Control movements

[Chuck Feldman]

The days gone have shown that back then they used 45 degree deflections. Today it is more like 15 degrees. So how did this evolution take place.  I hope to see a lot of discussion on this from higher skill fliers on this subject.

Thanks Chuck

[Brett Buck]

The days gone have shown that back then they used 45 degree deflections. Today it is more like 15 degrees. So how did this evolution take place.  I hope to see a lot of discussion on this from higher skill fliers on this subject.

Thanks Chuck

  What they said was going on, and what actually happens is two different things.  They didn't use 45 degree deflections in the good old days either. There wasn't anything like the necessary line tension to deflect the controls that far in flight. You could get them to move that far on the ground, but you can do that today, too. In-flight pictures of classic planes flown today show tiny deflections even when flown hard. As with most technical questions, Bill Netzeband knew all about this back in the 50s when it was going on, and it's called the "Netzeband wall" - the line tension and mechanical advantage of the control system was inadequate to deflect the surfaces into the airstream very far.

   If you tried to overcome this with engines and engine setup of the day by altering the control system, it didn't work anyway. If you manage 30 degrees of flap and elevator on a Fox 35/Nobler you are going to damn near stop dead in the air due to the huge drag and feeble power and speed stability. So even if you could have done it, you wouldn't want to, because you might get one great corner but the rest of the maneuver suffers or can't be completed. Do that at the first corner of the hourglass, and you aren't making it to the top.

    Current airplanes need less deflection for the same sort of corner because the CG is generally much further aft, made possible by much larger tail volume. They area also capable of much more in-flight deflection if needed, because the mechanical advantage of larger bellcranks and generally more solid line tension - and generally smaller control surfaces. At this point you have to limit the tightness of the corners for reasons other than the Netzeband wall, mostly, you still have to control the exit angles. Still, at the highest levels of competition, we are still cranking the corners about as hard as we think we can reliably do. Steve Fitton has posted a picture of me at the 2011 NATS with a handle angle corresponding to maybe 35-40 degrees of control deflection.

    An interesting effect of the Netzeband wall is how the airplanes responds through a corner. Old airplanes take a long time to start the corner, due to the low angular acceleration. While it's cranking up the corner, it's also slowing dramatically - which reduces the control loads and for given control pressure, allows the controls to deflect more. Which makes it accelerate in rotation even faster. So when flying them, the effect is that the entry is very "swoopy" and then it tends to "wrap up" into an ever-tighter corner. Then you have to stop it, and the reverse tends to happen - although not as dramatically since you are already slow and the engine (if set per standard 4-2 break practice and working perfectly) is starting to crank up. So the shape of the corner is nothing like a constant radius curve, it's sort of a lopsided hyperbola. You can see this even in the old pictures where they put flashlights on the wingtip, or use movie frame stills.

     More modern airplane, say, the prototypical top gun airplane, the Impact/40VF, as trimmed and flown by Paul Walker, has close to none of this. You can achieve the desired radius almost immediately, hold a constant tight radius , the kill the rate almost immediately. There's not a lot of drag, there's excellent speed stability and you have enough control margin to get whatever deflection you need and hold it. Paul is an absolute master at taking advantage of this which is why he was completely untouchable in the early 90's and still one of the the guys to beat today - after some of us caught on to the trick to one degree or another, and managed to use it to varying degrees of effectiveness.

    This type of performance has a lot of implications that you wouldn't necessarily expect. The engine performance and response in a corner is *absolutely critical* to the turn performance. That's why you see such endless talk about engine setup on TP engines - it's not because they are fussy or unreliable, it's because we are trying to get the engine to respond just right to make it as easy as possible to do the corners. Because if you get it wrong, and someone else gets it right, you lose. The other effect is that it puts *extreme* stress on the airplane, particularly the wing. We have had more wing failures since 1990 than I saw or heard of in the preceding 40-50 years. We keep building them stronger and stronger and they last a lot longer now than they did with old-style construction and the "lighter is better" theory, but there are still problems sometimes. Ever piped airplane I ever built was stronger than the last, but they have all ended up needing some sort of repair due to flight stresses.

