Question on the Netzeband wall

[55chevr]

Steve,

? care to explain that?



Joe

[john e. holliday]

Say there was a guy that built, oh, say, four new stunt planes every year for 40 years. That would be a lot of planes! But I don't see how that one guy would be able to fly enough to get 4 new planes really well trimmed each season. Without getting them well trimmed, will he be able to correctly evolve his design, even if they are all basically the same? I don't see that being possible, plus, they are not all basically the same. Hypothetically speaking, of course...  Steve

Easy, you are a millionaire with unlimited resources and tons of time on your hands.   When it is day time you go fly.   When it is night or weather really unbearable, you work on planes.  But, it is fun watching people through the years.  Someone win the NATS with a big plane.  The next year there are more big planes.   Then the pipes came out and now it is electric.   I am waiting for someone to win the NATS with a 40 size plane with electric propulsion.

[Eric Viglione]

I am waiting for someone to win the NATS with a 40 size plane with electric propulsion.

Really? I thought that already happened last year, no?
If you are going to dream, you need to dream bigger! How about winning the Nat's with the new YS63 C/L 4 stroke in a Bipe!     
Then everyone will run out and build one...right? The pits full of 4 stroke Bipe's, now how cool would that be?
(I think I just heard Milton fall out of his chair, sorry buddy, that was low, wasn't it?)

[Igor Burger]

   Finally, you can just make the airplane heavier. The weight is more-or-less a linear relationship to the line tension. Add 20% to the mass, and you get 20% more line tension, all other things being equal. That way, you can definitely deflect the controls further at a given speed. Or, you can get the same deflection at a lower speed.

Brett, are you sure wit that? Because I am not so sure :- )))

Yes you can get 20% more line tension, but 20% heavier model will need also aproximately 20% stronger hinge moment as it comes from pressuse on wing (that pressure which must overcome 20% centrifugal force necessary to carry 20% heavier model on the same flight path). So I would say just oposite, heavier model will not make situation with Netzeband wall any better :- )) ... well it can, if you push the weight to partial (flap) stall, then yes, it could help because hingemoment on flap can drop to 1/2 of normal but airflow separated on hinheline is last think I would like to have on my model :- )))

[Ted Fancher]

Brett, are you sure wit that? Because I am not so sure :- )))

Yes you can get 20% more line tension, but 20% heavier model will need also aproximately 20% stronger hinge moment as it comes from pressuse on wing (that pressure which must overcome 20% centrifugal force necessary to carry 20% heavier model on the same flight path). So I would say just oposite, heavier model will not make situation with Netzeband wall any better :- )) ... well it can, if you push the weight to partial (flap) stall, then yes, it could help because hingemoment on flap can drop to 1/2 of normal but airflow separated on hinheline is last think I would like to have on my model :- )))

Igor,

A hypothetical question.  If you had a zero ounce 700 square inch flapped stunter flying on 70" lines at five second a lap how much line tension would there be?  How much control deflection would be achievable?

How about 10 oz?  20 oz? thirty ?

Ted

[Brett Buck]

Brett, are you sure wit that? Because I am not so sure :- )))

Yes you can get 20% more line tension, but 20% heavier model will need also aproximately 20% stronger hinge moment as it comes from pressuse on wing (that pressure which must overcome 20% centrifugal force necessary to carry 20% heavier model on the same flight path). So I would say just oposite, heavier model will not make situation with Netzeband wall any better :- )) ... well it can, if you push the weight to partial (flap) stall, then yes, it could help because hingemoment on flap can drop to 1/2 of normal but airflow separated on hinheline is last think I would like to have on my model :- )))

  If you were running out of lift before, no. If the problem is that you can't deflect the elevator enough (which was the case here), you don't need more just because it is heavy.

     Brett

[Igor Burger]

  If you were running out of lift before, no. If the problem is that you can't deflect the elevator enough (which was the case here), you don't need more just because it is heavy.

     Brett

I thought we are speaking about flapped stunter, in that case most of feedback comes from flaps and if weight is 20% higher, also feedack is aproximatey 20% higher ... more in next post ...

[Brett Buck]

I thought we are speaking about flapped stunter, in that case most of feedback comes from flaps and if weight is 20% higher, also feedack is aproximatey 20% higher ... more in next post ...

\

    Bear in mind that the lap times also go up, from 4.6 to 5.2 secs/lap or so, so you get a 27% reduction in the control torque at a given deflection.

     Brett

[Igor Burger]

Igor,

A hypothetical question.  If you had a zero ounce 700 square inch flapped stunter flying on 70" lines at five second a lap how much line tension would there be?  How much control deflection would be achievable?

How about 10 oz?  20 oz? thirty ?

Ted

I do not think it makes too much sense to explore singularities like 0 weight needed 0 lift to make a corner, but if you take 10 oz stunter which needs some lift L for making some given radius which is on edge of Netzeband wall (means line tension equals force necessary to make that radius) and having some C centrifugal force, then if you go to 20oz stunter, you will have 2*C centrifugal force and thus twice stronger input to controls. That is where we agree.

