear tries to stretch. Kind of self-defeating. The
'tight' vs 'loose' issue is really an issue of moving resonant frequency
above the operating rpm range or moving it below the operating range.
'Tight' moves it up; 'loose' moves it down. Manual transmission cars are
'loose', moving resonance below normal operating rpm. You've probably
experienced the automotive version of torsional resonance if you've put a
manual trans car in 2nd or 3rd & let the idling engine try to pull the
car. If the engine continues to run, the car will move forward in big
surges. That's the resonant frequency of the drive train. I don't remember
Powersport ever having a problem with broken props or gearboxes; my memory
is that they went straight from their dyno problems to the big internal
spur gear. They did have a gearbox failure when competing in time-to-climb
at SNF because they were using nitrous & over stressed a bearing in the
gearbox. I think they were producing somewhere between 350-400hp (13B
without turbo) when that happened.

Damping torque pulses with belt slippage: inefficient & produces a lot of
heat.

I think Corky mentioned the nightmare of an intake manifold on RX-7 13B's;
fortunately a much simpler & lighter intake works fine for aircraft since
low rpm torque isn't needed.

Eggenfellner: I believe they've recently had the 1st failure of one of
their gearboxes.

Charlie
(Rusty's 'hangar away from home' for the next hurricane)

Gordon Arnaut wrote:
Charlie,

Thanks for the insightful post. What exactly is the source of the torsional
vibration in the rotary? And how does it compare in severity to a piston
engine?

Regards,

Gordon.


The source is the same as in a piston engine. The fuel/air mixture is
has to be compressed right before it is ignited. The ignited mixture
then only provides power for about 60 degrees of rotation.

The negative peek is very small in the two rotor for two reasons.
First, the 8 to 10 pound rotor plus very large cranshaft provides a
significant flywheel effect. And second, the negative pulse is slightly
overlapped by the postive pulse of the other rotor; eg. the pressure
graph is actually the result of the overlay of the graph from two
independant rotors.

--
This is by far the hardest lesson about freedom. It goes against
instinct, and morality, to just sit back and watch people make
mistakes. We want to help them, which means control them and their
decisions, but in doing so we actually hurt them (and ourselves)."

Charlie,

I did get some info from Paul Lamar on the rotary list, including some
graphs of torque peaks for the 2-rotor.

However, I just looked at some torque graphs of piston engines and even a
six-cylinder will go below the zero axis into negative territory -- although
the amplitude below zero is less than for a 4-cylinder piston.

Only when you get to an 8-cylinder piston does the wave stop going into
negative territory.

I'm not really clear from Paul's graph whether the 2-rotor goes negative or
not. I would think it does, since Paul describes it as similar to a
6-cylinder piston, in terms of torsional excitation.

It is clear of course that the torque pulses from the single-rotor will be
even more uneven and will go deeper into negative territory.

Still, as you pointed out, the cure to torsional excitation is to dampen it,
not to build a stronger transmission. It should also be noted that one way
to avoid torsional vibration is not to run the engine continuously at the
rpm where excitation occurs (as with the non-counterweighted Lycoming).

With a single-rotor engine, I would think that if you dampen the harmonics
at the offending rpm, you should not need a heavy gearbox designed for the
power of the bigger engine.

Also as far as poly-v belts are concerned, the perception is that they are
not as efficient as cog belts, but this just isn't so. A properly tensioned
poly-v belt is as efficient as a cog belt, between 95 and 98 percent.
(Goodyear and other belt makers have info on this on their websites).

This is as efficient as most gearboxes -- or even slightly better.

Also the system should be properly tensioned so there is no slippage at full
power. The only time slippage will occur is when torsional excitation causes
a big torque pulse. Of course, there is no reason to run the engine at that
rpm anyway (whatever it may be on a rotary, but probably below 2000 rpm).

As far as heat is concerned, both cog belts and poly-v belts generate
similar amounts of heat -- the two or three percent of power that is lost
goes to heat. So does a gearbox.

The belt manufacturers have made great strides in poly-v technology in
recent years. This is definitely not just a bunch of v-belts strung
together.

Regards,

Gordon.



"Charlie" wrote in message
.. .
Gordon Arnaut wrote:
Yes, the Egg redrive has no failures yet, from what I know.

However, he seems to have taken the "build it strong as hell" approach
and doesn't use any kind of damping, such as elastomers, sprag clutch,
etc. He is also using a heavy flywheel that helps to smooth out the
torque spikes.

The result is quite a heavy unit. Still his FWF package is competitive
with Lyc on a power-to-weight basis. Not bad at all.


Regards,

Gordon.


"Corky Scott" wrote in message
...

On Thu, 28 Jul 2005 22:08:10 -0400, "Gordon Arnaut"
wrote:


Your story is just another reminder that gearboxes are one of the big
bugaboos of any auto engine conversion -- and torsional vibration (or
resonance) is always the culprit. I know that in the Subaru community
there
is not really a box that I would consider completely trustworthy.

Really? Not even Eggenfellner's? I haven't heard of any failures of
his design yet, but I haven't been actively following Subaru
conversions.

