In these happy modern ARF times it’s very easy to get a new airplane - you can just buy one. Many people don’t even know that a model aircraft can be built. As a result of this the knowledge about how to build one - let alone how to design one - begins to fade away or even get lost. For those daring souls who insist on designing and building their own creations the situation is getting tough. It’s difficult to find information. Questions in forums indicate that even at low level there’s a lack of basic knowledge. So I’ve tried to put together what I have learned and experienced during many years. Of course most of what is written below are not my own ideas. Like with the majority of us it was found out by trial and error, read in books and in magazines, copied from plans, stolen from other flyers, learned from correspondence, and mostly from talking to competent friends with a glass of good wine and a discussion about aerodynamics (maybe I’d better call it hydrodynamics). So - what follows is really simple, basic, low level aerodynamics. Experts are allowed to ignore this page.                                          
     
  There are not many books about control line aerodynamics, not to mention Aerobatics . The problem is that most are written in a technical language not possessed by the average reader, and some are loaded with complicated formulae. Formulae are okay but they are only half the truth. If, for example, a formula states “the stability factor is the product of moment arm multiplied by tailplane area”, nobody tells us that we cannot simply double the moment arm to achieve twice the measure of stability. If we did this, the result would be just a tail-heavy model , difficult if not impossible to fly. Instead we need to know which measure will create which behavior in order to get the best compromise (whatever we get it’s always a compromise!).
When George Aldrich designed his NOBLER he was centuries ahead of his time. Not by chance is the NOBLER design called the father of modern aerobatic tools. It’s not easy to improve on this design. Of course modern creations are the result of constant subtle improvements and some increased knowledge about aerodynamics. As one Swiss flyer has put it: a good flyer can only improve if he happens to get a better airplane. The reverse is true, too. Only when a flyer has reached a certain level of skill (and knowledge) is he able to design a better airplane. With this in mind the following thoughts were gathered and put in order. This story is no exciting reading for experienced experts. It’s just plain-simple-basic –easy to understand knowledge about kindergarden aerodynamics. Though most aspects of model design are connected to others and cannot be looked upon separately, I’ll try to explain them in logical order for clarity. There’s one thing an airplane cannot do without, so let’s begin with the

     
                                                                                               
  The wing is what makes an airplane fly, and one of the most important aspects to good performance is the wing section. Several design parameters should be considered when choosing the AIRFOIL.      
 

Section thickness is one factor that strongly affects lift and drag. Within reasonable limits lift increases as does airfoil thickness, but so does drag. Thinner airfoils produce less drag, but they also produce less lift. Also, the boundary layer (the airflow close to the wing section) separates at lower angles of attack with thinner airfoils. When performing the corner of a typical stunt manoeuvre, the “angle of attack” ( AoA from now on ) soon reaches a degree where the wing can stall if it is insufficiently thick. However beyond a reasonable thickness there’s a point of no return and things will get less positive. Very thick wings have lots of drag and call for a powerful engine. The NOBLER design used an airfoil derivation from NACA 0018 which has an 18% thickness - BUT with an added Flap part, which reduced thickness to about 13% (of course the flap width is included to the wing chord). A thickness of lower than about 12% seems not quite suitable for acceptable aerobatics, and about 20% seems to be the upper limit. As a very simplified rule and within reasonable limits let's say that lift inreases with thickness and forward location of the point of maximum thickness.
We also have to consider Camber. Oh yes, we use symmetrical airfoils for aerobatics - but we also use flaps. As soon as we deflect the flaps we change the symmetrical airfoil into a cambered one. This will drastically change airfoil characteristics and “Angle of Attack” . The sketch shows that even without changing body angle of the airplane, AofA can change considerably.

