Low G Push over illustrated
Thread Starter
Passion Flying Hobby Science Sponsor Work
Joined: Apr 2004
Posts: 461
Likes: 0
From: Belgium
Low G Push over illustrated
Disclaimer
----------
The results presented here are obtained via a complex mathematical simulator developped by myself.
The results in itself are not new, other people have explained the underlying principles.
What I think is new that they are confirmed by a full physical dynamic model with a target precision between 0.5 and 2%.
Although I performed numerous tests to check the validity, the model, nor the results have been checked by professional third parties including official instances. All technical and mechanical data has not been provided for by the constructor but result from own measurements.
The results should not be used to develop pilot instructions different from the instructions published in the official manuals and safety notices.
The results may be used by instructors to show the danger of push overs. This is one of the motivations of the simulator, namely avoiding showing it in real, something that in this particular case has rightfully been counter advised by official instructions.
Simulated flight profile
------------------------
- R44, two front pilots each 80kg, 140 lit of avgass, ISO
- initial level flight 90 knts
- pull up with constant collective to 80 knts, 1900 fpmin
- push over with constant collective
- left cyclic to counter the right roll
During the simulation a tail rotor governor was used assisting the pilot.
(flying a simulator is far from easy...) It is my believe that this governor actually
reduced the impact, because it reacts faster in adapting tail rotor trust than an average pilot would do in this scenario avoiding extra side-slip effects.
The following film summarizes the flight:
(quicktime 350 kb)
http://www.e-sign.be/private/heli/R44_FlightFilm06.mov
(media player 350 kb, lower resolution)
http://www.e-sign.be/private/heli/R4...ilm06_Fast.avi
The controls over time (playing with the decimal system to have same y-axis, sorry USA)
remark the desparate left cyclic at the and that does not seem to stop the rolling.
Discussion
----------
A push over starts with bringing the cyclic forward which provoques a forward pitching.
This pitching reduces rapidly the aerodynamic loading of the main rotor disk. The airflow comes much more from above, reducing the effective aerodynamic angles of the rotor blades
even though collective is kept steady.
In the scenario the level trust is about 9300 N, going over 10000N in the pull up
and dropping rapidly to 4000N during the push over. During that same periode
main rotor SHP goes from 110 down to 100 and next up to 120shp. (this does not include mechanical losses nor tail rotor power)
In level flight appropriate anti torque trust is provided by the tail rotor and equally important by the assymetric tail fin.
In the scenario this is approximately 2050 Nm in total.
These moments are evened out with the main rotor torque to provide yaw-stability.
Equally important is roll-stability.
The tail rotor trust and the tail fin trust create a positive roll moment
(positive = clockwise or to the right for a forward facing pilot) about the center of gravity that is much lower than centres of these forces (by more than one meter)
In level flight this rolling moment of approximately 250 Nm is countered
by a small left cyclic of about 0.7?, the main rotor being above the centre of gravity by more than 2 meters.
For the sake of completenes: one minor extra positive roll moment is produced by the down force of the assymetric tail wing, during pitching this increases (50 Nm).
Some frames from the film (apparently some technical problem with the colors..)
published in case download of the film is too long
Start of push over and roll
Because of the rapid drop of main rotor trust shortly after the start of the push over two desequilibriums come up.
Clockwise roll moment
The drop in main rotor trust also takes aways the left component and the resulting negative rolling moment. This produces a clockwise rolling. This seems to be the most important factor.
Yaw moment
The change in the air flow of the main rotor influences the main rotor yaw moment. In general a small increase in SHP is withnessed in this case. At speeds above 80 knots the tail fin will be fastest to react by increasing its moment (and doing so increasing slightly the roll moment)
A right roll in forward flight is instinctively compensated by left cyclic action. Because of the reduced trust the left cyclic does however not produce enough anti roll moment to stop the
rolling. At the end of the scenario with left cyclic at more than 7 degrees, lateral flapping of the hub approaches 10 degrees and mast bumping is in the make.... (to avoid sensationalism I stopped the flight before that)
Or viewed from the outside (in plot the heli is rolled back to level to better assess rotor angles)
From different simulation attempts it appeared to me that aggravating factors are apparently small inbalances prior to the low-G: such as slightly off cyclics and small prior rolling rates.
Some engineerings plots at the end of the flight:
Flapping angle of the hub starts to approach 10 degrees (blade flap is higher because of coning)
Remark the very high mechanical angles of attact (mesh multicolor plot) trying to get the rotor to tilt fast to the left, this
creates also some resistance of the rotor to move that fast as seen from the much smaller effective aerodynamic angles of attack (blue solid plot). The lateral rotation of disk creates resisting relative air flows, on top of the inertial slowness.