Corky Scott



Sorry for coming in a little late on this; I usually frequent the
Flyrotary list & Rusty mentioned that this thread was alive over here. The
non-existent email address is there because I got tired of a steady diet
of spam.

Several things come to mind about the previous few messages in this
thread, from the stuff I've read in about 10 years of following
Powersport, then Tracy Crook's development trials & tribulations. This is
from memory & I never claim to have a good memory. :-)

Gearbox strength for 1rotor vs 2rotor: The big deal about a 1rotor is that
the torque curve actually reverses (goes negative) with a 1 rotor, like a
4cyl 4stroke piston engine. With a 2 rotor, the torque curve never
actually reverses so the gear box isn't stressed as much in the torsional
resonance dept. even though there's twice the power. If you frequent Paul
Lamar's list I'm sure he will be happy to show you the torque curve for
the 2rotor. IIRC, the torque curve for a 1rotor looks like a 4cyl piston
engine, going negative between each positive torque peak. If the system
resonates & you continue to excite it without damping the resonance, no
amount of strength will keep it from breaking.

The 1st incarnation of Powersport are the guys in the northwest with the
rotary powered RV-4 that had such horrendous torsional resonance problems
*on a dyno*. Current thinking is that they had a problem with resonance on
that particular dyno with that particular engine/dyno coupling (it was
built to test V-8's) They also had severe problems getting their P-port
engine to idle properly. Others have had no problem at all getting them to
idle smoothly. The developers had racing V-8 backgrounds & some of that
stuff doesn't transfer well to the rotary. Their internal tooth ring gear,
designed to keep the gearbox 'tight', like Rusty mentioned, is very heavy,
very expensive, & if it isn't heavy enough will actually loosen up as rpm
comes up & the ring gear tries to stretch. Kind of self-defeating. The
'tight' vs 'loose' issue is really an issue of moving resonant frequency
above the operating rpm range or moving it below the operating range.
'Tight' moves it up; 'loose' moves it down. Manual transmission cars are
'loose', moving resonance below normal operating rpm. You've probably
experienced the automotive version of torsional resonance if you've put a
manual trans car in 2nd or 3rd & let the idling engine try to pull the
car. If the engine continues to run, the car will move forward in big
surges. That's the resonant frequency of the drive train. I don't remember
Powersport ever having a problem with broken props or gearboxes; my memory
is that they went straight from their dyno problems to the big internal
spur gear. They did have a gearbox failure when competing in time-to-climb
at SNF because they were using nitrous & over stressed a bearing in the
gearbox. I think they were producing somewhere between 350-400hp (13B
without turbo) when that happened.

Damping torque pulses with belt slippage: inefficient & produces a lot of
heat.

I think Corky mentioned the nightmare of an intake manifold on RX-7 13B's;
fortunately a much simpler & lighter intake works fine for aircraft since
low rpm torque isn't needed.

Eggenfellner: I believe they've recently had the 1st failure of one of
their gearboxes.

Charlie
(Rusty's 'hangar away from home' for the next hurricane)

Gordon Arnaut wrote:

Still, as you pointed out, the cure to torsional excitation is to dampen it,
not to build a stronger transmission. It should also be noted that one way
to avoid torsional vibration is not to run the engine continuously at the
rpm where excitation occurs (as with the non-counterweighted Lycoming).

No. The only cure for torsional excitation is the move the system's
harmonic into a range that will only be seen in low power operations and
then won't be used for very long. Tracy's PSRU moved the excitation
range down to between 500 and 800 rpm (working off the top of my head
here, so the numbers may be off). A well tuned rotary will idle in the
900-1000rpm range (prop at about 1/3rd of that).

Dampening doesn't exist. Elasticity in the system may shorten the
peaks, but you'll be left with a fatter mountain. The energy has to go
somewhere. The problem is that you'll still be hitting the shaft at
it's resonant frequency, causing a vibration in it. Each hit adds a
little to the vibration. If you repeatedly hit anything at it's
resonant frequency, each hit will add to the system vibration until it
either wears out very quickly or comes apart catastrophically. That
applies to reduction units, crankshaft, and control surface skins (yes,
flutter is a form of resonant vibration). Adding some elasticity to the
shaft may make the gearbox last longer, but you won't be able to do much
in those few seconds 8*)

Here's someone who says it much better than I:

http://rotaryaviation.com/PSRU%20Zen%20Part%202.html


--
This is by far the hardest lesson about freedom. It goes against
instinct, and morality, to just sit back and watch people make
mistakes. We want to help them, which means control them and their
decisions, but in doing so we actually hurt them (and ourselves)."

Ernest,

You are right that springs and elastomers do not technically dampen kinetic
energy, they simply store and relase it at a later time (how much later
depends on the frequency at which it is tuned).

However, both springs and elastomers can achieve our objective of clipping
of destructive harmonics if they are tuned to the resonant frequency of the
object that we want to protect from resonance.

Also, fluid dampers do exist for both engines and propellers which actually
do dampen kinetic forces by turning them into heat, just like the shock
absorber on your car.

