| Wow, what a busy
month it's been; I better prescribe myself a pick-up. Ah,
that's better, good ol' VB. There is almost too much to report.
My new F5B (that's a weird R/C electric class using folding
props and 1600 watt motors) prop threw a blade on launch.
The poor guy holding it copped blood blisters under every
fingernail as a result: hope he doesn't sue. My new Unlimited
scale 23X30 prop hit over 220 MPH: landed with one blade ready
to drop off. Perhaps I should write about prop failure modes.
I would too, except somebody suggested my super-thin
F2A prop was too thin and likely to flutter. Hell, he could
be right: somebody has to be the bunny, I don't own a $900
FAI motor to do these tests. You, dear readers, are my unwitting
test pilots. And, like Uncle Sam, I reckon test pilots are
expendable. So what is the story with flutter? Well, if you
fly F2A and haven't had prop flutter, you are just not trying.
Tilleys' law states that "the thinner the prop tips, the faster
the airplane". Hell, he's a big tough guy, I'm not going to
argue with that! And my props are very thin at the tips, you
can depend on that, too. Such props are very likely to flutter
while they are run up on the deck. The phenomenon is "stall
flutter": you'll know if you've got it, the noise is like
banging a Scotsman around in a 44-gallon drum. But in F2A,
the RPM on the deck are usually low, as the motor does not
get on pipe until the model is airborne, a condition where
stall flutter is unlikely to occur. Still, you can get flutter
when airborne due to vibrations in the prop, and vibrations
coupled into the prop from the engine. These latter are unpredictable,
you just gotta get out there and fly. However, there is an
interesting equation for flutter due to propeller properties
alone. The maximum flutter speed Vf is given by:
Vf = K * c/R * SQR((t/c) ^ 3 * G / (p * (Xcg-.25))
Here, K is a constant dependent on blade plan-form (taper,
tip rake etc.), c is mean chord, R is prop radius, G is shear
modulus of prop material, p is air density, and Xcg is the
X abcissa of the c/g position expressed as a fraction of the
chord.
Now don't be put off by this equation: it's
quite useless to you, so I won't be testing you. However,
it says some very interesting things.
* If you increase the chord, you raise the flutter speed
* If you thicken the prop, you raise the flutter speed
* If you toughen up the material, you raise the flutter speed
* If you mass balance the prop, you raise the flutter speed.
Huh, what was that last one? That last term (Xcg-.25) is very
interesting indeed. It says that if the c/g of the propeller
airfoil is at the quarter chord line, then the prop won't
flutter! Hey, that's radical! This is why helicopter rotor
blades have a very heavy leading edge: they are made so that
they balance at the quarter chord position. With such narrow
chords, rotors would have a very low flutter speed if it were
not for this feature. The same applies to wings, when the
condition is referred to as "mass balance". Some wings, and
many control surfaces, are designed with this in mind.
Indeed, my light, hollow, carbon 32X12 aerobatic
props are moulded with massive fibreglass leading edges, rather
than with spars. With the mass balance condition met, very
little material is required in the propeller skin.
Unfortunately, F2A props are too thin to make hollow. Xcg
for F2Aairfoils lies at about .4 (40% of chord back from the
leading edge), so they are not mass balanced: hence the concern
about flutter. So what to do?
We can't increase the chord, since then the
prop is then excess load for the engine.
We can increase t/c, but not at the tip. The
thickness-to-chord ratio up to 3/4 radius can be as much as
13%, but this must drop to 4% at the extreme tip if thrust
losses due to shock-waves are to be avoided. So what is left?
The thing we haven't said is that the flutter is primarily
due to torsion: that is, twisting of the blade (not bending).
The parameter G, in the equation above, really refers to how
stiff is the propeller material. In the case of the carbon
fibre F2A prop, stiffness depends heavily on the carbon-to-resin
ratio, and the orientation of the carbon fibres. Careful lay-up
to minimise resin content is essential.
Diagonal-weave carbon cloth, laid in the prop
surface, front and rear, places the fibres directly in the
torsion field. This greatly increases torsional strength,
at the expense of bending stiffness. This is the same principle
as geodetic structure in wings (Vickers-Armstrong Wellington
bomber, for example).
That leaves parameter K, the blade plan-form parameter. If
you make a prop with swept forward tips, it will flutter horribly:
that is, it is flutter divergent. If, on the other hand, the
tips are raked back, the effect is to dampen out the flutter.
So if you are mad enough to try Supercool Kerr03 F2A, you
will find it is really very thin, and very bendy. But just
try twisting it. It also has the raked anti-flutter tip. The
main worry with this prop is that it is thin all the way to
the root. It could fail in torque or centrifugal modes, but
who knows without running it?
That leaves F2ACW001 F2A prop. It is the same
as Kerr03, but thickens up considerably inboard from the 80%
radius. This, plus its dedicated counterweight feature make
it the prop of choice at this time. At least until you bunnies
out there get hopping!
To return to stall flutter for a moment: if
you get flutter on the ground with one of my C/F props, just
ignore it and release the model anyway. The flutter will cease
when the airplane gains some speed. The props are very tough
and will not fail due to this flutter.
You may be aware that aluminium props are banned
from use on models. Aluminium is not so tough. To finish,
I quote from Den Hartog (1934, Mechanical Vibrations).
"Since the introduction of aluminium-alloy
propellers in airplanes, a number of fatigue failures have
occurred. Some of these were noted in time to avoid
failure, being seen in the form of cracks, but in other cases
either an entire
blade or the tip of a blade has blown away in mid-air. The
fact that these
failure were unmistakably due to fatigue makes it certain
that they were caused by vibration." |