    Brett

[Jim Thomerson]

I have read in a Frank Zaic yearbook that 45 degree deflection stalls the elevator and widens the turn, rather than tightens it.  I did some experiments using movable stops on the pushrod, and got results as Zaic predicted. 

[Brett Buck]

I have read in a Frank Zaic yearbook that 45 degree deflection stalls the elevator and widens the turn, rather than tightens it.  I did some experiments using movable stops on the pushrod, and got results as Zaic predicted. 

    If you look at it from a static sense, sure, if you just pop the elevator up to 45 degrees (or actually a lot lower) instantly with no pitch acceleration, it will stall. The same theory would suggest that if you put in a tiny deflection, the pitch rate would continually increase into a tighter and tighter radius.

     What is missing in this analysis (and seems to have been consistently missed by even the master himself) is that the pitch rate affects the tail AoA - drastically in our case. As soon as you start moving the elevator, the pitch rate starts accelerating, tending to *reduce* the AoA of the tail. Stop at some small fixed angle, and the pitch rate starts reducing the AoA of the tail until the torque falls to zero, at which point it stops accelerating in pitch and goes at a constant pitch rate. So the  response of the airplane and the angle of attack of the tail is highly dependent on the size of the tail, the moment of inertia, the tail moment, the speed, the speed of deflection of the elevator, etc. Older airplanes with tiny and short tails accelerate poorly (due to low pitch torque) and would have much more tendency to stall the tail than something like an Impact.

    These sort of dynamic effects are the missing link. People tend to focus on the aerodynamics and what the air does, and use full-scale practice and data as analogies. These are fine as far as they go. What is not analogous and what seems to have been consistently overlooked is that no full-scale airplanes ever operated in this sort of "non-equilibrium" conditions that we do, where the acceleration in rate is highly relevant, until very recently.

     Brett

[Howard Rush]

And even the static stuff wasn't well understood by Netzeband, maybe because he didn't have calculating machinery to see what is happening.  He was also focusing on different problems, like pushrod flex, that are no longer an issue.

[Brett Buck]

And even the static stuff wasn't well understood by Netzeband, maybe because he didn't have calculating machinery to see what is happening. 

  Amazing how many people managed to design acceptable airplanes (model and full-size) before CFD, though. Technical people underestimate Bill's knowledge because he wrote it for Joe Bellcrank. Zaic's stuff was a lot more speculative than Bill's.

    The only exception I ever had with Bill's stuff (and I talked to him about it) was not that he was solving the problems wrong, but that he was solving the wrong problems. He got frustrated over stunt, in particular, because he didn't want to accept, or at least didn't like, that it was not primarily an engineering exercise.

      Brett

[Tim Wescott]

     What is missing in this analysis (and seems to have been consistently missed by even the master himself) is that the pitch rate affects the tail AoA - drastically in our case. As soon as you start moving the elevator, the pitch rate starts accelerating, tending to *reduce* the AoA of the tail. Stop at some small fixed angle, and the pitch rate starts reducing the AoA of the tail until the torque falls to zero, at which point it stops accelerating in pitch and goes at a constant pitch rate. So the  response of the airplane and the angle of attack of the tail is highly dependent on the size of the tail, the moment of inertia, the tail moment, the speed, the speed of deflection of the elevator, etc. Older airplanes with tiny and short tails accelerate poorly (due to low pitch torque) and would have much more tendency to stall the tail than something like an Impact.

If you're talking about Mr. Ziac, he does explain that effect in "Circular Airflow".  Not quite in the language that you're using (although yours is clearer to me).  Instead, he plots an airplane on a flight path that is curved in pitch, then does sort of an eyeball conformal mapping that straightens out the flight path, curving the airplane in the process -- all this to explain why you need lots of elevator deflection, etc.

