Normal horizontal flight and inverse horizontal flight differences

[Matt Piatkowski]

Hello,
One of my models exhibits a very good and stable INVERSE horizontal flight - the inverse horizontal flight path tracking is excellent.

The same model does not show the same quality of tracking in normal horizontal flight. It needs the pitch angle mini-corrections to stay in the horizontal path. 

What to check, modify, and/or adjust to make the normal and inverse horizontal flight paths tracking identical?

The model is not mirror imaged w/r to the X-Y plane of the model's Cartesian Coordinate System with 0/0/0 in the model's C.G.

Happy Flying,
M


 

[Brett Buck]

Hello,
One of my models exhibits a very good and stable INVERSE horizontal flight - the inverse horizontal flight path tracking is excellent.

The same model does not show the same quality of tracking in normal horizontal flight. It needs the pitch angle mini-corrections to stay in the horizontal path. 

What to check, modify, and/or adjust to make the normal and inverse horizontal flight paths tracking identical?

The model is not mirror imaged w/r to the X-Y plane of the model's Cartesian Coordinate System with 0/0/0 in the model's C.G.

Happy Flying,
M


 

      I presume that the hinge lines are sealed, and if not, go seal the hinge lines and start over. If its an offset-hinge thing, then you can't do that, you are stuck with whatever you have and whatever differential flow it causes -  which is a good reason not to use offset hinges.

      In any case, it probably needs the flap/elevator neutral shifted slightly. From just the description, I would *guess* it needs to go down, so adjust the pushrod length for one turn of "down" elevator with the flap at neutral, see if that makes it better or worse. If better, adjust more the same way until it goes away or is unstable in inverted, and then center the adjustment in the "good" range. If worse, adjust it the other way, same deal, find the setting that makes it the same either way.

    Most people use 4-40 control hardware, and typically 1" or 1 1/4" lever arm length on the elevator, so that will adjust it by about 1.4 degrees, or maybe 3/32" at the trailing edge of the elevator. That's a pretty good-sized adjustment, so it will definitely do something, and you should be able to detect the effects and whether it made the problem better, worse, or causes no change (unlikely).

    Usually, I can see whether the airplane is flying nose-up or nose-down upright an inverted, and that will usually tell you which way it needs to go. Of course, the airplane has to fly very slightly nose-up with respect to the ground both ways, and usually it turns equally and is equally stable in either direction when that nose-up angle is more-or-less the same upright and inverted, unless the airfoil is asymmetrical, or there is some other misalignment .

      It's possible (or even likely), that this elevator adjustment is correcting for some other misalignment somewhere else. That is inevitable, you cannot measure or maintain perfect alignment with typical modeling materials, and everything is flexing all the time, and distorting from thermal and other shrinkage or expansion effects.

    Some design features tend to make the airplane more sensitive to these sorts of issues, particularly the stab airfoil, which has been the topic of extensive discussion and argument over the years. But you have what you have now, adjust it for the best performance, and then see what happens.

      Brett

     

[Tim Wescott]

Brett scooped me, but I'm going to leave all these lovely words standing.

Trimmed level flight is at a different position for upright and inverted, with more variables in play for a plane that's not 0-0-0.

The first thing I'd check for is a bind in the controls around neutral -- if there's a bind around neutral in upright level but not inverted, then the controls will never settle there, so you'll always have to control around that spot.  I think you fly at a pretty high level, so it doesn't have to be obvious -- just a small sticky spot, or one place where it barely goes "tick" right around where things need to be in upright level flight.

Next -- are all your hinge gaps sealed?  Unsealed gaps do weird things as the airflow changes, and the bias on the stab will be different from upright to inverted.

Next -- does it fly level in both upright and inverted?  If it's nose up both ways you need more flap/elevator; if it's nose down both ways you need less; if it's nose up in upright and nose down in inverted then the flaps need to trail down a bit more, if it's nose down in upright and nose up in inverted then the flaps need to trail down a bit less (for a typical wing-lower tail-higher plane the flaps need to trail down a bit).

After I did those checks, I'm write a plaintive post here, hoping that some expert would jump in and help out!  While I was waiting for an answer I'd check for other asymmetries, like the stabilizer leading edge not being symmetrical, any twist in any surfaces (I think you'd have already checked -- but still.  Don't just check the wings -- check the flaps, the stab, and the elevators), and whether you have any up- or down-thrust.

I think for a typical wing-lower tail-higher plane you want just a hair of down thrust, if any.  But that's getting outside of my expertise.  Again, it's something I'd check (and "correct", to between 0 to 1 degree of down) while waiting for an expert to tell me what works for them.

