netzeband wall

I have read the July/August Stunt News article about the Netzeband Wall, but, I am not sure I understand what the article presented. I would appreciate any help or comments as to what was intended.
Thanks in advance.

LAST EDITED ON Aug-04-02 AT 09:30 PM (CDT)

>I have read the July/August Stunt
>News article about the Netzeband
>Wall, but, I am not
>sure I understand what the
>article presented. I would appreciate
>any help or comments as
>to what was intended.

It takes torque to deflect the controls. The available torque is limited by the line tension*. This means you can only deflect the controls so far, and thus can't turn it any tighter.

This is only an issue in a limited number of practical circumstances, where the tension is light, or where you have the CG far forward, large control surfaces, and/or small bellcranks. Larger bellcranks help this situation (see footnote).

Brett

*actually, the available torque = cosine(bellcrank angle)*line tension* half the bellcrank span, {hung on one line}

I have actually performed the Impact broken-line maneuver described in the article. I don't know about a wall, but the model did hit a floor.

What I'd like to know more about is how Charles Buffalano is doing the model simulation for the Impact for his plots. It looks like fantastic stuff.

For example, Figure 3 shows the Angle of Attack of the Elevator vs Elevator Deflection. The latter is a "steppy" type function and the former seems to show a little "ringing" on the transitions. (Kind of what I would expect..)

Also, the plots for the Triangle seem quite realistic (e.g. no 5' radius corners), and the bellcrank deflection is realistic at about 10 degrees.

Looks like quite a nice mathematical model.

Larry wrote:
> What I'd like to know more about is how Charles Buffalano
> is doing the model simulation for the Impact for his plots.

He had an article in Stunt news a while back talking about it.
Not too much detail. The aim of the article was to discuss
a way to estimate the drag of a plane by measuring it's
deceleration as it coasts in for a landing.

As a side note, he talked a bit about his simulation work.

>What I'd like to know more
>about is how Charles Buffalano
>is doing the model simulation
>for the Impact for his
>plots. It looks like fantastic
>stuff.
>
>For example, Figure 3 shows the
>Angle of Attack of the
>Elevator vs Elevator Deflection. The
>latter is a "steppy" type
>function and the former seems
>to show a little "ringing"
>on the transitions. (Kind of
>what I would expect..)
>
>Also, the plots for the Triangle
>seem quite realistic (e.g. no
>5' radius corners), and the
>bellcrank deflection is realistic at
>about 10 degrees.
>
>Looks like quite a nice mathematical
>model.

I'm sure he'd send you the scripts if you were interested.

Hi All,

First of all I would like to say that I do not think the world “wall” is the best – and if than “rubber wall”. It is because it is definitely not an absolute limit. We have still kinetic and potential energy of model mass inertia useful to make an input stronger than centrifugal force allows. It is not only theoretical possibility – many of use it. I will make an example. I have a spreadsheet calculating many important numbers of stunt models. Two of them are also line tension necessary for wanted corner radius (3.5 m – sorry for metric units) and available line tension - both overhead (worst place - hourglass). I will mention 3 known models as they are described on public building plans - Trivial pursuit – available 30N necessary 14N; Dreadnought – available 30N necessary 24N; Cardinal – available 30N necessary 29N. They are all of the same weight and lap time – but it does not play a role here. (the line tension as well as necessary line tension depends on weight and speed – both comes from centrifugal force out of some radius) So the example shows that while TP has a lot of reserve to do a corner in any cases, the Cardinal is almost on its border (wall?). The feeling from handle is different – TP is controlled by deflection and Cardinal by “impulse” – means – does not matter what deflection of handle is, you do the same radius. I know many pilots prefer such an impulse system of flight – they “kick” model to the corner. (I am not a member of that group) They are mainly older guys – probably those grown up on 3” bellcranks. But question was if it is really a limit. I say not. Here is an example: let us take such a Cardinal and try to do 0.5m tighter corner. There is not enough line tension, but we can pull handle little bit and use a potentional energy of the model hinging on tight lines. Here are data for such corner: The bellcrank must go to 37deg instead of 32 deg. Necessary line tension is over 33N what is 3N more than available 30N. The time for the 90deg corner is .2s instead of .25s. For that time you need energy coming from pull which will move the model only 7cm what is less than 3” (it is distance for what the tight line must go back – it includes the deflection angle itself, real pull takes only an inch) – it is easy to live with as the Cardinal needs an impulse controlling anyway. Question is how often we can do it – the recovery path to get tight lines (means accumulate potentional energy back) again is only 1m (3 ft) – it is much less than path to the next corner.

>So the example shows that
>while TP has a lot
>of reserve to do a
>corner in any cases, the
>Cardinal is almost on its
>border (wall?).

What is it about the TP design that allows it to cirner with so much less line tension?

Tail moment?