 
The Point of Maximum Thickness largely governs the shape of the airfoil. A forward location necessarily makes a blunt front end which induces a fast high pressure build up at the front part of the airfoil. In turn this causes an early change from laminar to turbulent airflow. Turbulent airflow can overcome speed and pressure increases without flow separation, so higher AofAs are possible. While turbulent airflow produces more drag, wing stall occurs at higher AofA (= later). For high point location a figure of about 25 to 30% chord is acceptable. This point also effects the rear part (shape) of the airfoil. More about this later.

This will point our airplane into the opposite direction of where we'd like to go. This is the reason why (all other things being equal) airplanes equipped with flaps at first seem to fly more softly. To allow tight corners we of course need flaps and thus have to compensate with CG location and suitable elevator deflection (see sketch at left).

In this context it is interesting to note that whenever good lift is desired, a nice smooth curvature of the UPPER airfoil contour line (!) with a camber of 9% chord has proved successful (don't confuse this with airfoil camber !), and this is true for all model categories, maybe even for mancarrying aircraft. No wonder this easily translates to our 18% thickness stunt airfoils!

As can be seen in the sketch at right, lift produces a force which will turn the wing (our airplane) nose down.With deflected flap ( now a cambered airfoil) this "negative moment" becomes even bigger, because a) cambered airfoils can produce higher lift and b) because the "Center of Lift" on cambered airfoils is located more rearward.
 
 
     
  Nose Radius is another topic discussed by the experts. Usually a sharp nose will make for a more sensitive airplane, while a rounded leading edge should make for a more tolerant flight behavior. But things are not that easy. Al Rabe once stated that he flew a model with a very blunt leading edge - and it was extremely sensitive. Many fliers have reported good results with airfoils with a nose radius only slightly larger than that of the NOBLER design. Nose radius of the NACA 0018 (18% thickness) is approximately 2.5% ( percentage of airfoil length). The effect of nose radius can be seen in the sketch. When a gust hits the wing, the AofA changes and with it the stagnation point. Now imagine the same change of AofA on a blunt nose (big nose radius) and on a small nose radius. On the blunt nose the same change of AofA will have a much smaller effect. On the small nose radius the stagnation point will move to completely different places and radically change the airflow; thus changing lift and drag drastically.
 
Drawing thanks to Mr. Hepperle, Germany. Center of Pressure (= Center of Lift) on symmetrical and on cambered airfoils (= deflected flaps). It can be seen that on cambered airfoils the Center of lift is located more rearward than on symmetrical airfoils.
   
     
All factors work together and influence each other. A change in one direction will alter everything else . If a design is to be changed, only one modification at a time should be tried, and the results of this alteration carefully assessed. Much knowledge has been gathered over the years by many pilots, but still not everything can be explained scientifically. However you cannot doubt the evidence reported by skilled flyers.    
      For example, Jim Mannall (England) noted that his airplane flew more smoothly if the airfoil section was flattened aft of the point of maximum thickness and in fact featured just a straight line. Years later Gilbert Beringer (France) has tried an airfoil which he had borrowed from the full size CAP aerobatic aircraft which is an even more drastic version of this layout. Here the high point occurs very early, at 20% of the chord. It’s very doubtful to transfer design features of full size aircraft to model sizes since Reynolds Numbers differ quite drastically. Consequently characteristics from the full size area cannot be transferred easily to model dimensions without careful considerations.
                       