Follow up with a picture of your stab leading edge -- the west coast USians feel that a pointy leading edge leads to less hunting in level flight; if you've got a rounded leading edge that hunting effect may only be induced in upright.  They're calling out a 1/16 to 1/8" diameter radius on the leading edge (Brett Buck reports that what convinced him was taping 1/16" wire to his leading edge).

[Tim Wescott]

... Of course, the airplane has to fly very slightly nose-up with respect to the ground both ways, ...

I haven't done the experiment, but Paul Walker's trim chart certainly implies (and my in-brain simulations agree) that with a high enough flap/elevator ratio the airplane will "plow" nose-down in both directions, as the flaps induce enough lift that the wing needs to be pointed downward to keep the plane level.

Are you saying this doesn't happen, or that such a plane would fly badly enough that no sensible pilot would get there in their trimming?

[Matt Piatkowski]

I am pleased to inform that I have solved the tracking problem in the normal horizontal flight.
The root cause: the elevator's horn brass hinges were loose and horn could move slightly aft and back while the elevator was moved.  The horn itself is not glued in the elevator's holes and this amplified the effect. I could not see this aft and back movement as it was very small...less than one millimeter (!!). This linear movement caused, however, uncontrolled changes of the elevator's angles around the desired position - something like "fluttering" of the elevator.

The elevator's horn hinges were glued with epoxy four years ago and one of two things happened: either the epoxy was "bad" or it failed due to vibrations.   

Comment: I have fixed the problem but I still do not understand why only the horizontal normal flight was affected. Perhaps the lack of the model's X-Y plane symmetry indeed caused such differences in the airflow field that the "flutter" manifested itself only in the normal horizontal flight and the inverted was good?
We will never know for sure but the effect counts: eParrot is a competition level stunt model now.

Regards,
M

[Ken Culbertson]

I am pleased to inform that I have solved the tracking problem in the normal horizontal flight.
The root cause: the elevator's horn brass hinges were loose and horn could move slightly aft and back while the elevator was moved.  The horn itself is not glued in the elevator's holes and this amplified the effect. I could not see this aft and back movement as it was very small...less than one millimeter (!!). This linear movement caused, however, uncontrolled changes of the elevator's angles around the desired position - something like "fluttering" of the elevator.

The elevator's horn hinges were glued with epoxy four years ago and one of two things happened: either the epoxy was "bad" or it failed due to vibrations.   

Comment: I have fixed the problem but I still do not understand why only the horizontal normal flight was affected. Perhaps the lack of the model's X-Y plane symmetry indeed caused such differences in the airflow field that the "flutter" manifested itself only in the normal horizontal flight and the inverted was good?
We will never know for sure but the effect counts: eParrot is a competition level stunt model now.

Regards,
M

Matt:

I would like to know if you ever find the cause of the differential,  Is it possible the flutter was asymmetrical?  Reason I ask is that we cured hunting back in the 70's by adding slop in the elevator, normally about 1/8"  However it was symmetrical.  NO, I don't do that anymore!

It is very possible that you needed enough control to track inverted to consume the slop but not so in level flight.

Ken

[Air Ministry .]

Some people say downthrust is neccessary .

Two with diheadral , when you go inverted half asleep , you snap awake smartly.
Think its the lift at the tips, inerted- supporting below vertical C.G. making it more sensitive .

Weaves with overcorrection , but if you ' let it fly ' , grooves nicely .
Co-incedently , both are high on the priorty list - as to what to build ( or finish ) next . ( Hurricane and big Mewgull ) .

Building in some assymetry can give it a ' bias ' to set against ( tends to ere one way ) for differental  stability . Touchier one way up .

[Matt Piatkowski]

Hi Ken,
In the 70's the stunt models were different than the top models now.
Current models of the top stunt/F2B pilots have the C.G and the centre of lift (COL) more aft.
Current models of the top stunt/F2B pilots are more pitch sensitive and more difficult to fly level because all top people need the maximum sensitivity to make the pivoting, very tight corners.

I am not even mentioning here the flight dynamics differences between the models equipped with ICE and the models equipped with the electric motors.

The stunt models from the 70's, being generally nose-heavy when compared to the current top models had "naturally occurring good tracking". Slop in the elevator made controlling the pitch in such models more forgiving...the pitch input mistakes were "softened" and the model kept merrily tracking.