TP has less percentual flap to wing ratio 16% instead of 19%, so less lift is transferred to the pushrod. Also bellcrank to flap ratio on TP allows more bellcrank deflection - 55 deg instead of 32 deg for the same 3.5m corner, so lines moves more on TP compared to Cardinal – it asks for less differential tension.

>TP has less percentual flap to
>wing ratio 16% instead of
>19%, so less lift is
>transferred to the pushrod. Also
>bellcrank to flap ratio on
>TP allows more bellcrank deflection
>- 55 deg instead of
>32 deg for the same
>3.5m corner, so lines moves
>more on TP compared to
>Cardinal – it asks for
>less differential tension.
>
>igor

I don't see how that can be right. It appears that you have not taken the amount of deflection required to create a specific moment to turn the tail around the wing. If the ratio of the control throws were all that mattered, we would all have super slow controls and super short tail moments.

Are you also saying that flapless airplanes require less line tension to do the pattern? In practical terms I do not see how that can be true. We would all be flying flapless airplanes to accomplish a superior hourglass.

I have no problem believing that the TP may require less line tension, but not due to slow controls. Long tail moment, yes.

>Are you also saying that flapless
>airplanes require less line tension
>to do the pattern?
>In practical terms I do
>not see how that can
>be true. We would
>all be flying flapless airplanes
>to accomplish a superior hourglass.

I think you need to analyze Igor's data a bit more on the other topics, but simply turning tighter corners in the hourglass isn't the driving motivation to do anything. You can already turn tight enough with existing airplanes. The Netzeband wall, while an interesting phenomena, isn't really a factor with properly designed and trimmed modern airplanes, at least not the ones I would recommend.

I think whether very light flapless airplanes might be overall better stunt planes is a somewhat open question. I don't know the answer, but it won't be for Netzeband wall reasons.

Brett,Ted,

Brad and I were talking last night and have both seen this before; Ted takes off and gets a lean or fast run or the wind whips up hard in his flight. His pattern suffers NOT one bit.(I am sure alot of this is due to many hours learning to fly in all conditions) It doesnt look like he is getting control bind or nothing. It just goes faster but it still turns like a laser and has a very clean look. There is no forced look. We have been discussing what is the cuase of this. He(Brad) is under the impression that it is the tail moment that is causing this for the most part. The tail moment is very long creating stability, more power from the elevators with the same amount of deflection as in a shorter tail moment.

Is Brad correct in his observations? I also think there is something else to it. CG vs CL. The CG aft the CL. Will this help keep the controls smooth and free during high speed high wind flights?

OK, I understand, I use too much numbers instead of explanations, but you know – one number is easier to write than lot of letters …. … so here are those missing letters:

First of all – if we speak about two models of the equivalent weight, speed and corner diameter, than both wings must make EQUIVALENT LIFT – does not matter what is the wing area, linkage ratio of flap percentage, because the lift must be equal to centrifugal force out of the corner. You must properly pitch the model to necessary AoA to make that known lift. This pitching makes feedback to the lines what is transformed to the differential line tension. If that line tension exceeds available line tension what comes almost from centrifugal force out of the circle (that force is also depending on known equivalent weight of models and its speed). So the reserve in line tension depends on amount of feedback from pitching, not on line tension itself (which is equivalent).

The wing and elevator produce a lift in corner and both of them are applied on movable parts – flaps and elevator – what makes a torque on horn axles. So the feedback is that torque transferred to the force by linkage ratios. The question is what is the torque and what is the ratio. If we know lift and flap to wing ratio (percentage). Than we can estimate (by expecting constant air pressure distribution – it is not, but not too far from true) what is that feedback on flaps. In any case 20% flaps to wing ratio of first model takes 20% of lift back to the horn. 15% ratio of the second model makes 5% less feedback. AGAIN – does not matter what is the wing area – the second model will probably need little more wing area, but it does not change torque on flaps. And yes, it really does not matter what is the deflection of flaps - it depeneds on what lift the wing makes. The pressure distribution over the airfoil is also important (there are some peaks on LE, at hingeline and at TE) but not so much as the lift itself and you do not do too much error if you take it constant.

The elevator is much smaller and makes much less lift, so it is not so important, but in any case less flaps make less negative pitching of wing and thus less necessary elevator lift, so yes, it is true, less flaps makes pitching much easier and gives more reserve in line tension. And this is also reason why PT got more reserve.

Linkage ratio is another story. More angular movement of bellcrank and its longer arms gives less tension of lines necessary for the same torque on flap horns. It also gives more reserve.

BUT!!! It is everything only about PITCHING itself. It is not about real lift production. It gives you chance to pitch model easy, but to keep that wanted corner radius, you need also wing ABLE to produce that lift, so you must solve also another side of equation: wing with those small flaps must be still able to create that lift. It means it is always about some compromising and iterations.

Wouldn't aerodynamic balances on the control surfaces reduce the hinge moments and the torque required?