Al Rabe (USA) experimented with a section that had a highly cambered airfoil contour aft of its maximum high point. He wanted to create a smooth transition of the basic airfoil to the deflected flaps. All of these gentlemen are top flyers, all of these systems apparently work very well - so who can argue their point !?                
I do know that even some experts do not question the contour of a shoe sole as a practical shape for an airfoil ( I strongly doubt that they really use such methods !! ). Basically I’m convinced that extremes are never a good solution. Starting with something like a NACA 0020 by all means change to whatever you have a liking for, but bear in mind the consequences described above.
        a small selection of airfoil templates    
Lofting means the change of airfoil over the wing span (= from root to tip). In applications outside of Control Line Aerobatics this can mean some kind of twisting the wing (change of AofA). In our case it can mean a change of airfoil only. By changing from one root airfoil to another tip airfoil we can try to influence the airflow along the wing span, thus flight characteristics. The root airfoil may benefit from another chord/ thickness than the tip airfoil. For example: because of tip vortices caused by the air curling round from the bottom to the top of the wing tip, the wing tips will stall sooner than the wing roots at high AofA. To avoid early flow separation , we can use a different section which will allow a higher AofA before this happens. This can
mean a thicker airfoil and/or a more forward high point. Conversely we can use a thinner section (with a more rearward high point) at the root, maybe gaining the benefit of less drag where the wing chord is the widest. Please note that the Detroiter wing construction method produces a wing with exactly the opposite airfoil distribution! Since Les McDonald has won three World Championships with this configuration, we must accept that his airplane flies well. We must decide for ourselves what we want to believe.
            Wing Planform
Apart from (hopefully) enhancing looks the wing planform is also responsible for the way our airplane performs. It has an influence on whether we have an efficient wing, enough lift, low drag, a rigid construction, and nice flying characteristics. Let’s compare different shapes.

                                             
      An attempt to try a constant chord wing; RYAN by Claus Maikis, ST 60. Enlarged version from Charly Parrot's original 35 size design from the Sixties.                  
Rectangle
The easiest way to build a wing is the rectangular shape. Since our airplanes fly in the tethered mode, this shouldn’t cause much of a problem. Actually many airplanes - from simple trainer models to decent stunt ships - use this layout. As long as centrifugal force will keep the lines tight, we’ll not run into problems. The problems will arise as soon as we want to fly slowly! And that’s what we want to do. Horizontal flight cannot give us headaches, but when we get to 45 or even 90 degrees, gravity will take its toll; let alone situations (gusts, lack of engine power) when the lines go slack. Then our airplane is on its own, and we will be very happy if it knows how to behave well.

   
On constant chord wings, pressure exchange at wing tips causes big vortices. Tapered wings have better lift distribution = close to elliptical, which means less vortices and less induced drag.Conditions can be enhanced by choice of airfoil.
 
Tapered wing
The most common wing shape - and for very good reasons. It has several advantages:
It is easy to build.
Its lift distribution is very close to that of an elliptical wing planform (which is considered the most
effective shape: high lift, low drag).
Good load distribution, like elliptical shape. Wing thickness is biggest at root where bending loads
are high. Bending load is low at tips ( small chord = low lift = low bending force)
Good weight distribution: most weight concentrated close to Center of Gravity, light wing tips
(we don’t need that like our RC brothers do, but it’s good to have it in case of a crash !)
Small tip vortices = low drag.
Taper ratio is the relationship of dimension of wing root chord to wing tip chord and is written in percentage. For example: the NOBLER has root chord of 33 cm and a tip chord of 24 cm. The ratio is 72%, and this is a very popular figure. One notable exception: Gilbert Beringer’s “Caudron” (one of his first designs) sports a 60% ratio.

Elliptical shape
Aerodynamic and structural wise the perfect solution, but a pain to build. The small gain in efficiency is not proportional to the time and effort needed to construct such a pleasing contour. A semi-elliptical shape comes very close in quality ( Smoothie, Thunderbird, Grondal etc.). Easy to do with a rounded tip plate; more challenging with one, two, or more reduced size ribs.

   
 

Above: Paul Tupker with his "Grondal" design. With a little "cheating" (mixture of tapered shape plus rounded tips ) a semi elliptical planform can be achieved, which comes pretty close to elliptical planform in aerodynamics and looks.