This is not the case in modern stunt/F2B competition. Top fifteen fly the models that are very difficult to fly because the pitch sensitivity reached the maximum possible for each individual flyer. These models were laboriously trimmed and sit now in a very narrow trim range (ref: Paul Walker's TrimFlow Manual).

I wanted eParrot to be "narrow trimmed" and it was done.

What cannot be done easily is to convert eParrot to the fully X-Y symmetrical shape - it does not make sense to spend time to modify this model - it is better to build a new one having almost 100% X-Y symmetry (retractable L.G will be included).

Such monstrosity is in my plans for the next winter.

Happy Flying,
M
 

[Air Ministry .]

Current models of the top stunt/F2B pilots are more pitch sensitive and more difficult to fly level because all top people need the maximum sensitivity to make the pivoting, very tight corners.

Theyll have to fly one of THESE , Matt. A baby could fly it & it'll still outurn a Yatsenko . ! 





Its a variation of the ' HARA ' ( incorrect ) drawing in the Aeromodellor, in the 80s , at 112% and + 1 1/2 tail moment - stops the exit bounce . Airfoils ' As Sketch ' & does not match Harra true plan ( See N-30 - HARA )

Intesting DIXONs KA10/16 Plan , in compaison to Kolesnikovs complex original . Particularly on AIRCRAFT simplicity oftens lighter .
Not to mention taking half the time to build . If the Weight Matches ( or betters ) The simplistic one wins. Perhaps .

DONT be blinded by the latest is the bestest theory .Theres U tube of a 650 ( Pre 1971 ) Triumph 650 outcornering the latest 250 rice burners . On Track .

Evolution often goes in circles. Or is that spirals . Traction gives Adhesion = Cornering . None of this Japanese Flexing , on the bumps . 

Further too & insofar , Bob Hunt was/is  a firm advocate of NO elevator freeplay, perhaps his schnurle foam wing ' 5 footers ' such as the Genisis & the derivates published in F.M.
are the bigining of the ' modern ' trend, in the West . !?
Though the odd one specifies a thicker stab. than elevator .

checked the blue one above , I have used 5 m.m. at the trailing edge - slack . So the FLAPS may move without disturbing the elevators / pitch , initially .
I find the flat rear flank wing sections ' groove ' in round manouvres well .
A accidental incorporation of elevator free play flew/ grooved with C.G. 3/4 in AFT of plan designation . removing the slack had it skiving off all over the place .

the thickish airfoiled tailplanes seem to give large aft C.g. capeability , even with NO SLACK . For a good tight square corner .this ones 20 mm deep at root , 16 + at tip .



========================================================================================================

The REPORTS on this Adamisin Special all speak of very tight corners & good groove . He mentions enourmous C.G range variability . If your after ' Something New ' .



Post on " Tri - Canard " is here : https://stunthanger.com/smf/stunt-design/arch-adamisin%27s-tri-canard/

[Matt Piatkowski]

Hi Air Ministry and thank you for the photos.
The canard configuration is an interesting idea. The additional pitching moment from the forewing is, of course, desirable for corners but the downwash affecting the main wing is not.

On the other hand, the forewing sits in a highly turbulent flow behind the propeller that affects the main wing anyway...
Very tight corners & good groove sounds tempting though...

Re: horizontal stabilizer thick and sharp LE airfoil: Thundergazer has it. Max Bee II has it but less thick, Russian and Ukrainian RTF F2B models have it to a different degree.... The pressure distribution for such airfoils is such, that the centre of lift (COL) is moved slightly aft w/r to the, still typical, blunt LE and basically flat stabilizers. This COL shift increases the tail moment arm, assuming the wings COL stays in the same place. The sharpness of the stabilizer LE ( ref: Tim's latest remark: " the west coast USians feel that a pointy leading edge leads to less hunting in level flight..") effect is presented in the attachment and supported by the JavaFoil simulation I did.

In my never-ending quest to understand and improve, I will build several new, sharp LE (to a different degree) and thick (to a different degree) horizontal empennages for three of my full-size stunt/F2B models that are take-apart and flight test them one by one.

I feel I need a full-time assistant as my to-do list is growing at an alarming rate but she must be a good Salsa dancer.
Stay Safe and Fly Safely,
M

[Air Ministry .]

Yep , the ' traditional ' leading edge -> 1/4 Sq on edge , 1/16th sheet .

Vertical point through ' diamond ' = depth at max L E ( structure ) depth .

SO , if we use THAT -> halved as a radius for L.E. Radius , thats what we get .