That heavier stunter will certainly need also twice the lift of that lighter 2*L at the same speed and the same radius. It is because of the same equation for centrifugal force which gives higher line tension 2*C. Just the radius is not line length, but radius of corner instead. (note: I ignored gravity to make it simpler)

The pressure distribution chordwise (the quality) on wing will be very similar on both models because it does no change too much with changed lift, just that heavier, which needs 2*L lift will have twice higher pressure (its quantity, but distribution will be very similar). It means the pressure of flaps will be twice higher what will make twice higher hinge moment which will be transformed to twice higher force necessary to make the same corner, so we re back on the same minimal radius.

Most of the feedback comes from flaps, I am now not able to say hom much, because i am on trim in CA now and all my calculators are home, but elevator hinge motment is really small. But ok Brett is right, we have also elevator. Elevator must counterbalance moment of CG to AC. That moment will need some lift and thus pressure distribution on elevator. But is very similar to flaps. Twice heavier model will have twice higher CG moment and thus it will need twice higher elevator lift and thus also absolute pressure. However we have still the same elevator and stab, so its distribution is still similar, it means also elevator hinge moment will be twice stronger on heavier model then on lighter.

So from this point of view weight does not change the situation. However if you count that heavier model will probably need little more deflected controls (that twice higher lift must be gained somewhere somehow – probably by more deflected flaps) means lift will be concentrated more on back side of airfoil, making worse pressure distribution and also higher airfoil moment which must be counterbalanced by elevator, so it can easily happen that lighter model could be even little better in meaning minimal flyable radius.

So this is probably what you did not want to hear : )))) ... however it is still only static situation. We have also some dynamic effects here allowing little tighter corner then limited by Netzeband wall - you can pull handle during comer and heavier model will give you better mass inertia to do it, but it has nothing to do with Netzeband wall, which is defined statically.

[Igor Burger]

\

    Bear in mind that the lap times also go up, from 4.6 to 5.2 secs/lap or so, so you get a 27% reduction in the control torque at a given deflection.

     Brett

Could be, but as you aleady wrote and I fully agree, the lap times also do not change it. Nececcary line force goes up with speed at the same magnitude like line tension - it is still the same equation for centrifugal force.

[Brett Buck]

Could be, but as you aleady wrote and I fully agree, the lap times also do not change it. Nececcary line force goes up with speed at the same magnitude like line tension - it is still the same equation for centrifugal force.

    Right, if you merely speed it up or down, it doesn't get you anywhere. Adding mass and reducing the speed alters the required control torque for a given deflection more than it alters the line tension. That's the whole point.

     Take the example above, the line tension goes down to 94% of the original but the aerodynamic loads go down to 79% of the original, so you are ahead of the game substantially. And, in fact, that's pretty much exactly what happened in the experiment (and the many similar examples over the years).

   Since Howard always wants me to show my work,

 Original = 36 oz at 4.6 sec laps, 89 fps, .069 slugs, 65 feet; line tension = 8.52 lb, q = 9.3 psf
  final = 44 oz at 5.2 sec laps, 78 fps, .086 slugs, 65 feet; line tension = 8.04 lb, q= 7.3 psf

   ratio of line tensions = .94, ratio of q = .79

    As an aside, note also that all the aerodynamic forcing functions for out-of-trim conditions also go down substantially, while the line tension that acts to stabilize it (essentially all the roll restoring torque and maybe 50% of the yaw) changes only a little bit. This would have drastically assisted the 585 square inch/54 ounce airplane since it also had some serious aerodynamic trim issues that were not that easy to fix, without taking a saw to the fin.

     Brett
 

[Howard Rush]

One curious thing is that Brett, Igor, and Ted are within a kilometer or so of each other now.

[Mark Scarborough]

well thankfully they are not sitting at a table sipping a beer talking about this, or we would all miss out on the exchange!!

[Steve Helmick]

Those 70" lines have me concerned, tho!  Steve

[Igor Burger]

Original = 36 oz at 4.6 sec laps, 89 fps, .069 slugs, 65 feet; line tension = 8.52 lb, q = 9.3 psf
  final = 44 oz at 5.2 sec laps, 78 fps, .086 slugs, 65 feet; line tension = 8.04 lb, q= 7.3 psf

   ratio of line tensions = .94, ratio of q = .79

Hmmm ... I am lost in numbers and especially units and where they come from, as I am here only with handy ... so I affraid I can continue when I come back home, however it is no queston that proper weight is propper weight, as much as wing can carry and motor can pull, but now I do not see directly how it can help to push Netzeband wall, especially if heavier and slower model will need larger controlls deflection, I will try it home on my stnt calculator, I am already curiouse : -))

BTW I am now some 50 miles from SF .. exactly 53 from Alcatraz .. at least that is waht google maps told me when I started my way yesterday :- )))  ... I am in Mountain Vew visiting my son :- )))

[L0U CRANE]

GREAT thread!

Only two thoughts:-

1.) The term 'hinge moment' may confuse some of us.

Try it this way: A moment is a force at an arm distance, tending to cause rotation about a point. When we move our control surfaces off neutral, they meet airloads trying to return them to neutral. A point where, in effect, all the airload acts on a surface can be SWAG'ed or estimated. The distance from that point to the hingeline is the moment arm, and the total force x that distance is the hinge moment.