This is all an appendix in the back of the book -- he's clearly more interested in free flight, but he's willing to stand on the sidelines and render opinions about C/L.

[Howard Rush]

  Amazing how many people managed to design acceptable airplanes (model and full-size) before CFD, though. Technical people underestimate Bill's knowledge because he wrote it for Joe Bellcrank. Zaic's stuff was a lot more speculative than Bill's.

    The only exception I ever had with Bill's stuff (and I talked to him about it) was not that he was solving the problems wrong, but that he was solving the wrong problems. He got frustrated over stunt, in particular, because he didn't want to accept, or at least didn't like, that it was not primarily an engineering exercise.

Bill explained stuff well for high school kids like me at the time.  He oversimplified things, though, which probably contributed to his solving the wrong problems. 

I ain't talking about CFD.  If Bill had a PC at the time he wrote his control system stuff, he could have done a lot more.  For example, control system geometry is a three-dimensional problem that would have taken a lot of work and a heap of chase-around charts to calculate.  Now it's trivial.   

[Jorge de Azevedo]

......If you look at it from a static sense, sure, if you just pop the elevator up to 45 degrees (or actually a lot lower) instantly with no pitch acceleration, it will stall. The same theory would suggest that if you put in a tiny deflection, the pitch rate would continually increase into a tighter and tighter radius.......

As a new pilot, I woul like to hear from the experts about the way you achieve the maximum deflection. In other words, I would like to hear about linear versus exponential control systems and its efects on pitch rate .
Thank you in advance,
Jorge de Azevedo

[Chris Wilson]

 What they said was going on, and what actually happens is two different things.  They didn't use 45 degree deflections in the good old days either. There wasn't anything like the necessary line tension to deflect the controls that far in flight. You could get them to move that far on the ground, but you can do that today, too. In-flight pictures of classic planes flown today show tiny deflections even when flown hard. As with most technical questions, Bill Netzeband knew all about this back in the 50s when it was going on, and it's called the "Netzeband wall" - the line tension and mechanical advantage of the control system was inadequate to deflect the surfaces into the airstream very far.

   If you tried to overcome this with engines and engine setup of the day by altering the control system, it didn't work anyway. If you manage 30 degrees of flap and elevator on a Fox 35/Nobler you are going to damn near stop dead in the air due to the huge drag and feeble power and speed stability. So even if you could have done it, you wouldn't want to, because you might get one great corner but the rest of the maneuver suffers or can't be completed. Do that at the first corner of the hourglass, and you aren't making it to the top.

    Current airplanes need less deflection for the same sort of corner because the CG is generally much further aft, made possible by much larger tail volume. They area also capable of much more in-flight deflection if needed, because the mechanical advantage of larger bellcranks and generally more solid line tension - and generally smaller control surfaces. At this point you have to limit the tightness of the corners for reasons other than the Netzeband wall, mostly, you still have to control the exit angles. Still, at the highest levels of competition, we are still cranking the corners about as hard as we think we can reliably do. Steve Fitton has posted a picture of me at the 2011 NATS with a handle angle corresponding to maybe 35-40 degrees of control deflection.

    An interesting effect of the Netzeband wall is how the airplanes responds through a corner. Old airplanes take a long time to start the corner, due to the low angular acceleration. While it's cranking up the corner, it's also slowing dramatically - which reduces the control loads and for given control pressure, allows the controls to deflect more. Which makes it accelerate in rotation even faster. So when flying them, the effect is that the entry is very "swoopy" and then it tends to "wrap up" into an ever-tighter corner. Then you have to stop it, and the reverse tends to happen - although not as dramatically since you are already slow and the engine (if set per standard 4-2 break practice and working perfectly) is starting to crank up. So the shape of the corner is nothing like a constant radius curve, it's sort of a lopsided hyperbola. You can see this even in the old pictures where they put flashlights on the wingtip, or use movie frame stills.

