SMALL - espritmodel.com SMALL - Telemetry SMALL - Radio
Reply
Thread Tools
Old Jul 01, 2013, 11:32 PM
Registered User
The Willamette Valley, Oregon
Joined Dec 2008
1,234 Posts
Good points Bruce; I am definitely going to fly with a camera looking at the wing. Whatever the results, I feel like I have learned at least one or two significant things from this conversation... Steve
aeronaut999 is offline Find More Posts by aeronaut999
Reply With Quote
Sign up now
to remove ads between posts
Old Jul 01, 2013, 11:46 PM
Registered User
The Willamette Valley, Oregon
Joined Dec 2008
1,234 Posts
Quote:
Originally Posted by ShoeDLG View Post
This is known as "apparent dihedral effect" because you can't isolate the effects due to sideslip from the effects due to rudder deflection. I have been improperly using the terms interchangeably.
Well... you could deploy a drogue chute from one wingtip, as an alternative way to yaw the aircraft!

I did that with the hang glider... but not for fear that a rudder would be so high above the cg that it would create a strong direct roll torque... rather because a keel-mounted rudder would shift the keel to the side and thus change the shape of the airfoil on each wing...

Talk about effects due to flexing...

Unfortunately my drogue chute design was not stable enough, it tended to orbit in a large circle... passing in and out of the tip vortex no doubt...

I'm having trouble visualizing exactly around what axis the aircraft rolls. For example, if the rudder is tall, but the aircraft is flying at a very high angle-of-attack, so that the rudder's center of area is not higher or lower than the cg, in horizontal flight, is the rudder therefore not creating any direct roll torque?

Steve
aeronaut999 is offline Find More Posts by aeronaut999
Reply With Quote
Old Jul 02, 2013, 06:00 AM
Registered User
ShoeDLG's Avatar
Germany, BW, Stuttgart
Joined Mar 2012
864 Posts
Quote:
Originally Posted by aeronaut999 View Post
I'm having trouble visualizing exactly around what axis the aircraft rolls.
It may help to put your hand (or a model) up to represent an aircraft in level flight at about 30 degrees (or more) AOA.

If you roll the aircraft 90 degrees about its x (nose-to-tail) axis, the nose will still be pitched up 30 degrees, but the roll will have converted all of the AOA into sideslip. If the aircraft has positive dihedral effect it is likely that opposing torque would have stopped the roll (even with full aileron deflection) long before getting to 90 degrees.

If you were able to continue the roll through 90 degrees bank angle, your upper foot would be on the floor (to keep the nose up), but you’d be deflecting the stick full downward. In other words, performing a roll about the aircraft x axis starting from significant AOA would require almost full cross control inputs at certain times… not something you’re likely to do unless you have copious roll authority and a tolerance for discomfort (fast aileron rolls can be bad enough without high-amplitude back and forth side force excursions).

If you start from level flight at 30 degrees AOA and instead roll the aircraft 90 degrees about its velocity vector, the nose will end up on the horizon (pointed 30 degrees from the direction of travel), and you will have maintained 30 degrees of AOA and zero sideslip. Rolls about the velocity vector can be sustained without encountering an opposing torque due to dihedral effect. You just need to keep the ball (or yaw string) centered throughout the roll.

Which way do airplanes roll? It depends, but rolls about the velocity vector are likely the case unless you’re performing some unnatural aerobatic maneuver.

Airplanes that fly at large angles of attack can be departure prone if sideslip is allowed to build up (aerodynamic moments will often exceed control authority at high sideslip). For this reason, the flight control systems are designed to keep the sideslip close to zero (through both sideslip and rate of sideslip feedback). Something that was interesting to do in an F-18 was to smoothly apply full aft stick from level flight. Depending on initial airspeed, you could be at more than 40 degrees AOA by the time you reached 60 degrees nose-up (with your flight path about 20 degrees up). If you then applied full lateral stick and rudder in the same direction (while holding full aft stick), the nose would essentially describe a 45 degree cone around your direction of travel (the nose would be about 40 degrees below the horizon as you passed through inverted even though your flight path was about 5 degrees up). If you did it right, you’d finish the roll in a few seconds with the nose 45 degrees up having lost very little (if any) altitude. The flight control system did all the work to keep the sideslip close to zero. All the rudder input did was command additional roll rate about the velocity vector.