To hold the control surfaces to angles needed to perform figures, we need to match the airload hinge moments, by shifting pull between the lines, sharing the full value of centrifugal force - split equally at neutral - to more on the 'loaded' line and less on the 'unloaded' line. The pull difference, led through the series of levers (bellcrank line radius, bellcrank pushrod radius, horn radii at each control surface) applies the matching counter moment.

The Netzeband Wall is hit when the pull (basically CF) cannot reach the values needed to hold or increase control surface deflection for conditions. The model does not turn tighter, however much we want it to.  If too low, and it opens up... Next model please...

Note: It IS possible to add briefly to pull's CF value by yanking the handle away from the model. (Inertia permits a very brief boost to CF. Paul Walker, in a video I've seen, and in seeing him fly, uses this, quite violently, for corners. It does work!)

2.) Line curvature due to drag: (weight IS uniform, but very small and not sigificant for the following remarks.)

I've seen the curve our lines take in flight called an "accelerated catenary." A simple catenary is the curve we see in a cable suspension bridge, supported at both ends, for a cable of uniform weight load along its length. Our lines, in flight, do not meet a uniform load. Air drag varies with the square of velocity - which, itself, varies from (in effect) zero at the handle, to whatever value exists at the leadout guides.

Incidental to this: If the line pull direction, when it reaches the leadout guides, can be aimed at the CG, we gain. "Moments," again... LINE I and LINE II help for these concepts...

If the line pull aims ahead of the CG, it tends to pull the model's nose "in" toward the flier. And vice versa...

These are 'yaw' disturbances, and at extreme, can cause bad hinging. It is ideal when yaw and roll tendencies are at minimum.

Gyro effects from the propellor and crankshaft, for conventional (CCW rotation props, seen from in front,) ""nitro"" models (flown CCW), cause a nose-out tendency on inside turns, and, again, vice versa.

(It may be possible, if good numbers exist, to spread the leadout guides chordwise enough to reduce, or even counter, the yaw tendencies! For years, I've tried to cancel these for the worst case I could estimate, hoping that that would deal with "lesser included" conditions. I'm no rocket scientist -Howard and Brett are - but my approach seemed to work for my level of seriousness and potential...)

Over to you,Chet...

[Brett Buck]

Hmmm ... I am lost in numbers and especially units and where they come from, as I am here only with handy ... so I affraid I can continue when I come back home, however it is no queston that proper weight is propper weight, as much as wing can carry and motor can pull, but now I do not see directly how it can help to push Netzeband wall, especially if heavier and slower model will need larger controlls deflection, I will try it home on my stnt calculator, I am already curiouse : -))

BTW I am now some 50 miles from SF .. exactly 53 from Alcatraz .. at least that is waht google maps told me when I started my way yesterday :- )))  ... I am in Mountain Vew visiting my son :- )))

     That would be about 5 miles from here...


    I can convert it to old-world units when I get a chance, but the ratios will be the same.

      Brett

[Peter Germann]

[I will try it home on my stunt calculator, ... [/quote]

A truly interesting thread, indeed! How about all of this when I fly corners above level, such as upper square, triangle and hr. glass corners? Could it be the Netzeband wall is moving with line elevation angle?

Peter Germann

[Phil Krankowski]

[I will try it home on my stunt calculator, ...

A truly interesting thread, indeed! How about all of this when I fly corners above level, such as upper square, triangle and hr. glass corners? Could it be the Netzeband wall is moving with line elevation angle?

Peter Germann

While the line weights will balance each other out due to having the same drag and weight to them, at elevation (directly overhead) the weight of the model is removed from the line tension, and at a steep angle a large fraction of the weight of the model is removed from line tension (apply some vector math, trig and all), so yes, the upper portion may have less tension compared to the level portion at shoulder height.

Phil

[Igor Burger]

[I will try it home on my stunt calculator, ...

A truly interesting thread, indeed! How about all of this when I fly corners above level, such as upper square, triangle and hr. glass corners? Could it be the Netzeband wall is moving with line elevation angle?

Peter Germann

Yes it does, All we wrote until now is in level flught. Overhead we have liine tension less gravty and it makes situation completaly diferent. Minimal radius is narualy larger due to lower line tension, but it chages also effect of different weight and different speed. While the weight does not change situation + that similar like in level, the the higher speed wit improve it, because the higher speed we fly, the higher centrifugal acceleraion we have and the lower proportional effect gravity has.  ... at least that it how I see it now, just without couning numbers. So I think that quicker lighter model from Bretts example will have advantage overhad.

However there is still effect of gravity for necessary lift, which is present in level, but not overhead, so it is little more complicated, and the difference is not so large ... it needs really count numbers then just guess.

[Igor Burger]

Note: It IS possible to add briefly to pull's CF value by yanking the handle away from the model. (Inertia permits a very brief boost to CF. Paul Walker, in a video I've seen, and in seeing him fly, uses this, quite violently, for corners. It does work!)