Yes it will, but question is if we want it. I personally prefer to have rather more feedback - it makes flying smoother if you have loaded hand. Anyway, any such a device hurts air around and it makes it unpredictable. I think flaps and elevator should end smoothly – without any tricks like balancers or fixed tip blocks – even if you consider that model never flies tangent – one or two degrees out of circle.
There are enough ways to get more line tension reserve. The Cardinal is very popular here in Europe just because it loads the handle. However the France almost flapless Suchoi with such balancers on elevator also works as we saw on WC, but I saw a guy who did such model and he was not able to fly with such model and he went back to his previous successful model. I think it is question of preferences.

LAST EDITED ON Aug-08-02 AT 12:58 PM (CDT)

To cut to the chase, the "wall" is the force which can be applied to the controls using centrifugal force as limited by the airplanes speed and weight. It is a highly theoritical, exotic, esoteric and largely useless exercise in logic. Our stunt ships are not limited by speed and weight considerations in our ability to apply control. There are a number of forces at work which increase line tension beyond weight and speed. "Wall" computations do not take into consideration any of the other factors which increase line pull such as engine offset, rudder offset, tipweight, fuselage side area, leadout location, torque, thrust, movable rudders and last, but far from least, whipping.
The only time I ever encounter the "wall", which is to say, run out of control, is starting down from the top of the vertical eight in a strong wind. This is a very tough spot, particularly the second repitition where the airplane is losing energy through consecutive vertical maneuvers involving hard climbs above 45 into the wind. When starting down, the airplane has lost much of its speed, is being pushed down by the tailwind, and what little centrifugal force remains is largely offset by gravity. With the airplane in an awkward location near the top of the circle effective whipping is nearly impossible. For me, this spot is more troublesome than the hourglass because I can add so much energy through whipping the entry, it is nearly posible to do the hourglass deadstick. OK, so that is an exageration, but you get the idea. Whipping can add a great deal of energy, and the hourglass has no second repitition.
In summary, anything which adds line tension beyond speed and weight push the "wall" into never-never land. This being the case, why the interest?
Al

lots of flyers have a terrible time with the outside loop of the overhead eight. The plane loses speed, and drifts way wide, down to 30 degrees or less because it doesn't have enough CF to move the controls. Especially bad in a wind. The second corner of the hourglass is a problem too.

Part of the problem on the vertical is that in a wind the plane picks up a huge boost from the wind and comes around the second half of the top loop much faster than the first half. There isn't time to react sometimes. Also, the plane is being blown down into the right hand side of the bottom loop. It can pick up so much speed that it can't make the turn unless it's started high, similar to the fourth turn of the hourglass.

One thought in addition to Al Rabe's post - The fuselage side area config can also produce very significant lift, outward, so as to increase line tension. I have a 15 year old flapless flying tail lightweight, 34 oz on 540 sq in. wing area. It is twin boom, with very minimal side area. Even though the stabilator is mounted on an axle at 20% of the chord, and is to that extent self actuating, the thing will still wallow and yaw if you pull too hard in a square or triangle corner. What it feels like is that the momentum does not carry it through the corner like a more heavily loaded, deep fuselage flapped ship, and the engine by itself can't keep the speed up. My theory when I built it was to have the 'power steering' effect of the stabilator aerodynamic counterbalance lessen the need for the side area, because if the surface was self actuating, you'd need less line tension to generate/overcome the turning moment at the elevator. That was supposed to mean that the Netzeband wall was further away, and you could get a harder turn before reaching it. It didn't turn out that way, although it might if we had a power train which would add a lot of torque in response to a drop in RPM. Maybe a RPM governed ignition system? Certainly looks as if the 2 strokes fail here, in that if overloaded, preignition robs the thrust just when we need it most. Anybody have any ideas, or should I be more careful to wear my hat out in the noonday sun? Tom Hampshire

Al,

My limited knowledge of English language does not allow to understand all nuances, but what you wrote in your last message:
>>>It is a highly theoretical, exotic, esoteric and largely useless exercise in logic.<<<

It is jewel in my eyes. I must to save it somewhere – it is just great!

But back to rubber wall: Yes I agree. As I already wrote it is definitely not an absolute limit and I also do not know how to enumerate its value (minimal radius?). But anyway if we will speak only about “reserve” between available and necessary line tension (what is much easier to understand and evaluate) and if we will simplify really only to the centrifugal force and ignore any other inputs, we can get at least overview what is going on and what we can expect out of some model. And what is most positive, we can do it BEFORE building and flying – and what is even more positive I can try 20 different models one evening without wasting any piece of balsa. It is not CAD, it is CATS – no it is not about girls – it is Computer Aided Time Saving.