You could do something similar at the top of an Immelman… You could pull to full aft stick as the nose approached the horizon and apply full lateral stick and rudder in the same direction. The aircraft would roll about the velocity vector (which was about 45 degrees up as you started), and you’d consistently end up upright and about 50 degrees nose-high. This could be disorienting at first because the airplane’s initial response to the lateral stick input did not appear to be a roll, but pure sideways nose movement. It took a while for it to register that (for a zero sideslip roll) this motion actually corresponded to what you had commanded (and you were not causing the airplane to depart). Someone seeing this from the back seat for the first time would usually offer: “whoa, what just happened?”

Bottom line: most full-scale flight control systems designed to operate at high AOA won't let you roll around the aircraft x axis... they will force you to roll around the velocity vector. Things are different for RC (or full-scale) aerobatic airplanes:

-These planes have significant top-bottom symmetry, meaning they are unlikely to experience much dihedral effect (even at high AOAs and sidelsip anlges)... This is evidenced by the fact that you can sustain level knife-edge flight without holding much lateral stick.

-The moments of inertia (particularly for RC) are such that you can transition through high-sideslip flight regimes without developing angular momentum that can't be quickly overcome/nulled by the controlled application of aerodynamic forces. In a full-scale fighter, brief excursions to high sideslip angles are likely to generate angular momentum that can't be quickly checked by aerodynamic forces (meaning you're likely to end up out of control).
ShoeDLG is online now Find More Posts by ShoeDLG
Last edited by ShoeDLG; Jul 02, 2013 at 12:44 PM.
Reply With Quote
Old Jul 04, 2013, 01:13 AM
Registered User
The Willamette Valley, Oregon
Joined Dec 2008
1,234 Posts
Quote:
Originally Posted by aeronaut999 View Post
Good points Bruce; I am definitely going to fly with a camera looking at the wing. Whatever the results, I feel like I have learned at least one or two significant things from this conversation... Steve
PS- it's just that I can't see any reason that the wing would be more bent away from the usual positively-loaded shape when flying at high negative angles-of-attack, than when flying at low negative angles-of-attack. (-1G loading in either case.) If anything I would expect the opposite- the tips should bear more of the (negative) load at the less strongly negative a-o-a, when the central sections may be nearly unloaded, generating near zero lift. (Consider washout.) That's why I don't think my results can be explained entirely through bending. But, it certainly deserves a look with the camera...

Steve
aeronaut999 is offline Find More Posts by aeronaut999
Reply With Quote
Old Jul 04, 2013, 11:30 AM
B for Bruce
BMatthews's Avatar
The 'Wack, BC, Canada
Joined Oct 2002
11,607 Posts
Big changes to how the model acts can come about with one or two degrees of deflection. If the camera trick is going to tell you anything I'd suggest that it needs to be used in conjunction with some indicator of some sort that makes it easy to see small angular changes. It may be something as simple as two wires poked into the foam at the root and tip that line up in the camera view but which can be analyzed for bending deflection and angular flexing later on in still captures.
BMatthews is online now Find More Posts by BMatthews
Reply With Quote
Old Jul 04, 2013, 02:22 PM
Registered User
ShoeDLG's Avatar
Germany, BW, Stuttgart
Joined Mar 2012
864 Posts
This is a video of the Immelman with a loaded roll at the top that I tried to describe in post #48. Unfortunately the airplane vibrates quite a bit at high AOA and I hadn't figured out damping for the camera support yet... It's looking backwards, but you can see the yaw kick in as the roll to upright starts.

Roll (0 min 23 sec)


Some (jittery) high AOA video from the same flight (with good condensation in the LEX vortices).

High AOA (1 min 21 sec)
ShoeDLG is online now Find More Posts by ShoeDLG
Last edited by ShoeDLG; Jul 04, 2013 at 04:31 PM.
Reply With Quote
Old Jul 10, 2013, 03:35 PM
Registered User
The Willamette Valley, Oregon
Joined Dec 2008
1,234 Posts
C - 17

Quote:
Originally Posted by ShoeDLG View Post
This is known as "apparent dihedral effect" because you can't isolate the effects due to sideslip from the effects due to rudder deflection. I have been improperly using the terms interchangeably. As you point out, the "dihedral effect" can be different in magnitude (and even sign) from the "apparent dihedral effect".
Along those lines-- I just "flew" the C-17 on RealFlight 6.5 -- killing one or both engines one one side and observing the aileron input needed to prevent roll in straight-line (banked) flight with no rudder deflection-- the "effective dihedral" is mildly positive (at least at the airspeeds / a-o-a's I was using) -- yet the plane shows a strong "wrong-way" roll response to rudder inputs-- surely due to the rudder acting so high above the CG--

This makes it REALLY interesting to try to steer the aircraft in a straight line with rudder alone, leaving the ailerons neutralized -- it can be done-- to stop a turn, move rudder stick toward direction of the unwanted turn and wait for roll response-- but it's "interesting". Before the roll response, there's an initial increase in the rate of the unwanted turn, due no doubt to sideforce from the fuselage...