Yes but it has also its limits, if you fly square loop, you can use little enegy from model mass inertia, but you must allow model to gain it back, it means you need some time and some distance between corners to release handle back. There are many pilots flying like that. And not only because pure design of their models, some people simply like such feeling, they can easier feel impulse given to handle then deflection and time. Once I even saw friend of mine from Ukraine (they typically fly like that) trying to fly my 200g indoor like that  ... well without too much success n first corners :- )))

[Igor Burger]

     That would be about 5 miles from here...


    I can convert it to old-world units when I get a chance, but the ratios will be the same.

      Brett

5 miles? it must be somewhere close, I see you have perfect weather for flying whole year, you not need need winter sleeping like we do :- ))))))))))))


yes, may be it is not question of units, may be if you can tell how you got those numbers ... the thing which I cannot get is how you got that difference if you combined speed and weight difference together, if speed and also weight difference alone does not make any difference ... hmm ... I am missing my notebook and my spread sheets ... we are going o visit some aquarium, may be I will get it looking to blue water :- )))))

[Brett Buck]

[I will try it home on my stunt calculator, ...

A truly interesting thread, indeed! How about all of this when I fly corners above level, such as upper square, triangle and hr. glass corners? Could it be the Netzeband wall is moving with line elevation angle?

Peter Germann

   Absolutely. The worst case is probably the hourglass. But don't overlook the effects of varying speed and the airplane getting "out of shape" due to trim issues during other maneuvers, particularly the square 8. If everything is not perfect, the available corner can vary dramatically from corner to corner.  That's the reason nearly no one flies the maneuvers the right size - these sort of transient effects have to mostly go away before you can successfully do a clean corner. They aren't too big because the corners are too big, they are too big (for the most part*) because you have to let it recover after the corner. Once you get to a certain skill level, this is what separates the competitors in many cases.

    That's why trim matters so much and why getting it perfect is so important - and takes up the vast majority of the practice time.

    We are only talking about quasi-steady-state, for the most part, above. That's why I mention some improvement in the restoring forces above, more is going on that just running out of control torque.

      Brett

*check some videos of Igor's WC flights, and compare the sizes to what you see at local contests! When Ted/David/I practice, we spend time from flight to flight on correcting shape errors, but we talk about sizes over periods of decades.

[Brett Buck]

5 miles? it must be somewhere close, I see you have perfect weather for flying whole year, you not need need winter sleeping like we do :- ))))))))))))

    What do you mean, it rained for about 5 minutes several times this week, and the highs are only in the 60s? They call that a major winter storm around here. I laugh at it, too, I grew up in the Midwest, where brass monkeys are an endangered species.

yes, may be it is not question of units, may be if you can tell how you got those numbers ... the thing which I cannot get is how you got that difference if you combined speed and weight difference together, if speed and also weight difference alone does not make any difference ... hmm ... I am missing my notebook and my spread sheets ... we are going o visit some aquarium, may be I will get it looking to blue water :- )))))

    m(V)**2/r  (line tension) is the same no matter where you are!   same with 1/2*rho*v**2 (q, AKA dynamic pressure). I am using the values shown with because those are roughly the parameters of the Tucker Special tests mentioned above.

4.6 seconds a lap and 36 ounces ("before") 
5.2 seconds a lap and 44 ounces ("after", same airplane with 8 ounces added)
 r= 65 feet in either case (~60ft lines+length of Ted's arm+length of inboard wing), circumference = 408 feet

 m = mass in slugs, mass= weight in lbf/32.174, weight in lbf = weight in ounces/16

v= velocity in feet per second = 408 feet/(lap time)

rho = density of air = 0.00237 slugs/cubic foot (I assumed STP, which is close enough since we were in the Mission College parking lot about 8 miles from your current position, and you will pass right by if you are headed to the Monterey Bay Aquarium, elevation is about 8 feet - it's by Great America amusement park {which is closed for the season, sorry}).

psf = pounds per square foot

   And before I get another lecture from the "USA Metric in '78!" crowd... Anyone reporting their weight in Kg is committing all the same offenses, because the SI unit of force is Newtons, not Kgf. and if you try to get it all right, you have to put in a fudge factor of 9.8 somewhere - just like the 32.174

     Brett

[Igor Burger]

I see we have different approach here, I think rho does not play any effect here, as the wing pressure is given from centrifugal force and wing area, so the line tension from centrifugal force out of the circle and centrifugal force out of the flight radius goes up at the same ratio, so that is why I think change of mass does not change situaton ... ok, at least i have enough for think about :- )))

[Igor Burger]

    What do you mean, it rained for about 5 minutes several times this week, and the highs are only in the 60s? They call that a major winter storm around here. I laugh at it, too, I grew up in the Midwest, where brass monkeys are an endangered species.

be happy, we have only one rain during winted ... 3 monthes long :- ))))))))))))))))

[Howard Rush]

It may be possible, if good numbers exist, to spread the leadout guides chordwise enough to reduce, or even counter, the yaw tendencies!

I had an interesting discussion on this yesterday.  It was about forwards props and backwards props and making a new plane with up on the aft line (or maybe up on the forward line) instead of the other way around, all to fix a funny on the third lobe of the clover.  Anyhow, I lost track of all the sign changes.  