Regarding your example with vertical figures – here I cannot agree with you. I do not think that the lost of line tension in vertical figures has what to do with this “Netzeband wall” effect. What you wrote is clear – the model loses speed until the gravity wins and model falls down. You can even help him somehow by whipping or stepping back or what ever to keep enough speed. But any of such actions does not change the minimal allowed radius – it is because the “wall” or “reserve” is speed invariant – because - if you slow down or speed up the “available” and “necessary” line tension grows together as they both depend on the same speed and weight. It means if you fly slower, you have less line tension, but you also need less line tension by the same amount (proportionally) for the same radius. The hourglass in wind is known example – I think everybody remember it – you can do second corner much easier (I mean impulse to the handle) just because model loses its speed – the model flies and turns even you feel only “microscopical” line tension. … Yes there is also gravity in game but I think we can dismiss it for simplification.

>LAST EDITED ON Aug-08-02
>AT 12:58 PM (CDT)
>
>To cut to the chase, the
>"wall" is the force which
>can be applied to the
>controls using centrifugal force as
>limited by the airplanes
>speed and weight. It
>is a highly theoritical, exotic,
>esoteric and largely useless exercise
>in logic. Our stunt
>ships are not limited by
>speed and weight considerations in
>our ability to apply control.
> There are a number
>of forces at work which
>increase line tension beyond weight
>and speed. "Wall" computations
>do not take into consideration
>any of the other factors
>which increase line pull such
>as engine offset, rudder offset,
>tipweight, fuselage side area, leadout
>location, torque, thrust, movable rudders
>and last, but far from
>least, whipping.
>The only time I ever encounter
>the "wall", which is to
>say, run out of control,
>is starting down from the
>top of the vertical eight
>in a strong wind.
>This is a very tough
>spot, particularly the second repitition
>where the airplane is losing
>energy through consecutive vertical maneuvers
>involving hard climbs above 45
>into the wind. When
>starting down, the airplane has
>lost much of its speed,
>is being pushed down by
>the tailwind, and what little
>centrifugal force remains is largely
>offset by gravity. With
>the airplane in an awkward
>location near the top of
>the circle effective whipping is
>nearly impossible. For me,
>this spot is more troublesome
>than the hourglass because I
>can add so much energy
>through whipping the entry, it
>is nearly posible to do
>the hourglass deadstick. OK,
>so that is an exageration,
>but you get the idea.
> Whipping can add a
>great deal of energy, and
>the hourglass has no second
>repitition.
>In summary, anything which adds line
>tension beyond speed and weight
>push the "wall" into never-never
>land. This being the
>case, why the interest?

I tend to agree that this isn't really all that much of a problem in the modern world, given decent trim and a reasonable CG position.

But one really good thing about discussing it (and a lot of other things that don't directly translate into a specific goal or action) is that trying to understand it leads to more general understanding. For instance, a lot of people have never considered the concept of hinge moment translating to line tension differential. It's not that they are dumb, it may have simply never occurred to them. I am sometimes surprised at the things that come up here I feel are really fundamental, that appear to be completely unknown or come as stunning revelations to others. It's not like you have to take a course and have to pass an exam to fly stunt! You just never know what might be new information. Even if it doesn't lead directly to anything.

>Brett,Ted,
>
>Brad and I were talking last
>night and have both seen
>this before; Ted takes off
>and gets a lean or
>fast run or the wind
>whips up hard in his
>flight. His pattern suffers
>NOT one bit.(I am sure
>alot of this is due
>to many hours learning to
>fly in all conditions)
>It doesnt look like he
>is getting control bind or
>nothing. It just goes
>faster but it still turns
>like a laser and has
>a very clean look.
>There is no forced look.
> We have been discussing
>what is the cuase of
>this. He(Brad) is under
>the impression that it is
>the tail moment that is
>causing this for the most
>part. The tail moment
>is very long creating stability,
>more power from the elevators
>with the same amount of
>deflection as in a shorter
>tail moment.
>
>Is Brad correct in his observations?
> I also think there
>is something else to it.
> CG vs CL.
>The CG aft the CL.
> Will this help keep
>the controls smooth and free
>during high speed high wind
>flights?
>
>DMoon

Doug and Brad,

First, thanks for the nice comments on my patterns. Not sure they are totally accurate but it's nice to hear such things (even if there's a bit of "smoke blowing" going on).

There is some truth to all of what you say. The most important, in my mind, regarding maintaining flyability in windy or fast run conditions are the control geometry and the CG/CL relationship.

Most relevant to this thread's discussion is the control geometry. Whatever you want to call it...the wall is as good a name as any...our ability to fly maneuvers with a tethered c/l type airplane is dependent on our ability to deflect the control surfaces an amount appropriate to the radius of pitch change desired. No matter what esoteric assumptions are made about whipping and out thrust etc. the foundation material from which the deflecting flaps and elevators is accomplished is the result of having enough mechanical advantage built into the control system (lots of bellcrank rotation for a smaller amount of control deflection)to allow us to deflect them the desired amount even when line tension is not so great (top of the circle, vertical eights, hourglasses, etc.).