Swept wing
Maybe a temptation for (semi) scale like adventures. Aerodynamically positive sweep (wing swept backwards) is not a big problem except for a somewhat less than optimum airflow across the wing. The downside is that the positive swept wing has to be placed much more forward in order to allow for an acceptable Center of Gravity location. This might cause problems with the tank location/ room.
    However negative sweep should be avoided. The reason: if the airplane yaws, the forward (moving) wing panel of the positive swept wing has more “projected” frontal area, thus creating more drag; this will push this wing half backwards, thereby restoring the original flight direction and establishing and improving yaw stability. In case of the negative sweep, forces act in the opposite way. The “backward” wing half produces more drag because of more drag (increased frontal area), so the yaw angle is even increased, thereby de-stabilising the airplane in the lateral direction.
                                                                                 
For clarification: a swept backward leading edge doesn’t necessarily mean positive sweep and a swept forward trailing edge need not be negative sweep. However a swept forward flap hinge line may cause problems. Since we have two different “moving axis” for both flaps now, an easy solution is to use a separate horn for each flap. Each horn requires its own pushrod so the logical solution is a “fork shaped” pushrod. The problem is shown in the accompanying sketch. Because the pushrod is also moving sideways (it follows the bellcrank arch), both horns will be moved a different travel length, resulting in different deflection. The problem can be lessened by attaching the “fork” at a place where the sideways movement of the elevator pushrod is smaller; for instance aft of the flap horns.
Laser with ST 46 by C. Maikis. Problems with swept forward flap hinge line was discovered after the airplane was built only. Sketch at left shows problem.
 
    Wing tip shapes are probably chosen more for aesthetics than for rational reasons. For those looking for a scientific excuse: NASA has done research on wing shapes and has found out that those shapes with a backwards curved front edge were superior in performance over other shapes - and of course this includes the wing tip. It is interesting to see that those birds with the best “aerobatic” flight performance in speed and manoeuvrability - like swallows and gulls - are using this feature: the swept back leading edge. In the sketch several tip shapes are shown. Version A with the “hacked off” wing end seems to be the least effective since tip vortices are expected to be big. The strongly rounded trailing edge as in B is said to be no noticeable improvement. C represents the above mentioned solution, and D is the bird wing shape. Drawing E shows a typical tapered wing layout with a swept back wing tip (the well known “FliteStreak” shape) and the flaps NOT running till the very end of wing (more about this later).
                                                             
What about wing tip plates? The theory behind this gadget is to reduce the tip vortex by preventing the exchange of pressure difference between top and bottom surface of the wing. In reality this solution doesn’t seem to reach its intended goal. Only one Chinese flyer has ever appeared with this feature on a highly competitive level (World Championships). Instead of solving one problem the tip plates obviously create their own new ones by causing some airflow interference plus new little vortices, especially if not perfectly aligned! Add to this the disadvantage of having additional weight at the wing ends (a situation we’d like to avoid), not to mention the undesired hindrance when handling (carry around) the airplane. Of course all this can be said about the tailplane, too.
Aspect Ratio
This facet of design has never bothered stunt flyers too much. The basic figure of about 5 :1 has never been questioned. After all this is roughly the “NOBLER number” ( Aspect ratio = AR = is the ratio of “chord to span”, or “wing area to square of span”) . Actually it’s the same ratio, both formulae are useful for different applications. There have been many attempts to diverge from these proportions. The Adamisin brothers (USA) have realized their ideas as has Henning Forbech (Denmark). High AR has a place in gliders (full size and models) where they drastically reduce induced drag, which is necessary to improve glide performance (duration). The effect of high AR: the lift-coefficient doesn’t increase proportionally with increase of AofA. Which means: for a given certain increase of AofA the lift increases not linear, but over-proportional; in practical terms and in simple language: for a little more elevator deflection we get much more lift. Whether such an attribute will help your flying you’ll have to decide for yourself.

High aspect ratio design by Henning Forbach, Denmark.Henning also developed highly sophysticated control system to compensate for high AR shortcomings.
   