Good explanation of hinge moment.    Hinge moments are kinda hard to predict, which is another bone I'd pick with Bill Netzeband's article about them.  He overgeneralized the calculation.

[Howard Rush]

A truly interesting thread, indeed! How about all of this when I fly corners above level, such as upper square, triangle and hr. glass corners? Could it be the Netzeband wall is moving with line elevation angle?

Yes, see the second graph in reply 29 above (sorry about the proprietary measurement units).  There are also the destabilizing effects on line tension of thrust and Cl_β.  Less line tension or more wind cause the airplane to point toward the pilot.  The effect of the prop disk being at an angle to the flow is a lot more than thrust * sin β.  These effects are exacerbated close to the Wall, as seen in that graph.

[Paul Walker]

be happy, we have only one rain during winted ... 3 monthes long :- ))))))))))))))))

Be happy is is only rain  !!!
We get snow here!

[Bruce Perry]

We get snow here!
[/quote]

"that's not Snow, This is snow...."  says Polarbear Dundee.

hahaha

B

[Paul Walker]

We get snow here!


"that's not Snow, This is snow...."  says Polarbear Dundee.

hahaha

B


Bruce, get back to charging your Li-Po's.....

[Paul Walker]

Brett, are you sure wit that? Because I am not so sure :- )))

Yes you can get 20% more line tension, but 20% heavier model will need also aproximately 20% stronger hinge moment as it comes from pressuse on wing (that pressure which must overcome 20% centrifugal force necessary to carry 20% heavier model on the same flight path). So I would say just oposite, heavier model will not make situation with Netzeband wall any better :- )) ... well it can, if you push the weight to partial (flap) stall, then yes, it could help because hingemoment on flap can drop to 1/2 of normal but airflow separated on hinheline is last think I would like to have on my model :- )))

Igor,
Where does the extra 20% lift come from?

[L0U CRANE]

Howard, to your #70, thanks for the kind comment.

As to CW shaft v. CCW shaft (as seen from out front), my only practical application was in the early 1980's on an OT AA, Sr., Fox 35Stunt, factory shaft.

Main difference noted at my cuneiform in clay skill level computer was that the basic engine torque reaction on the fuselage, being reversed, held the inboard wing up better. No tip weight needed. Takeoffs were easier, and the bellcrank connections could be Jim Walker Basic (down line forward.) No tabs or warts needed.

All other practical handling actions the same both ways, except that flipping that CW prop with the wheels on the ground consisted of punching the pavement, Yes, ouch, and with the flying hand...

To Igor - earlier post:

Of course the 'yank' to increase line pull was - I thought - understood to be only for that brief moment when it was needed to move Netzeband's wall back a small distance. Unless we are whipping the model continuously, we should have enough CF pull to be comfortable almost everywhere.

Howard, estimating the force and effective chord distance from hingeline where it can be considered to act is something I'd gladly leave to you, Brett and Igor... For my rather crude uses, I applied what the NACA wind tunnel pressure profiles in approx WW2 basic texts seemed to imply for deflected flap surfaces. SWAG, not rigorous, but a more credible start point than none at all. At my highest level of performance, that was about a match...

[Igor Burger]

Less line tension or more wind cause the airplane to point toward the pilot.

Depends what is the model and who is the pilot. ... certainly not mine to me :- )))

[Igor Burger]

Be happy is is only rain  !!!
We get snow here!

we get snow afterwards, March and April :- )))

[Igor Burger]

Igor,
Where does the extra 20% lift come from?

If it should be the same model, then the only way is elevator and flap deflecttion, but it is hard to predict without more data, so I sinply ignored it. However as I wrote before, especially flaps which will be more deflected will push center of pressure more back (if it does not stall), it means the pressure distribution will be moved more to flaps and so neavier model will fly even more opened corners, instead of tighter. But I think that modified pressure distribution will make very small difference.

[Howard Rush]

   Right, if you merely speed it up or down, it doesn't get you anywhere. Adding mass and reducing the speed alters the required control torque for a given deflection more than it alters the line tension. That's the whole point.

I don't think so, nor do I think that's the point.  I shall attempt to explain.

Take the example above, the line tension goes down to 94% of the original but the aerodynamic loads go down to 79% of the original, so you are ahead of the game substantially. And, in fact, that's pretty much exactly what happened in the experiment (and the many similar examples over the years).

   Since Howard always wants me to show my work,

 Original = 36 oz at 4.6 sec laps, 89 fps, .069 slugs, 65 feet; line tension = 8.52 lb, q = 9.3 psf
  final = 44 oz at 5.2 sec laps, 78 fps, .086 slugs, 65 feet; line tension = 8.04 lb, q= 7.3 psf

   ratio of line tensions = .94, ratio of q = .79

These numbers are a tad off--Brett must have been using his 5" slide rule, rather than the big one--but pretty close.  I get line tension 2 / line tension 1 = .96 and q2/q1 = .78 .  But as Igor was saying, the heavier model requires more lift for a given loop radius.  His simplifying assumption was that the pressure distribution on the wing and flaps had the same shape, but was multiplied by the ratio of lift required for the heavy airplane to lift required for the light airplane.  Thus the hinge moment as well as the line tension is proportional to mV².  The ratio of hinge moments is therefore .96 also, given Igor's simplifying assumption.  So at best, the ratio of line tension to hinge moment is 1 ratio (line tension 2/hinge moment 2) to (line tension 1/hinge moment 1) is 1.  Actually, for the same elevator/flap ratio, you'd need a bit more flap deflection to get the increased Cl you'd need for the same loop radius with the heavier model, so hinge moment would go up.  Thus adding weight decreases the ratio of line tension to hinge moment.  But it worked; it made the Tucker fly better.  