A more artistic plot of the lift and drag distribution.
A final note
------------
This phenomenon has been associated during the last decade by some people with the R22/R44.
I am myself a happy R44 owner. People have blamed in particular the hub design, which is imho not to blame.
All my test showed that the rotor remains very stable even partially stalled. What happens is not 'a wild flapping', but 'a wild body rocking' provoqued by the pilot.
This is intrinsic to theetering rotors that have no excentricity, so no stiffnes, so no direct roll control.
Delta3
----------
The results presented here are obtained via a complex mathematical simulator developped by myself.
The results in itself are not new, other people have explained the underlying principles.
What I think is new that they are confirmed by a full physical dynamic model with a target precision between 0.5 and 2%.
Although I performed numerous tests to check the validity, the model, nor the results have been checked by professional third parties including official instances. All technical and mechanical data has not been provided for by the constructor but result from own measurements.
The results should not be used to develop pilot instructions different from the instructions published in the official manuals and safety notices.
The results may be used by instructors to show the danger of push overs. This is one of the motivations of the simulator, namely avoiding showing it in real, something that in this particular case has rightfully been counter advised by official instructions.
Simulated flight profile
------------------------
- R44, two front pilots each 80kg, 140 lit of avgass, ISO
- initial level flight 90 knts
- pull up with constant collective to 80 knts, 1900 fpmin
- push over with constant collective
- left cyclic to counter the right roll
During the simulation a tail rotor governor was used assisting the pilot.
(flying a simulator is far from easy...) It is my believe that this governor actually
reduced the impact, because it reacts faster in adapting tail rotor trust than an average pilot would do in this scenario avoiding extra side-slip effects.
The following film summarizes the flight:
(quicktime 350 kb)
http://www.e-sign.be/private/heli/R44_FlightFilm06.mov
(media player 350 kb, lower resolution)
http://www.e-sign.be/private/heli/R4...ilm06_Fast.avi
The controls over time (playing with the decimal system to have same y-axis, sorry USA)
remark the desparate left cyclic at the and that does not seem to stop the rolling.
Discussion
----------
A push over starts with bringing the cyclic forward which provoques a forward pitching.
This pitching reduces rapidly the aerodynamic loading of the main rotor disk. The airflow comes much more from above, reducing the effective aerodynamic angles of the rotor blades
even though collective is kept steady.
In the scenario the level trust is about 9300 N, going over 10000N in the pull up
and dropping rapidly to 4000N during the push over. During that same periode
main rotor SHP goes from 110 down to 100 and next up to 120shp. (this does not include mechanical losses nor tail rotor power)
In level flight appropriate anti torque trust is provided by the tail rotor and equally important by the assymetric tail fin.
In the scenario this is approximately 2050 Nm in total.
These moments are evened out with the main rotor torque to provide yaw-stability.
Equally important is roll-stability.
The tail rotor trust and the tail fin trust create a positive roll moment
(positive = clockwise or to the right for a forward facing pilot) about the center of gravity that is much lower than centres of these forces (by more than one meter)
In level flight this rolling moment of approximately 250 Nm is countered
by a small left cyclic of about 0.7?, the main rotor being above the centre of gravity by more than 2 meters.
For the sake of completenes: one minor extra positive roll moment is produced by the down force of the assymetric tail wing, during pitching this increases (50 Nm).
Some frames from the film (apparently some technical problem with the colors..)
published in case download of the film is too long
Start of push over and roll
Because of the rapid drop of main rotor trust shortly after the start of the push over two desequilibriums come up.
Clockwise roll moment
The drop in main rotor trust also takes aways the left component and the resulting negative rolling moment. This produces a clockwise rolling. This seems to be the most important factor.
Yaw moment
The change in the air flow of the main rotor influences the main rotor yaw moment. In general a small increase in SHP is withnessed in this case. At speeds above 80 knots the tail fin will be fastest to react by increasing its moment (and doing so increasing slightly the roll moment)
A right roll in forward flight is instinctively compensated by left cyclic action. Because of the reduced trust the left cyclic does however not produce enough anti roll moment to stop the
rolling. At the end of the scenario with left cyclic at more than 7 degrees, lateral flapping of the hub approaches 10 degrees and mast bumping is in the make.... (to avoid sensationalism I stopped the flight before that)
Or viewed from the outside (in plot the heli is rolled back to level to better assess rotor angles)
From different simulation attempts it appeared to me that aggravating factors are apparently small inbalances prior to the low-G: such as slightly off cyclics and small prior rolling rates.
