Mini-HowTo Elevator on top or below tailboom?

[martin kopplow]

It has been discussed, where to best put the elevator on a DLG: On top or below the tailboom. I always was a supporter of the 'on top' variant, but what do we know? There were theories.

Today, we made a little experiment.
We fixed a lenght of fluffy wool to the TE of a DLG, about 10cm long, extended by a thin thread that goes all the way to the elevator. There, a vertical scale was attached, with the extended chord line of the wing marked as a reference. A miniature camera was mounted at one wingtip and the plane flown a few times at a local slope, going through all the usual flap and speed configurations.

The video can be see here:

<http://www.youtube.com/watch?v=dAbloetwppU>

It comes clear that the elevator should be mounted on top (Or at least reasonably high above the extended chord line of the wing).


Best regards,
Martin Kopplow

[markdrela]

Pretty cool.

Originally Posted by martin kopplow

It comes clear that the elevator should be mounted on top (Or at least reasonably high above the extended chord line of the wing).

If you reflex the flaps so that the rear bottom of the wing is flat, this flat surface then makes a very convenient reference line which can be extended with a long ruler. Where does this bottom-surface extension fall on your wake-height scale?

[martin kopplow]

Originally Posted by markdrela

Where does this bottom-surface extension fall on your wake-height scale?

Mark,

I removed the height scale this morning, but when I put a long ruler under the wing as you requested, I get a point 25mm below the elevator LE, which is 5mm higher than the chord mark I used in the experiment.

I was surprised the wake was so high when circling. This sequence is unfortunalely rather short, because the camera battery failed after this take, but it is obvious that in this situation, the thread is rather unsteady, and even appears to be in a curve, pointing upwards at the end. I was expecting this would happen, but not as evident.

The elevator surface of my own DLGs is probably not high enough to escape the wake in this situation, given the dispersion or spread it exhibits. Again, one lesson learned.

Martin

[RCPC]

I agree, this is pretty darn cool. your experiment is quite interesting and is a good demonstration of downwash angle as a function of cl.

what are the fundamental disadvantage of having the elevator in the wake of the wing? i could see that the turbulance created by the wing in landing mode might be advantagous so as to not stall the elevator at low airspeeds? is my reasoning completely erroneous?

once again, very nice experiment. i wish i saw this a while ago when i had a powered plane with a T-fin/T-canaliser (the mini-oxalys). i would have loved to do some experiments to see if it really did clean up the airflow over the rudder. i still have a couple other ideas. you have inspired me.

thanks for sharing,
paul

[martin kopplow]

Originally Posted by RCPC

what are the fundamental disadvantage of having the elevator in the wake of the wing?

I see two reasons not to have the elevator in the wake of the wing. The one is as you expected: More predictable control and trim. The other one is to improve performance by ensuring a clean air flow around the surface.

Martin

[markdrela]

Originally Posted by martin kopplow

I was surprised the wake was so high when circling. This sequence is unfortunalely rather short, because the camera battery failed after this take, but it is obvious that in this situation, the thread is rather unsteady, and even appears to be in a curve, pointing upwards at the end. I was expecting this would happen, but not as evident.

Without any wing downwash, the thread would follow the curved path of the glider in a turn, which curves "upward" as seen by the plane if the airplane is banked. At a turn radius R , the upward wake deflection at a distance d behind the wing is
delta(z) = 0.5 * (d^2 / R) * sin(bank) = 0.5 * (g d^2 / V^2) * tan(bank) * sin(bank)
Using typical numbers for a slow, tight turn, I get
g = 9.8 m/s^2
d = 0.65 m
V = 5.5 m/s
bank = 30 deg

delta(z) = 0.02m = 2 cm

Not huge, but not neglible, either.