Here's where it gets weird.  I took the program that generated the plots in reply 29 and did runs for the Tucker Special at the two different weights, but at the same speed: 5-second laps.  I don't know how much differential line tension it takes for a given maneuver.  I'll tell you when I determine hinge moment.  However, for Igor's simplifying assumption of the same pressure distribution shape, the same maneuver would require the same difference in line tension for both airplane weights.  On the attached plot, you can see that for a given difference in tension between the lines, the 44 oz. airplane has a less perverted response to control inputs than the 36-oz. airplane.  Even if hinge moment coefficient goes up some for the heavier airplane, as it will, there could still be a big advantage to adding the weight.  I picked a couple of points as samples.  The square on the 44-oz. line requires more differential line tension to react hinge moment than the square on the 36-oz. line, but still gives better control response and is further from the Wall.

To include the speed difference in the comparison, I'd need to change the plot to normalize the X axis by V², but I have parts to sand and glue, so it can wait until after August when I finish the short monograph on hinge moment.

My conclusion from all this is that an airplane probably does have an optimal weight, as folks have said above, but that weight should be determined by turbulence response, rather than hinge moment issues.  If you have a good control system, you won't have to worry about getting the plane too light for the Netzeband wall.  Hint: you won't get there looking at the leverage of control surface to bellcrank.  A little added side force can help, too.  Sparky is on the right track.  

Edited to make a little more sense.

[Brett Buck]

I don't think so, nor do I think that's the point.  I shall attempt to explain.

These numbers are a tad off--Brett must have been using his 5" slide rule, rather than the big one--but pretty close.  I get line tension 2 / line tension 1 = .96 and q2/q1 = .78 .  But as Igor was saying, the heavier model requires more lift for a given loop radius.  His simplifying assumption was that the pressure distribution on the wing and flaps had the same shape, but was multiplied by the ratio of lift required for the heavy airplane to lift required for the light airplane.  Thus the hinge moment as well as the line tension is proportional to mV².  The ratio of hinge moments is therefore .96 also, given Igor's simplifying assumption.  So at best, the ratio of line tension to hinge moment is 1.  Actually, for the same elevator/flap ratio, you'd need a bit more flap deflection to get the increased Cl you'd need for the same loop radius with the heavier model, so hinge moment would go up.  Thus adding weight decreases the ratio of line tension to hinge moment.  But it worked; it made the Tucker fly better.  

    You appear to be  assuming that all the additional lift comes from deflecting the flap more, in which case you are correct. But I don't think that's what happens when you have coupled flaps. I think the flaps are deflected less than you think on the heavy airplane, but more than on the light airplane.

     Do the same thought experiment with no flaps at all, in which case the hinge moment per turn rate is almost unaffected by the weight of the airplane. Then do the same experiment with the flaps set to move only as far as they do in the "light case" and then stop (while the elevator can continue to move). This is pretty much the same as doing the same experiment with the flaps glued in place as some fixed angle.

     The heavy airplane ends up flying at a higher angle of attack for a given turn radius, but I think my ratio holds. The required pitch rate at "equilibrium" is actually lower on the heavier but slower airplane, you end up needing less deflection to achieve it. In fact the same theory can hold until your AoA in the corner exceeds the stall angle.

    This is now verging on the Imitation experiments with variable-sized flaps, and why heavy airplanes "hop" when the flaps are too small or have the wrong flap/elevator ratio. 

   Perhaps we aren't on the same page with what happens when you are in the middle of the turn in quasi-steady state conditions and/or how it gets to a particular angle of attack. 

      Brett

[Bill Johnson]

    I touched upon it above, but I was already droning on a bit. Running the CG aft has the effect of requiring less elevator deflection for a given rate of turn, and thus also deflecting the flap less. It doesn't help you move the controls further, just makes it less important, so you are still better off. The part I barely mentioned above about Paul's slow control ratios - which are enabled by the aft CGs, and the slow control ratios improve the margins because you get better translation of the bellcrank torque to control horn torque.

    The downside to the aft CGs is exactly the same as the advantage. By requiring less control deflection, you get less flap deflection, and less camber in the wing from the flap deflection (for a given pitch rate). If you were going to run out of lift, moving the CG aft and doing nothing else would hurt you and make the stall happen sooner. Al Rabe used the inverse of this to good effect, theorizing that running out of lift was his problem, he moved the CG forward (in the days of unadjustable controls) so he would would have more flap deflection at a given pitch rate, so he didn't run out of lift. The same side effects happened, of course, but it's better than stalling. So he ended up improving his corners by running the CG forward.
Brett

Delete this if I muddle up this excellent thread.