Some engineerings plots at the end of the flight:
Flapping angle of the hub starts to approach 10 degrees (blade flap is higher because of coning)
Remark the very high mechanical angles of attact (mesh multicolor plot) trying to get the rotor to tilt fast to the left, this
creates also some resistance of the rotor to move that fast as seen from the much smaller effective aerodynamic angles of attack (blue solid plot). The lateral rotation of disk creates resisting relative air flows, on top of the inertial slowness.
A more artistic plot of the lift and drag distribution.
A final note
------------
This phenomenon has been associated during the last decade by some people with the R22/R44.
I am myself a happy R44 owner. People have blamed in particular the hub design, which is imho not to blame.
All my test showed that the rotor remains very stable even partially stalled. What happens is not 'a wild flapping', but 'a wild body rocking' provoqued by the pilot.
This is intrinsic to theetering rotors that have no excentricity, so no stiffnes, so no direct roll control.
Delta3
Iconoclast
Joined: Sep 2000
Posts: 2,132
Likes: 0
From: The home of Dudley Dooright-Where the lead dog is the only one that gets a change of scenery.
I promised not to mention the numbers but I can't help myself.
I promised not to reference the numbers 18 and 72 but I just can’t help myself. I could say that the devil made me do it. I have long contended that with the 18-degree offset on the rotor system that the helicopter would fly to the left if the cyclic is moved forward from the rigged neutral position. Because of this I contended that in a pushover resulting in blade loss of lift that it would be necessary to move the cyclic to the rear and slightly to the left. This is supported by Tim Tuckers presentations in the safety course. Although this is my contention many of you including NickLappos have stated that the Robinson blade has a 72-degree phase angle as opposed to the 90-degree phase angle offered in my proposal..
If this is true the pilot will in order to counter the 18-degree offset will move the cyclic to the right. In a low G situation this placement of the cyclic will add to the roll induced by the tail rotor during the low G condition placing the helicopter into a snap roll, which an inexperienced pilot might not be able to counter.
Delta 3 in his post has suggested that the pilot move the cyclic to the left while pulling it back. I agree but how far to the left. If it is moved too far to the left the rotor will become dynamically unstable and result in a rotor incursion or mast bumping. However the POH says that the stick should be moved straight back. Who is correct?
If this is true the pilot will in order to counter the 18-degree offset will move the cyclic to the right. In a low G situation this placement of the cyclic will add to the roll induced by the tail rotor during the low G condition placing the helicopter into a snap roll, which an inexperienced pilot might not be able to counter.
Delta 3 in his post has suggested that the pilot move the cyclic to the left while pulling it back. I agree but how far to the left. If it is moved too far to the left the rotor will become dynamically unstable and result in a rotor incursion or mast bumping. However the POH says that the stick should be moved straight back. Who is correct?
Thread Starter
Passion Flying Hobby Science Sponsor Work
Joined: Apr 2004
Posts: 461
Likes: 0
From: Belgium
Hi Lu
Your response at light speed, seems your getting well again as many whished and hoped in this forum.
In level flight the cyclic controls work 'straight', I'll post the fases for this particular flight when level (tomorrow or so, its getting late in the old continent now)
In forward flight some left cyclic is needed not because of rotor misalignement but to counter the lateral effects of the tail rotor and fin.
During pull up cyclic goes slightly to the right, during pushover I simulated a panic reaction trying to correct the roll by going very far to the left again (see time plots)
All the math suggest the main rotor is exactly doing what the controls asked
Delta3
In level flight the cyclic controls work 'straight', I'll post the fases for this particular flight when level (tomorrow or so, its getting late in the old continent now)
In forward flight some left cyclic is needed not because of rotor misalignement but to counter the lateral effects of the tail rotor and fin.
During pull up cyclic goes slightly to the right, during pushover I simulated a panic reaction trying to correct the roll by going very far to the left again (see time plots)
All the math suggest the main rotor is exactly doing what the controls asked
Delta3
Joined: Apr 2003
Posts: 3,012
Likes: 1
From: USA
delta3 and Lu,
That is a nice study of what happens, the reason why the control sneaks all the way to the stops is that the rotor control power is nil at zero g, since the thrust is zero. If you try that with an articulated or "rigid" system the control will be retained. It has nothing to do with any missing rigging angles.
Teetering rotors have zero cyclic control at zero g. Thus any unbalancing force will upset thye aircraft.
That is a nice study of what happens, the reason why the control sneaks all the way to the stops is that the rotor control power is nil at zero g, since the thrust is zero. If you try that with an articulated or "rigid" system the control will be retained. It has nothing to do with any missing rigging angles.
Teetering rotors have zero cyclic control at zero g. Thus any unbalancing force will upset thye aircraft.