(Just to be clear, none of this has much to with what I was originally describing with the Radian but... interesting to talk about nonetheless... )

Steve
aeronaut999 is offline Find More Posts by aeronaut999
Last edited by aeronaut999; Jul 10, 2013 at 03:51 PM.
Reply With Quote
Old Jan 27, 2014, 06:45 PM
Registered User
The Willamette Valley, Oregon
Joined Dec 2008
1,234 Posts
New observations

New observations-- I flew the plane with an onboard camera-- changes in twist are hard to evaluate-- changes in G-load (lift force) produce obvious changes in the wing's dihedral geometry-- the wing certainly has less dihedral in inverted flight than upright.

It wasn't really feasible to try to do tests at a constant altitude, which would have eliminated one variable. The aircraft becomes EXTREMELY roll-unstable in climbing inverted flight, and is rather roll-unstable in level inverted flight, so I decided to leave the power off for all the tests.

The NET lift force from the wings, either upward or downward (in aircraft's reference frame), is obviously less in a dive than in level flight.

In inverted flight in a rather steep power-off dive (maybe 35 degrees), the dihedral angle looked about the same as when the aircraft is at rest (being held in the hand by the fuselage) in a level, upright attitude. In inverted power-off slow flight, where the glide path was flatter (but still at least 10 degrees or more from horizontal), the wing was slightly flattened-- the dihedral angle was a bit less than described above. In all upright flight modes, the dihedral angle was considerably more than described above.

So the wing is SLIGHTLY flatter in inverted slow gliding flight than in inverted higher-speed diving flight. That would certainly contribute to the observed "backwards" rudder-roll coupling, or at least to a reduction in "normal" rudder-roll coupling. Whether the aircraft has enough sweep in the wingtips to produce the observed "backwards" rudder-roll coupling when the wing is "flattened" in inverted slow gliding flight, is another question.

Note though that as the total lift force increases (approaches the aircraft weight) as the glide path is flattened, this alone will cause the swept tips to contribute a greater "backwards" rudder-roll coupling effect, even if the wing shape stayed constant. Conversely as the total lift force decreases (becomes much less than aircraft weight) as the glide path becomes steeper, this alone will cause the swept tips to contribute less "backwards" rudder-roll coupling effect, even if the wing shape stayed constant. In a vertical dive (zero lift), to a first approximation the swept tips will contribute zero rudder-roll coupling. (Actually, due to washout, the tips will be slightly downlifting as seen from the cockpit, while the root is slightly lifting, so there still will be some backwards rudder-roll coupling contributed by the tips.)

Are all these effects too negligible to produce the observed behavior? Are the tips actually stalling during inverted slow flight? Possibly. Note however that I could reproduce these results even when I slowly fed in the rudder inputs, producing only a low rate of change in slip angle and a low yaw rate, rather than slamming the rudder to one side to produce a high rate of change in slip angle and a high yaw rate. I'm pretty convinced that what we're seeing here is NOT primarily roll-due-to-yaw-rate.

Still, when you are flying inverted with a wing with dihedral, maintaining a steady slip angle with the rudder will INCREASE the (negative) angle-of-attack of the wingtip toward which the rudder is deflected, which could cause a stall of that tip, followed by a roll toward that tip. This is a "backwards" roll response-- we want the aircraft to roll AWAY from the deflected rudder, as viewed from the ground . In other words, a steady rudder deflection causing a steady slip angle creates SIMILAR tip-stalling tendencies as does an abrupt rudder input causing a high, temporary yaw rate, and both of these effects act OPPOSITE to the normal rudder-roll coupling created by dihedral. When we are flying UPRIGHT in an aircraft with dihedral, this is no longer true-- an abupt left rudder-stick input might tend to slow down the left wingtip causing it to tip-stall, yet a steady left rudder deflection creating a strong slip angle but little yaw rate, would tend to increase the angle-of-attack of the RIGHT wingtip and certainly could not tip-stall the left wingtip.