     More modern airplane, say, the prototypical top gun airplane, the Impact/40VF, as trimmed and flown by Paul Walker, has close to none of this. You can achieve the desired radius almost immediately, hold a constant tight radius , the kill the rate almost immediately. There's not a lot of drag, there's excellent speed stability and you have enough control margin to get whatever deflection you need and hold it. Paul is an absolute master at taking advantage of this which is why he was completely untouchable in the early 90's and still one of the the guys to beat today - after some of us caught on to the trick to one degree or another, and managed to use it to varying degrees of effectiveness.

    This type of performance has a lot of implications that you wouldn't necessarily expect. The engine performance and response in a corner is *absolutely critical* to the turn performance. That's why you see such endless talk about engine setup on TP engines - it's not because they are fussy or unreliable, it's because we are trying to get the engine to respond just right to make it as easy as possible to do the corners. Because if you get it wrong, and someone else gets it right, you lose. The other effect is that it puts *extreme* stress on the airplane, particularly the wing. We have had more wing failures since 1990 than I saw or heard of in the preceding 40-50 years. We keep building them stronger and stronger and they last a lot longer now than they did with old-style construction and the "lighter is better" theory, but there are still problems sometimes. Ever piped airplane I ever built was stronger than the last, but they have all ended up needing some sort of repair due to flight stresses.

    Brett

Wonderful, wonderful reply and I applaud the effort that went into that tome.

Well done.

[George]

Shucks, I thought they used all that deflection to try for them five foot corners!

George

[Derek Barry]

"Steve Fitton has posted a picture of me at the 2011 NATS with a handle angle corresponding to maybe 35-40 degrees of control deflection."

Yes, and that is in my top ten favorite Nats pictures. Too bad we didn't have one of the plane at that very same moment. Sounds like a good project for this year's Nats.

Derek

[Brett Buck]

......If you look at it from a static sense, sure, if you just pop the elevator up to 45 degrees (or actually a lot lower) instantly with no pitch acceleration, it will stall. The same theory would suggest that if you put in a tiny deflection, the pitch rate would continually increase into a tighter and tighter radius.......

As a new pilot, I woul like to hear from the experts about the way you achieve the maximum deflection. In other words, I would like to hear about linear versus exponential control systems and its efects on pitch rate .

    You don't want exponential controls, at least on a conventionally designed stunt plane. All that does is make it softer around neutral, which makes it take longer to get in the necessary deflection to smartly start and stop the corner. Most of what we have done in the design of the airplane has had the effect, intentional or not, of making the airplane more sensitive around neutral instead of less.

    Brett

[Kim Mortimore]

.....Most of what we have done in the design of the airplane has had the effect, intentional or not, of making the airplane more sensitive around neutral instead of less.

    Brett

Does this have any effect on the ability of the less-than-expert pilot to fly the steadily in level flight, especially inverted?

[Howard Rush]

As a new pilot, I woul like to hear from the experts about the way you achieve the maximum deflection. In other words, I would like to hear about linear versus exponential control systems and its efects on pitch rate .
Thank you in advance,
Jorge de Azevedo

Control systems that have a high ratio of control surface movement to control line movement at large control surface angles ("exponential controls") are difficult to fly accurately because line flexibility and hinge moment make it difficult to get the intended control movement.  The opposite of the "exponential" system may be beneficial.  Even better may be the opposite of the "exponential" system applied only to the flaps.  That's what the current European champion and, I am told, a past US national champion used.   RC airplanes are a different matter: there is no aerodynamic feedback to the pilot, and there is no unpredictable flexibility between the actuator and the control surface.

[Howard Rush]

Does this have any effect on the ability of the less-than-expert pilot to fly the steadily in level flight, especially inverted?

Not much.  Any difficulty there probably comes from a trim problem or some nonlinearity like control friction.