The overiding secret, therefore, of great patterns in varying conditions is a control system which functions more or less perfectly even when the conditions ain't so great. This use of larger radius slower repsonse control systems is probably the single greatest advance in stunt model design and engineering since the boys discovered the various virtues of flaps many moons ago (no lunar puns intended).

This is only half the story, however, regarding the ability to fly corners well when the wind blows hard or the flight speed is greater than you might actually prefer. You are right, the Trivial Pursuit series is quite admirable in such conditions. The reasons why are basically what you addressed in your question.

Modern stunt ships are noteworthy and distinctively different from their predecessors in that they have longer tail moments and large tails (as a percentage of the wing area) than before. These design factors make the tail very powerful in influencing the direction in which the airplane flies (pretty much the whole idea).

One way of using that available force is what many did back in the early .46 to .60 years, Jimmy Casale is a perfect example. The very powerful tail was seen as a means whereby a ship could be trimmed more nose heavy and through brute force very tight corners still could be flown. The thought was that this would increase line tension and stability while retaining the tight corners for which the Casale pattern was well and properly respected.

The problem was that having the CG and the Center of Lift so far apart as a result of the forward CG that when the airplane accelerated as in the wind or simply flew faster by virtue of a bad engine run the ability to fly those same tight corners was compromised.

Two things happened. In those days Jimmy still used three inch B/Cs and short horns and very fast control systems. Under ideal conditions such systems were at the ragged edge of their limits...pretty much what the "wall" and Igor's more detailed discussions are all about. When the ship flew faster the control loads on the airplane were increased simply by virtue of the increased forces on the deflected surfaces.

To compound the problem, with the CG and the CL far apart when the ship starts to fly faster it also wants to open up the size of the maneuver it is performing. This happens because the accelerated weight of the entire airplane is increased as a function of the number of "Gs" the ship is pulling. This increased weight (a ten G corner would increase the load from the static four pounds in one G flight to a full "40#") acts at the CG. If the lift to support this 40 pounds is centered and inch or two aft of the CG this "couple" is a 40# force trying to stop the pitch change...a gruesomely large negative pitching moment!

This results in the demand for greater control deflection to maintain the desired radius which in turn demands greater control forces...which may not, in fact, be available through the control system. Very much a case of a dog chasing its own tail.

The better way (and the way the Trivial Pursuit was designed to utilize) was to use a further aft design CG (ideally at the theoretical location of the Aerodynamic Center of the wing, 25% of the MAC). Doing so eliminates the tendency in accelerating flight for the maneuvers to open themselves up simply because there is no "couple" between the CG and the CL. It simply doesn't take a lot more control deflection to perform the turn you want regardless of the speed of the ship. Please note that the G loads themselves will result in the need for a higher angle of attack and greater control deflection to produce the lift required. Thus, corners in bad conditions will still require more deflection. This is much more moderate however and isn't the result of the aerodynamics of the control system fighting the pilot!

When combined with the greater mechanical advantage of the modern control systems the inputs required for good cornering in high speed conditions are entirely doable and never require significantly increased amounts of muscle.

The result is a ship which can handle extremes with equanimity. Low tension? No problem, the control system doesn't need much tension in order to deflect the controls enough for any corner you might need...too much control is a much more likely problem than too little.

High speeds or high line tension in an accelerating environment? Still no problem. The ship's weight/lift relationship is such that it doesn't fight you and the control system has more than enough mechanical advantage to deflect the control surfaces as far as necessary to overcome the G induced tendency to open up corners. You still gotta hang on harder since the line tension increases as much as any ship with increased centrifugal force. The difference is that the biased control loading to obtain control deflection isn't significantly greater than that required under ideal conditions. Ergo, the ship turns pretty much as well under bad conditions as ideal.

Does any of this make sense???

LAST EDITED ON Aug-08-02 AT 08:40 PM (CDT)

>
> It didn't turn out
>that way, although it might
>if we had a power
>train which would add a
>lot of torque in response
>to a drop in RPM.
> Maybe a RPM governed
>ignition system? Certainly looks
>as if the 2 strokes
>fail here, in that if
>overloaded, preignition robs the thrust
>just when we need it
>most.

Huh? Maybe if it's already peaked out - then it sags. But adding load to a rich-running 2-stroke adds a tremendous amount of torque, and in fact causes the 4-2 break. Advancing ignition, if it's already severely retarded (as happens with a very rich setting on a 2-stroke) greatly increases the torque. And the whole purpose of a tuned pipe in stunt is to create a very steeply descending torque curve (which is exactly what you describe as desirable) at the operating point.

The only time your description applies is when the engine is already peaked out and sags lean, or the pipe is set too short and the engine drops on the low -rev side of the tuning peak. Or on a 4-stroke that's already peaked out, for the exact reason you describe.
Brett

>Does any of this make sense???
>
>
>Ted

Yes indeedy.

I have looked at all of my plans and you have the longest tail moment of all of the plans I have.

What are the advantages of the VERY long tail moment?