Summary-- tip-stall due to high slip angle (induced/ with rudder), if such a thing is possible, always tends to create a "backwards" roll effect. Tip-stall due to high yaw rate (induced with rudder) tends to create a "normal" roll effect in upright flight and a "backwards" roll effect in inverted flight.

So the observation that the "backwards" effect exists even when I gradually move the rudder to full deflection (keeping the yaw rate low) rather than snapping it abruptly to full deflection (giving a high yaw rate) doesn't really rule out tip-stall, I guess.

Anyway the main new data point that has been added here, is the observation that during high-speed inverted flight, we aren't seeing the wing being forced into a configuration where it has less dihedral than in low-speed inverted flight, as I thought might be happening (due to loads imposed by washout.) If this happened this would tend to produce an effect opposite to the rudder-roll coupling I actually observed, which would suggest that something else (swept tips, or ??) was creating a VERY strong "backwards" rudder-roll coupling in inverted flight at low airspeeds/ high (negative) angles-of-attack.

See however post #4 in this thread-- I must have been on more "on my toes" that day than in the more recent tests, as I describe exploring the glider's behavior at a wide range of power settings including (inverted) climbing flight, and always seeing the same "backwards" rudder-roll coupling at low airspeeds. Climbing under power will "unload" the wing to some degree, just as diving does. This is inevitably going to cause the dihedral angle to increase just a bit-- and yet the effect remains. I guess I need to go for another filming session, to see if there is any noticeable difference in dihedral angle between the climbing, low-airspeed case (where I did see the "backwards" rudder-roll coupling) and the diving, high-airspeed case (where I did not see the "backwards" rudder-roll coupling).

At any rate the observed changes in dihedral angle between the inverted low-speed glide and the inverted high-speed dive were small. Like maybe around a 1-inch change in the height of the wingtip, at most. (I need to film a ruler at the wingtip to come up with a more definite number.) And certainly in NO case was the wing forced into an actual anhedral shape.

So at this time I don't think a change in wing shape is what is causing this "backwards" rudder-roll coupling in low-speed inverted flight. It's got to be something like tip stall, or the effect of the swept wingtips, or a combination of both.

PS if you haven't been following this thread which has been dormant for a while, you might want to take a look at post #34...

PPS if "tip stall" is what is going on here, shouldn't the effect be even more pronounced on no-aileron version of the Radian, which has a great deal more dihedral at the tips, so a rudder input while inverted will produce a much greater increase in angle-of-attack of the wingtip toward which the rudder is deflected? Maybe there's no point in comparing a wing with smooth upswept tips with a wing with a sharp dihedral break, but... just a thought...

Steve
aeronaut999 is offline Find More Posts by aeronaut999
Last edited by aeronaut999; Jan 28, 2014 at 12:57 PM.
Reply With Quote
Old Jan 29, 2014, 09:09 PM
B for Bruce
BMatthews's Avatar
The 'Wack, BC, Canada
Joined Oct 2002
11,607 Posts
Well, when inverted washout becomes washin. So that also adds to the confusion.

It's a rare bird that flies polyhedral models upside down for long. I've only done it occasionally myself. If it did have any odd non linear issues of this sort I can't say I ever noticed. Both the models pretty much just worked as if they had ailerons. But I can't say I ever made a serious attempt to slow down to a slow inverted speed.
BMatthews is online now Find More Posts by BMatthews
Reply With Quote
Reply


Thread Tools

Similar Threads
Category Thread Thread Starter Forum Replies Last Post
Sold Parkzone Brushless Motor PKZ4716 for Radian & Radian Pro BRAND NEW! Paul1PA Aircraft - Sailplanes (FS/W) 3 May 23, 2013 11:03 AM
Sold Parkzone Brushless Motor PKZ4716 for Radian & Radian Pro BRAND NEW! Paul1PA Aircraft - Electric - Power Systems (FS/W) 2 May 22, 2013 05:59 PM
New Product CXN - Radian / Radian Pro Pusher Motor Mount...Version 2 crd FPV Talk 0 May 08, 2013 12:48 PM
For Sale Radian PRO wings/rudder/elevon/servos TRADE OR SELL flynfunny Aircraft - Electric - Airplanes (FS/W) 17 Dec 26, 2011 07:53 PM
Question Repairing a Radian Pro rudder.... NorCal2500HD Electric Sailplanes 3 Jul 20, 2011 06:25 AM