High AR airplanes tend to be more sensitive to turbulent weather: firstly because of the effect mentioned above, and secondly because of the longer wing: the distance ‘wing center to wing tip’ is longer , thus gusts work on a longer moment arm and can easily roll the airplane.
If the wing area is given and AR should be increased, the wing chord can be reduced and the span increased. At the same time this can improve roll stability: the line guide is located farther from the center of Gravity, and this should noticeably reduce tendencies for the airplane to shake and/or roll. This is especially interesting for biplane designs which usually have shorter wing spans (more details about biplane design can be found HERE ). Altogether high AR models are more difficult to trim, simply because they have a smaller chord which reduces the workable trim range. For windy weather low AR wings are recommended.

                                                                                                           
Wing Area
This is one factor by which wing loading can be controlled. The other factor of course is model weight. Many of us seem to have a problem with this, including me; the Gramms have a habit of concentrating on my airplanes. It’s generally accepted that an average wing loading of about 40 gm/sq. decimeter (that’s about 13 oz/sq.ft.) is an accepted figure. Performance may suffer a little with a high loading, but with modern high power engines this is not an unsolvable problem. Airplanes with very low wing loading can fly better on a calm day, they can be flown more slowly. But they are blown around badly when the wing starts to blow - and that’s what it usually does when we intend to go flying. Also a smaller wing is easier on the engine. Bearing all this in mind suggests using a wing area slightly smaller than average and accept the somewhat higher wing loading. For many years the Italians have followed this school of thought - and they have been quite successful with it.
Symmetry
respectively Asymmetry of wing halves is an age old controversy. The majority of airplanes is built with a longer inboard wing panel. But who can argue against a world champion (Bob Hunt) who prefers equal length panels. There are two reasons which recommend unequal panels. Firstly we have to compensate for the weight of the control lines which want to bank our airplane to the left (in anti clockwise rotation !). Secondly logic dictates that - because the outer panel is flying faster than the inboard panel - it will produce more lift. This again will raise the outboard panel which in excessive amounts can lead to hair raising moments. As a result ALL control line aircraft are equipped with (outboard) tip weight. Additionally most aerobatic models have a longer inboard panel. Those who stay with the equal panels always carry a hefty tip weight.
There are those who say that the inner part of the outboard panel (close to the fuselage) is shaded off by the fuselage, so the effective wing area is reduced anyway. I just cannot imagine that the fuselage is tilted outward far enough to really blank off part of the wing (make a reduced scale drawing and find out for yourself).
Another means to fight asymmetry in corners is to adjust flap size. More about this later.

Dihedral
Aerobatic airplanes don’t want dihedral. Half of their life they spend flying in normal attitude, the other half inverted, and they want to feel happy either way. Dihedral is absolutely necessary for free flight models because these must have built in roll stability. The principle: the down going wing half has a bigger projected area, thus has more lift, which raises this panel and restores the original flight attitude. Since our airplanes fly tethered on the lines, roll stability is not a problem and we don’t need dihedral.
Semi scale "Miles" design by Yves Fernandez, France, with dihedral.
However if somebody is tempted to allow for some semi scale appearance and to duplicate the shape of an original full size aircraft, the required dihedral will cause some problems. The center of gravity of our airplane and the line guide MUST lie exactly on the same horizontal plane. If they don’t the airplane will be tilted in or out; both situations will have disastrous consequences.
Vertical CG and line guide MUST match on the same level.
   