>Al,
>
>My limited knowledge of English language
>does not allow to understand
>all nuances, but what you
>wrote in your last message:
>
>>>>It is a highly theoretical, exotic, esoteric and largely useless exercise in logic.<<<
>
>It is jewel in my eyes.
>I must to save it
>somewhere – it is just
>great!

Igor, you beat me to the punch. I was going to post the EXACT same remark. AL is great, isn't he? You should talk to him in person. He always has the best dirty jokes.

Classic Al...

It's
>not like you have to
>take a course and have
>to pass an exam to
>fly stunt! You just
>never know what might be
>new information. Even if it
>doesn't lead directly to anything.
>
>
> Brett

Guys, this is a great topic, and thanks to all who have resonded and or asked questions. It seems to me that an understanding of the forces at work here will help anyone who designs their own planes to get closer to what they want.

A well designed control system is the key to a winning plane as Brett, Ted, Igor, and others have stated. It also eases the trimming process ten fold, because the plane flies truer. Coupled with a great powertrain, and almost any reasonable design, built true, can be a winner as well.

Again, well done guys and gals. There's lots of good information on this and several other threads here at SSW.

>High speeds or high line tension
>in an accelerating environment?
>Still no problem. The
>ship's weight/lift relationship is such
>that it doesn't fight you
>and the control system has
>more than enough mechanical advantage
>to deflect the control surfaces
>as far as necessary to
>overcome the G induced tendency
>to open up corners.
>You still gotta hang on
>harder since the line tension
>increases as much as any
>ship with increased centrifugal force.
> The difference is that
>the biased control loading to
>obtain control deflection isn't significantly
>greater than that required under
>ideal conditions. Ergo, the
>ship turns pretty much as
>well under bad conditions as
>ideal.

I think you are right as far as you go. But at the danger of being accused of being a sycophant, it takes some pretty impressive reflexes and pretty impressive hand-eye coordination to take advantage of the airplane's ability in these situations. I've watched you fly for a lot of years, and my airplanes have probably handled these sorts of situations as well as anyone's, I have never come close to matching you consistently in these sorts of conditions.

I think the airplane is certainly part of it, but I think a lot of what Doug and Brad see is the same thing I see on a regular basis - some pretty good performance at the other end of the lines, too.

LAST EDITED ON Aug-09-02 AT 00:15 AM (CDT)

>
>>Does any of this make sense???
>>
>>
>>Ted
>
>
>Yes indeedy.
>
>I have looked at all of
>my plans and you have
>the longest tail moment of
>all of the plans I
>have.
>
>What are the advantages of the
>VERY long tail moment?

Ted's are right in the middle of the range compared to a lot of other (dare I say it) West-coast-style designs. Too short is too short, right is right, and too long is too long. I have had the tail moment from 16" (standard Genesis) to almost 22 (first Infinity) and the best was the ~18" on the current Infinity. This is, coincidentally, about the same as Ted's recent designs (Temptation and Trivial Pursuit/Great Expectations/Star Gazer).

For fear of expelling more useless theorizing, longer tails:

make the damping (torques proportional or related to the pitch *rate*) higher (tail AoA drops more quickly with rate)

make the peak pitch rate lower (tail AoA goes to zero at lower rate)

make the angular acceleration higher (but not close to proportional due to moment of inertia reasons)

In practical terms the only consequential difference in Infinity #1 and Infinity #2 was that the tail moment got shortened 2.5" on the second one. The first one flew OK, but the response was too "regressive" even for my tastes. This means it was really sensitive around neutral, and the tremendous acceleration would easily delineate the straight/round sections, but it really didn't respond all that great through the whole corner. It always looked kind of awkward. The second one is a much better airplane, still strongly regressive but not awkward at all.

Brett

p.s. it's interesting to note that while we frequently talk about stuff like tail moments as some sort of "magic performance number", the actual performance difference from close to arbitrarily changing it by 2-3" with everything else the same isn't all that dramatic as long as it's long enough to begin with. A little too short is a lot worse than way too long. I think Bruce Perry's latest model is as long as they have gotten at something like 22", and while it may or not be optimal, it's certainly perfectly flyable and competitive.

Having just gone through the control thing in the wind, I can atest that having the CG at 25% really makes the ship turn easier in high wind. Ted is correct in mentioning that the line tension does get higher, but you can still turn the ship with reasonable handle forces. Howard mentioned to me that the vertical fin can be used to help with line tension upwind and downwind.

>Having just gone through the control
>thing in the wind, I
>can atest that having the
>CG at 25% really makes
>the ship turn easier in
>high wind. Ted is
>correct in mentioning that the
>line tension does get higher,
>but you can still turn
>the ship with reasonable handle
>forces. Howard mentioned to me
>that the vertical fin can
>be used to help with
>line tension upwind and downwind.

Just out of curiosity, how the trim changes you were going to try work out?