Shaves about 1.5 m.m. off the front of the 1/4 Sq.
--------------------------------------------------------------------------------------

Foreplanes ! A semi Scale Hawker Hurricane , Following the I.A.I. Kfir concept ,  , I used the semi scale Exhaust manifolds - about 16 mm wide , 5 mm deep , 80 mm long , as foreplanes .
The plane ( First flight , OS 40 FSR-S , Dusk , a stready 18 knots with sturdy excess gusts , way aft C.G. , Actually Control reversal after the engine cut - so the helicopter landing was ' accidental )
further flights had a curious ' nose ' tweak , down & left . Which didnt seem to be adverse to ' the pattern ' .
Curious , I eyed it up from head on . the inner ' foreplane ( ex. manifold ) was canted say 3 degrees to horizontal - fore & aft . Cut and refitted straight - twist tendancy was eliminated .

So, a small- 12 Sq Cm area , ' aerodynamic surface ' forward - would seem to have descernable influance on trim . Presume ' flow straightener ' effect on these ' Low A.R. devices . ( Compare to Kfir C.2 )

Good Salsa Dancers will likely want Union Fees ! . 

[Ted Fancher]

Hi Ken,
"snip"
Current models of the top stunt/F2B pilots are more pitch sensitive and more difficult to fly level because all top people need the maximum sensitivity to make the pivoting, very tight corners.

I am not even mentioning here the flight dynamics differences between the models equipped with ICE and the models equipped with the electric motors.


"snip"

Top fifteen fly the models that are very difficult to fly because the pitch sensitivity reached the maximum possible for each individual flyer. These models were laboriously trimmed and sit now in a very narrow trim range (ref: Paul Walker's TrimFlow Manual).

Matt

"ship"

Oh Dear, Matt. 

Have to disagree fairly strongly about the underlined part of your response...although our disagreement might be merely a language issue.

Modern stunt ships flown by the top fliers I'm familiar with are not "sensitive" in the pitch mode to the pilots.  Sensitivity of response is the result of a combination of factors, most specifically, the trim of the airplane, the control system linkage and, most particularly, the associated "trim" of the handle by which the pilot directs the path of the airplane.

Appropriate "arm length" control system  combinations of bellcrank input (leadout spacing) and output (pushrod drive "arm") and appropriately matched control horns are connected so as to allow leadout displacements for an appropriate/preferred degree of "aircraft" pitch response and the adjustable control handle completes that control system by providing a means to adjust response of the aircraft to be consistent with the preferences of the pilot.  In other words, the on board systems are consciously configured to be insensitive, i.e. requiring significant bellcrank/leadout rotation to achieve more "modest" flap/elevator deflection per unit of pilot input.  Significant amounts of leadout displacement are thus necessary to get the reduced elevator, flap deflection necessary for the more aggressive corners demanded in modern competitions.

Note, however, the importance of having a handle capable of providing the necessary adjustable line spacing and line attachment overhang to close the circle of the properly trimmed and configured aircraft system.  When done correctly and appropriately for the aircraft involved the combination will provide a more responsive airplane utilizing modest and repeatable control inputs from its pilot.  The very responsive aircraft will respond appropriately with modest, mostly wrist and finger inputs allowing the pilot to simply follow the aircraft's path with his hand/handle making necessary inputs with the pilot's "preferred" amount of input.

As usual, I apologize for the length of this post.  I believe, however, that it is important that it be widely understood that a stunt ship capable of flying competitive patterns at today's expected level of corner size and crispness can be achieved with an aircraft system that is NOT excessively sensitive.

Ted Fancher

[Ken Culbertson]

  I believe, however, that it is important that it be widely understood that a stunt ship capable of flying competitive patterns at today's expected level of corner size and crispness can be achieved with an aircraft system that is NOT excessively sensitive.

Ted Fancher

Matt, I go back to the late 50's flying Stunt and flying it at an "expert" level by the late 60's.  I can assure you that the planes of that era were considerably more difficult to fly than the PA ships of today.  They were light and underpowered.  Overhead work in wind was a religious experience.

Corners have evolved over the years.  They were once very tight, somewhere around 10' (we all swore they were 5 knowing all along that they weren't) and often times even the best fliers had a tinge of a stall.  Locking was something that happened most of the time if you were good but there was always that random wiggle.  Anybody tried to trim out a "random wiggle"?  Then the corners loosened up to around 14' and now they are headed back to the 10' range.  No, today's ships don't do tighter corners, they do better corners.  Locking is a given and you can't blame the plane anymore for "Random Wiggles".

Ken

[Matt Piatkowski]

Hi Ted,
Thank you for your opinion.