Drive belts do this as well, including the synchronous (cog) belts which
don't actually slip but which do absorb some energy as heat.

In fact, even a propeller can be tuned efectively to do this, because it is
kind of like a spring -- it flexes at its tips with the power pulses and
then whipsaws back during the lulls. A propeller thus tuned can do the same
thing as a harmonic balancer on an engine.

Of course the trick is in the tuning. More often we are concerned about the
opposite, where they propeller may in fact add to the excitations of the
crankshaft at its resonant frequency. We see this in non-counterweighted
Lycoming four-cylinder engines, which have operating restrictions in certain
rpm ranges.

So when we talk about torsional resonance issues, there are two aspects: one
is the engine itself, and specifically the crankshaft, which shoulders the
torsional forces. The other aspect is what the engine is connected to,
whether a propeller, a gearbox, or both.

We know of course that each object has a natural frequency at which it
resonates -- the point at which external excitations cause harmonics to
build. The object's mass, stiffness and damping determine how it will
respond to an excitation force.

Obviously having smoother excitations -- such as cylinders or rotors firing
at more frequent intervals -- will result in less extreme instantaneous
torque excursions, and the engine will run smoother. Dow that mean it will
be free of torsional vibration? No. Even an electric motor has torsional
vibration because the magnets arranged around its circumference which propel
the rotor are like firing pulses -- a motor with 12 magnets will be smoother
than a motor with two magnets.

However the engine can be as smooth as butter, but if the shaft on which it
is riding is very thin, it will still flex -- and at some point will reach
its resonant frequency, and if left there for the harmonics to build up --
will break.

So that's where mass and stiffness comes into play. Look at tuning forks,
the smaller ones will vibrate at a higher frequency. Even a small excitation
will cause a lot of flex in the thin tuning fork.

It's the same with crankshafts or other shafts -- such as those in a
gearbox, or a propeller -- which are exposed to torsional forces.

For example, most V-8 engines come with a harmonic balancer, even though
they have four power pulses for each crankshaft rotation. That's because
there is enough flex in the crankshaft that the crank can begin to resonate
at some rpm within the operational range.

But look at the four-cylinder Subaru opposed engine. It has never needed a
harmonic balancer. Why? It's crankshaft is quite massive and very stiff, so
its resonant frequency is below the engine's operating range. (It is stiff
because it has five main bearings, so each crank throw is supported by two
main bearings, one on each side; plus the crank journals are quite massive).

Although I'm not yet fully up to speed on the rotary, it is quite clear that
it is similar to the Subaru in that it does not need a torsional dampening
device. That e-shaft is quite rigid and it's throw is short enough that it
will not flex the way a crankshaft will flex. So more stiffness equals a
lower resonating frequency.

So why be concerned with torsional vibration? It's because we are putting a
gearbox and propeller on this engine. Just because the rotary does not
resonate itself does not mean it won't set a gear shaft into resonance.
That's the point of the damping at the engine-gearbox coupling.

When I asked what the source was of torsional issues with the rotary, I did
not have a full grasp of the dynamics with the rotary engine -- until I
watched some helpful animations.

However, as you can see, torsional vibration is not strictly a function of
power pulses. It is a function of the many things inside the engine --
including moments of inertia, stiffness of torque shafts, etc -- plus just
as many variable in whatever it is the engine may be connected to.

Regards,

Gordon.




"Ernest Christley" wrote in message
. com...
Gordon Arnaut wrote:

Still, as you pointed out, the cure to torsional excitation is to dampen
it, not to build a stronger transmission. It should also be noted that
one way to avoid torsional vibration is not to run the engine
continuously at the rpm where excitation occurs (as with the
non-counterweighted Lycoming).

No. The only cure for torsional excitation is the move the system's
harmonic into a range that will only be seen in low power operations and
then won't be used for very long. Tracy's PSRU moved the excitation range
down to between 500 and 800 rpm (working off the top of my head here, so
the numbers may be off). A well tuned rotary will idle in the 900-1000rpm
range (prop at about 1/3rd of that).

Dampening doesn't exist. Elasticity in the system may shorten the peaks,
but you'll be left with a fatter mountain. The energy has to go
somewhere. The problem is that you'll still be hitting the shaft at it's
resonant frequency, causing a vibration in it. Each hit adds a little to
the vibration. If you repeatedly hit anything at it's resonant frequency,
each hit will add to the system vibration until it either wears out very
quickly or comes apart catastrophically. That applies to reduction units,
crankshaft, and control surface skins (yes, flutter is a form of resonant
vibration). Adding some elasticity to the shaft may make the gearbox last
longer, but you won't be able to do much in those few seconds 8*)

Here's someone who says it much better than I:

http://rotaryaviation.com/PSRU%20Zen%20Part%202.html


--
This is by far the hardest lesson about freedom. It goes against
instinct, and morality, to just sit back and watch people make
mistakes. We want to help them, which means control them and their
decisions, but in doing so we actually hurt them (and ourselves)."