I am trying to design my own stunt plane. I came here to read all the esoteric stuff. I just love it when you guys talk dirty. It seems the answer lies in two directions. Build and fly the heck out a proven design, or go to hydraulic controls. We already fly by wire. a little humor there, darn little.

Ted,

I like to see everything in clear numbers so I will try to convert what you wrote to more theoretical view. The Netzeband wall just says what we can do with the model, but it does not tell what it costs us on side of handle - it is different story. You wrote that CG at 25% MAC does not make accessive negative pitching moment which must be overballanced by tail. It is not 100% true, because we have another sources of negative pitching momet, but for simplyfying we can count that it is really zero. So the tail does not make a lift in staedy radius. It means that if we accelerat it, we do not need to deflect elevator anymore. It is different if you have CG at say 20%. In this case the negative pitching moment from CG front of AC grows with square of speed while lift of deflected evevator (which is already angled little to counterballance CG) grows only linearly with speed – so we need to deflect it more. Until now is, I think, everything clear, but now comes the trouble – the feedback does not come form DEFELCTION itself – it comes from LIFT of wing and elevator – and that lift grows with acceleratig. It grows much less if CG is at 25% instead of 20%, but grows and hand feels it. It means that CG at 25% is definitelly better to keep flying similar in wind and calm, but it definitelly not enough to suppress growing of those forces at all.

But I know trick how to supress it at all. Guess what – yes it is my logarithmic device . I used it on my MaxII which I used on WC in Sebnitz. It has now at least 150 flights and I ended up with CG at 12% MAC. It should have all those symptoms of troubles in wind, but it does not. It is becouse of that negative pitching moment which must be counterballanced by tail. The trick is that if I deflect the elevator little more in wind, the logarithmick unit on flaps goes to place where it transfers much less feedback from flaps back to the bellcrank.

Brett - You are right in that if the needle setting is right, the torque goes up in response to added load. My findings with a low side area flapless lightweight seem to be that torque doesn't go up fast enough to hold the speed up, and the airplane slows below desirable flying speed. This is what mkes me think that the the heavier ships can be said to make it through a hard corner on a combination of thrust and momentum. The only way I can think of to measure this is with either high speed film/sound or a high speed tape recorder, so as to be able to time the individual exhaust pulses. Has anybody ever tried this? Tom Hampshire

Lift of deflected elevator grows only linearly with speed?

Stay tuned.

I heard the Netzeband Wall was torn down shortly after
the Berlin Wall, but you can still get bricks on eBay.

Oh no. It just occurred to me that the Expocrank will exacerbate the Netzband Wall. By increasing control sensitivies at larger flap deflections, it will transfer more force to the controls. So I'll hit that wall sooner.

Oh my.
[photo not recovered: 3d53dc8601dc815b.jpg]
L.

HUPSSSSSS....

Nice to see that somebody reads what I write here ... but now I am asking myself ... hmmmm ... and Ted also ... something goes wrong here ... ... OK I am going to do the homework ... so wait little bit .... sorry .... I am back soon ...

Sorry for that mistake again. Now I did what I had to do before. I gave holidays to my brain and I employed my simulator. I modeled 3 different models. TP with CG at 24%, CG at 20% and Cardinal at 20%. I tested both at 5.5s per lap and 5 s per lap. I modeled also a virtual handle with line spacing asking for 30 deg deflections to the standard 3.5m corner radius (means more lines separation for TP) on both models. Here are data:

TP24% at 5.5s lap – 1.24Nm
TP20% at 5.5s lap – 1.47Nm
Cardinal20% at 5.5s lap – 1.55Nm

TP24% at 5s lap – 1.5Nm
TP20% at 5s lap – 1.79Nm
Cardinal20% at 5s lap – 1.88Nm

The value “1.24Nm” means torque (momentum) on that virtual handle – sorry again for Metric units, but absolute value does not mean anything. Most important is, that the value grows by the same ratio: 20% if speed is 5s per lap. And yes – as Howard noted – the deflection is nonvariant – sorry again for my mistake and thanx Howard, so handle must deflected in both cases to 30 deg – does not matter what the speed is. (the simulation calculate also with airfoil negative pitching moment)

So the conclusion is, that 25% CG position does not bring any qualitative difference. Its advantage is only in smaller absolute value of feedback and thus – also those 20% of smaller value is smaller.

But I want note that it is everything only theoretical simulation expecting instant elevator deflection and circular path in corner. It is not true and so thanx limited speed of handle movement and higher speed of model we will after all end up at higher elevator deflection anyway. (… hmmm… yes … most important is to find acceptable theory )

I did some calculations several years back to try to understand why some of my planes wouldn't turn, and I discovered that it is possible to run out of line tension under certain circumstances with a 3" bellcrank. Running out of line tension is basically the same as hitting the Netzbend wall.
For 10 years I have been using 4" bellcranks and even larger on my Sig Magnum.
I gave a talk about it at the Knights of the Round Circle meeting a few years ago but never published it beyond. I will try to retrieve this article and send it in. It does involve calculus so it's not light reading.
Safe to say that all my future 60 powered ships will have a 4.5" or 5" bellcrank it them, whatever will fit in.
Happy Landings,
John

Brett,

Thank you. Coming from you, I value that compliment a great deal.