Pitch sensitivity I was writing about is the subjective thing.
Each stunt/F2B flier combines several geometric parameters you have mentioned, to repeatedly get the response of the model that matches his (or her) flying style and receive the highest scores.

The style is another subjective thing: we all have one if we are aware of this or not.

If someone's flying skills, his (or her) reflexes, general fitness level, ability to coordinate lower and upper body movements and so on is such that this person can repeatedly fly the pattern very well, the pitch sensitivity is optimal for this person but ONLY for him (or her).   

By writing "maximum sensitivity" I meant the maximum sensitivity for a certain flier having a certain style, skills, and experience. I did not mean the absolute maximum sensitivity possible to achieve by changing several geometric parameters of the entire control system in a certain stunt/F2B plane. I apologize for not being specific enough and hope that my explanations are sufficient and free of "language issues".

Stay Safe and Happy Flying,
Regards,
M

[Air Ministry .]

The Big Ace , here in N S W , speaks of his Yatsenko Elipse , as being trimmed as needing concentration in level flight , as its where its most ' un groovey ' Trim wise .
Steadies in rounds etc . Definately the ' pilot traslated ' bumps & wobbles , in the straight sections . Well Practised ' in tune ' pilot .

This seems to be the general agreement in Aus. regarding the later Yatsenko airplanes , you need to have them set up ' toey ' to get good sharp squares . Overheard in their pits .

I relate it as a bit like ' Toe Out ' to get the turn started . Driveable but not for casual attention . Boardering on divergent stability in straight flight .
Zero Control Friction .
Aerodynamic  Loads ? Balalance with forces in turns to steady things a mite . so max responce at light control loads . WIND does seem to overpower things , in a good breeze , perhaps .

[Air Ministry .]

The Bloke here who is big into Ted's designs , relates them as being easy to fly ( by anyone ) with good all round performance ,
whereas he states ' other designs ' can be suited to a narrow range of pilots. not everyone can fly them well , get good performance from them .

I regard the ' total airfoil ' as being equally significant to say half the rest of the design factors . Moments Areas etc .

In the case of a few , such as the Typhoon (Hawker )  with ' No Nose ' ( 4 1/2 inch L E to Spinner ) on a 60 inch span , A similar airfoil to my Folkerts is used . Blunter entry ( Fwd. L E , tho . )

Assume the blunter ' entry ' gets lower ' nervousness ' .

The Mustang used a Typical Hunt / Wewage Airfoil , again with a blunter entry . This ship only came right with elevator freeplay . 5 m.m. at Elev. T E . , and came better then as the C G went aft .
( yet to see if the lighter ST 46 now in it is a step to far, in aftness of C G yet .

The Typhoon has no tendancy to self diverge - The response is smooth and steady . Turn tightening exponetially with handle . Tho at 7 Ton , the effective speed range ( lower ) is narrow  ,
or it becomes a dray , falling off all over the place . The near completed Fuse Plug , combined with a 1/16th D tube wing , should loose maybe 20 ounces tho .
The Tempest ( Hawker ) wing compatable to this Fuse Plug , is same airfoil as the P-51 Mustang . So we have hih hopes for its capeability .



3.2 : 1 @ 42 M/Sec. https://www.f1technical.net/forum/viewtopic.php?t=14575



The ' Less Disruptive ' TEARDROP aerodynamic form . !

If one looks at a airfoil as a deriveitive of a ROUND ( say Bar ) things get intresting . AT LEAST IT WOULD TURN FAIRLY WELL . !

The Blue Hara Hurricane Derived device ( above ) has aerodynamically extortionate ! L E , flowing into flat flanks .
Still gets toey , or pilot induced , and wind induced divergence. and buffet .
But C G forward - flys like a tractor or digger . Pul it turns . let of it straightens . Very forgiveing - bar buffet in say 12 knot plus wind . Earlyer still stayed out in 18 knot .

The thing is , to Visualise . And not be entirely blinded by science . Tho no doubt the maths can be usefull . If the results dont suit the theory , change the theory .
As some renowned scholar here has stated .

P.S. I think the most important ' design ' feature is FLYING . Lots. Hundreds of hours . On Anything thatll do the schedulle . The worse the better . youll learn a lot more' coaxing ' it through .
AND a bullet proof control system . It should work like a rifle bolt - . NO DEFLECTION whatsoever , if possable , under any envisaged flight loads .

If the things straight and timbers matched side to side , away you go .  The best experiance is experiance . rain hail or shine . Adverse winds are most informative as to design / configuration inadequacies .