Generally, the CG will be somewhere forward of the CL (center of lift). Moving the CG closer to the CL makes the airplane more maneuverable, needing less force to rotate the aircraft around the lateral axis, but also makes it less stable.

I would think that at some point, the increased maneuverability would be offset by the instability, leading to over-controlling attempting to fly smooth maneuvers.

[Brett Buck]

Delete this if I muddle up this excellent thread.

Generally, the CG will be somewhere forward of the CL (center of lift). Moving the CG closer to the CL makes the airplane more maneuverable, needing less force to rotate the aircraft around the lateral axis, but also makes it less stable.

I would think that at some point, the increased maneuverability would be offset by the instability, leading to over-controlling attempting to fly smooth maneuvers.

   Oh, certainly, if you move the CG back enough, it will be unstable and/or unflyable. The theory in this thread is that weight was added, but the CG didn't change.

    Brett

[Igor Burger]

Yes I did not include different controls deflction for both weights, also for simplicity, and also because it can change situation depending on arms geometry, it can easily happen that line difference / hinge moment ratio changes with deflection (unequal bellcrank and flap horn arms) and there are also another effects which are not included to equaton like mentioned mass inertia but aslo transient effects of moment of inertia - heavier fuselage can be accelerated before it enters circular path and that will be different on lighter and heavier models. And also my logarithmic device (making feedback smaller at higher deflection necessary for heavier model) will make heavier plane controll easier then lighter. 

[Howard Rush]

    You appear to be  assuming that all the additional lift comes from deflecting the flap more, in which case you are correct. But I don't think that's what happens when you have coupled flaps. I think the flaps are deflected less than you think on the heavy airplane, but more than on the light airplane.

     Do the same thought experiment with no flaps at all, in which case the hinge moment per turn rate is almost unaffected by the weight of the airplane. Then do the same experiment with the flaps set to move only as far as they do in the "light case" and then stop (while the elevator can continue to move). This is pretty much the same as doing the same experiment with the flaps glued in place as some fixed angle.

     The heavy airplane ends up flying at a higher angle of attack for a given turn radius, but I think my ratio holds. The required pitch rate at "equilibrium" is actually lower on the heavier but slower airplane, you end up needing less deflection to achieve it. In fact the same theory can hold until your AoA in the corner exceeds the stall angle.

    This is now verging on the Imitation experiments with variable-sized flaps, and why heavy airplanes "hop" when the flaps are too small or have the wrong flap/elevator ratio. 

   Perhaps we aren't on the same page with what happens when you are in the middle of the turn in quasi-steady state conditions and/or how it gets to a particular angle of attack. 

You can get the extra lift from angle of attack.  If you want to do so without changing hinge moment between the light and heavy airplane, you could use fixed flaps and a balanced elevator.  I assume, though, that the experiment we are discussing is a given Tucker Special with the same control system at both weights.  That airplane will have more hinge moment as lift increases.  You did not improve the flyability of the airplane by reducing hinge moment via added weight.  You improved it by overcoming the control nonlinearity caused by line drag.   

[Steve Helmick]

From Brett's post: "Original = 36 oz at 4.6 sec laps, 89 fps, .069 slugs, 65 feet; line tension = 8.52 lb, q = 9.3 psf
  final = 44 oz at 5.2 sec laps, 78 fps, .086 slugs, 65 feet; line tension = 8.04 lb, q= 7.3 psf

   ratio of line tensions = .94, ratio of q = .79"


The subject Tucker Special has been ballasted up 8 oz and slowed from 4.6 sec. laps to 5.2 second laps. It will require a little more control input due to the increased weight, but due to the reduced speed/increased lap time, the hinge moments will be reduced quite a bit. It seemed to work; I was there.

One thing Gary Letsinger once told me was that sometimes the corner can be tightened by shifting the CG forward. Theory is this requires more control input, which brings more wing flap to bear. I suspect this would be only possible on heavy models and of course, modern adjustable controls would probably make this the wrong thing to do. But then, not everybody incorporates adjustable controls.  Steve

[Ted Fancher]

From Brett's post: "Original = 36 oz at 4.6 sec laps, 89 fps, .069 slugs, 65 feet; line tension = 8.52 lb, q = 9.3 psf
  final = 44 oz at 5.2 sec laps, 78 fps, .086 slugs, 65 feet; line tension = 8.04 lb, q= 7.3 psf

   ratio of line tensions = .94, ratio of q = .79"


The subject Tucker Special has been ballasted up 8 oz and slowed from 4.6 sec. laps to 5.2 second laps. It will require a little more control input due to the increased weight, but due to the reduced speed/increased lap time, the hinge moments will be reduced quite a bit. It seemed to work; I was there.