It also occurs to me that I might not have been as gracious as I might have in response to Doug's original nice words. I do appreciate the kindness and if I was flippant in response I apologize to both Doug and 'zilla.

Igor,

Yes, it is true that we must deal with negative pitching moments from other sources other than CG/AC relationship. Specifically the pitching moment from the cambered airfoil created when the flaps are deflected. I've spent some bit of time with that subject on another thread lately.

It is my feeling that this moment is primarily stabilizing to our control inputs...loading the controls to make it easier to fine tune our inputs. It does exist and must be accounted for in our control inputs but it doesn't present nearly the problem in accelerated conditions that a displaced CG/AC condition does.

I also respect that increased Gs will require greater flap deflection and will thus increase the camber generated pitching moment a measureable amount. Although I can't quantify it mathematically my experience tells me it is of a much lesser magnitude than that generated by an improperly located CG. The additional load on the controls (what is felt by the pilot) is nominal and not an issue with high mechanical advantage control systems.

I was a little imprecise when I spoke in the earlier comments about the additional loading caused by accelerated G forces. Not only does the exampled 10 G increase the lift requirement to 40#, the fact that the load might be displaced by as large a distance as, say, two inches in a dramatic case would result in a pitching moment of 80 oz/in in opposition to our desired direction of turn. This huge force must be entirely overcome by the tail surfaces before "any" pitch in the desired direction can take place!

I can't do the math (arithmetically challenged), but I can figure out that that's asking a lot of the tail surfaces!

Yes Ted, I did yesterday evening same calculation for several models also, and it really looks that if model accelerates from 5.5s per lap, the moment on handle necessary for the same corner grows by 20% up. Does not matter what is original moment and also CG position and linkage ratio or flaps size. (it is clear – it is because it comes from lift: 5,5^2/5^2=1,21) It means that if your TP with CG at 24% has small necessary moment and you speed it up, the 20% more moment can be still under problems, while another model which is pretty flyable in calm could come to serious troubles in wind. Paul with his B17 can probably comment it better .

I remember a thread year or two ago about lifting tail. There was a question how far we can go back with CG. The problem was that lifting tail would make problems. I believe yes, because a mass inertia will ask for negative lift of tail anyway (sometimes – when it accelerates pitching) and if stab goes from positive to negative lift it could be a problem with smoothness before, inside and after the maneuver - because no one can make 100% precise stab. The idea was, that if we use flaps, the pitching moment needs to be counterbalanced by tail even if CG is aft of 25% MAC thus the tail is still not lifting. So I tried to find the place where is that “zero” point. The result is surprising the pitching moment is so small that even 26% is no more safe. Here are some numbers – no more surprising anymore:

All for TP at 5.5s per lap:
-Negative pitching moment of wing is close to 1Nm
-Negative pitching moment of CG to AC if CG is at 24% (means 1% front of AC) is also close to 1 Nm
-Negative pitching moment of CG to AC if CG is at 20% (means 5% front of AC) is 5x more so 5 Nm

So now it is clear that aerodynamic pitching moment is really not important, so enlarging of flaps which I heard “leads to soft corners because of too much negative pitching of wing” is far from true. The truth is that it is because the feedback which is almost linear to the size of flaps. Just to open eyes here are another numbers:

The elevator pushrod of TP is stressed to only 3N
The flaps pushrod of TP is stressed to 34N
The deflection of elevator is 26.5 deg

It means flaps make 10x more feedback.

Here is same table for TP with CG at 20%:
The elevator pushrod is stressed to 6N
The flaps pushrod is stressed to 38N
The deflection of elevator is 26.5 deg

It shows that feedback from flaps is almost same (nonvariant to CG position), while feedback from elevator is 2x more (even only small). But you must count it needs more elevator deflection, handle will feel it much more than only those 4 Nm. Real value could be (thanx shorter flap horn arm to have same bellcrank and handle deflection for the same corner) 38*28/26.5=40Nm –> what is significant difference to 34N actually 20% more.

To make a conclusion: The CG at 24% is good way in my eyes (even I have better toy now). You are right that it needs more tail area. My Max (first version www.netax.sk/hexoft/stunt/images/mRim00014.jpg) was of the same idea. Instead of small line separation on handle I used much smaller elevator than commonly used. It was only 37% of tail area (27% of wing area at 18.8” tail moment). It was to allow flaps to go to 30 deg in corner. Later version was reworked to 40% elevator. That version was flown by German junior Richard Kornmeier on WC in Sebnitz (www.slovanet.sk/orsia/200207_008.jpg)

igor

Again – sorry for metric units, but I think absolute value is not so important