One thing Gary Letsinger once told me was that sometimes the corner can be tightened by shifting the CG forward. Theory is this requires more control input, which brings more wing flap to bear. I suspect this would be only possible on heavy models and of course, modern adjustable controls would probably make this the wrong thing to do. But then, not everybody incorporates adjustable controls.  Steve

Steve,

I would think that could only be true if the ship was stalling with the aft CG.  A better solution to that dilemma is to increase the flap movement relative to the  elevator as I did with the  porky original Trivial Pursuit which still flies pretty respectably.  In hot, rainy air in Muncie at one team trials it started to do that (stall) in triangle bottoms and scared me to death.  That's when we first experimented with the .018 control line turbulator randomly taped to the high point of the wing as suggested by--guess who--Mr B.  The problem disappeared and the plane flew competitively but it wasn't the year it won the trials...don't ask me which it was .

Of course, the taped on wire looked tacky and I couldn't live with that so I moved the elevator pushrod out a 1/16 of an inch or so and there hasn't been a stall since (where is the knock on wood smiley???).  There was, by the way, almost no difference in input required that I can recall from doing so and no significant difference in the ability to fly corners as tight as I'm capable of.

Ted

[Ted Fancher]

Delete this if I muddle up this excellent thread.

Generally, the CG will be somewhere forward of the CL (center of lift). Moving the CG closer to the CL makes the airplane more maneuverable, needing less force to rotate the aircraft around the lateral axis, but also makes it less stable.

I would think that at some point, the increased maneuverability would be offset by the instability, leading to over-controlling attempting to fly smooth maneuvers.

Bill,
On a more or less conventional canard the CG "must" be well behind the wing (the front lifting surface...i.e. the canard).  As the tail grows larger and larger in comparison with the wing the aft most "stable" CG moves with it.  As the aft surface approaches the area of the front surface the useable CG range quickly moves aft with it.

The result for a normal stunt ship design is that, if you want, you could make the tail any size and be able to fly it with stability in level flight.  When you start to maneuver in the  pitch axis, however,  lift generated forward of the CG will accelerate the pitch change (positive pitching moment) and (I think) make stopping turns and flying constant radius corners and/or round maneuvers more demanding, especially in the wind as the control inputs to maintain a constant rate pitch change will vary as well...more or less the opposite of the "enhanced" stability we used to seek using forward CGs.  Ever seen  a competitive  aerobatic canard?

Ted Fancher

[Brett Buck]

One thing Gary Letsinger once told me was that sometimes the corner can be tightened by shifting the CG forward. Theory is this requires more control input, which brings more wing flap to bear. I suspect this would be only possible on heavy models and of course, modern adjustable controls would probably make this the wrong thing to do. But then, not everybody incorporates adjustable controls. 

  I mentioned both sides of this earlier in the thread. As far as I can tell, Al Rabe first figured this one out, it was in the Bearcat article. It does require more line tension, particularly before large bellcranks were common. You only need it on heavier models, of course, because the lighter ones don't run out of lift.

     Brett

[phil c]

  ....
     The heavy airplane ends up flying at a higher angle of attack for a given turn radius, but I think my ratio holds. The required pitch rate at "equilibrium" is actually lower on the heavier but slower airplane, you end up needing less deflection to achieve it. In fact the same theory can hold until your AoA in the corner exceeds the stall angle.

      Brett

Nice thing about flaps.  When the flaps go down the AOA of the wing automatically goes up, even if the AOA of the plane doesn't necessarily change.

Found an applicable quote from Brian Hamton:

"It would seem the smallest possible flap deflection to produce the desired lift would be the best solution.
I quite agree. The model I designed had the flaps as part of the airfoil section but also had fully adjustable (independent) movement from zero to around +/-30 degrees. With flaps set at zero there was a trace of stall in the last corners of both triangle and hourglass even with a very light wing loading of 10.25 ounce/sq foot. Flaps set at +/-5 degrees eliminated the stall and kept added drag to a minimum to minimise loss of airspeed in hard turns. Higher wing loading would have required a bit more flap movement on this model."

Brian has shown an unusual flap horn arrangement that allows the flaps and elevator movement to be adjusted independently with virtually no interaction.  Seems we need to be building better adjustable controls for trimming.  They would solve most of the problems discussed here.

[Ted Fancher]

  I mentioned both sides of this earlier in the thread. As far as I can tell, Al Rabe first figured this one out, it was in the Bearcat article. It does require more line tension, particularly before large bellcranks were common. You only need it on heavier models, of course, because the lighter ones don't run out of lift.

     Brett

Remember the pictures of Al's modified Hot Rock handles with up to six inch spacing to "increase leverage"?  Given we depend on line tension no matter how much mechanical advantage we think we gain at the handle would such extensions have particular value unless inputs are accompanied by an aggressive shortening of the lines (pulling back while rotating the handle) to achieve an advantage???  Sure seemed to work for Al but it kind of goes counter to this thread.  Also, Al used three inch bellcranks for most of his competitive career although I heard a rumor he might have gone to the now preferred four inch ones.

Anybody know??

Ted

[Howard Rush]

You can't get leverage by handle spacing alone.

[Brett Buck]

You can't get leverage by handle spacing alone.

  Also as mentioned above. Of course, if you move the CG forward, you will need to increase the spacing just to get the sensitivity back. You don't get any more torque at the bellcrank. Once the slack line is really slack, it might as well be disconnected.

     Brett