Sometimes we can see an accident waiting to happen when a low wing airplane is built with a straight wing. Because the model’s vertical CG is higher than the line guide it will fly in a banked-outward attitude. In normal upright flight this is not yet a big problem. However as soon as the airplane gets get inverted the catastrophe begins - the airplane will try to chase us! To solve this low winger problem we’ll have to bring the line guide (= wing tip) up to match with the horizontal CG location.
Flaps
Our control line stunt pattern includes some square manoeuvres with very sharp corners. Because of the resulting G-loads we need additional lift. This would be difficult to achieve without flaps. Contrary to popular believe flaps don’t work because the airflow pushes against the deflected flaps and thus turn over the model. Instead the deflected flap changes our symmetric airfoil into a CAMBERED one. Cambered airfoils can produce much more lift even at low angles of attack (AofA). As with every control surface, efficiency depends more on reasonable area rather than on great deflection. More than the right deflection doesn’t create more lift but much more drag instead. For the best layout we have to decide between span and chord. Since the function of the flap is a ‘change of airfoil’ , it appears logical to change as much of the wing as possible; that means: full span flaps. For the flap chord Ted Fancher recommends up to 20% of the wing chord at each station. Please remember that the greater the flap area the more load is placed on the control system. The “feel” of flying the airplane is changed and flight characteristics alter. NACA data say that about 20 degrees of deflection is optimum.

Deflection of large control surfaces can put a high load on the control system (horns and pushrods). So we shouldn’t forget what amount of force is available. The pull on our lines is all we can get. If the load on the control surfaces (flaps plus elevator) exceeds that of line tension, more deflection is not possible and deflections stops - with the usual bad consequences.
I’ve said that long span flaps are desirable. On the other hand we shouldn’t forget wing tip design. At a place with strong vortices already, a deflected flap would destroy airflow even more and cause additional drag. So it may be wise to not let the flap run till the wing end. A practical solution is to use the rearward swept wing tip shape and leave the tip without flap. This allows us to use long span flaps without unwanted great vortices. In practical terms we have built in some “washout” which is a reduction in camber, thus reduction of AofA , which means less lift - ergo less drag. Some designers use this part of the wing as an adjustable trim tab.

Most beautiful T-bird by Jeff Reeves, Australia.
 
Many years ago Bob Palmer (“Thunderbird”) tried asymmetric flap deflection; he used two flap horns (one for each flap) and a fork shaped pushrod. The idea is to have more flap deflection on the inboard wing. This sounds nice as long as the airplane is in horizontal flight. Upright or inverted doesn’t matter, the inboard wing has more lift and tilts the airplane outward, thereby increasing line tension. However the situation changes as soon as we fly a loop. Now the airplane is banked inward and tends to fly towards US !!! The idea has never caught on because the results were not satisfying.
Wing tip weight is necessary to compensate for the weight of the lines. Also it’s an easy method to improve line tension. In big stunt ships it can be up to about two ounces. Because of the high G forces in tight corners the airplane may want to roll: the outboard wing tip tends to ”swing” out of the loop circle. There’s an easy cure for this problem. We simply increase the span of the outboard flap and run it more close to the wing tip, or we increase the chord of the flap there.

Line Rake
Modern stunters are built with adjustable leadouts, but many models have a fixed line guide. So it is helpful to know in advance where the right position is. To put it simply: our airplane is hanging on the lines just like a pendulum. Now a pendulum hangs with its center of gravity exactly vertical below the point of attachment. In our case centrifugal force takes the place of gravity. Handle, line guide, and the model’s CG position are roughly in one straight line. However in flight the drag of the lines will bend them backwards, depending on model speed and weight. They will enter the wing at an angle slightly behind that theoretical straight line. Position of the line guide will have to compensate for this deviation by locating it slightly backwards. About three degrees is a generally accepted figure. Since we can trim flight performance of our airplane by shifting line guide position according to line length, speed, and weather conditions, it’s highly recommended to use this feature.
 

A tailplane has two duties, it has to stabilize and to control our airplane. It has to compensate for the negative moment of the airfoil (= wing), and it has to change the AofA to change flight direction - in order to fly loops and corners. This task is done by creating lift in one or the other direction = up or down. Again we have to change a symmetrical section into a cambered one. The most practical method is to use the “flat plate” section which is very easy to build, and so far has proven as a satisfying solution. From casual judgement the airfoil shape looks rather crude when the elevator is deflected. As I said at the beginning in control line we don’t have much scientifically tested knowledge, so nobody knows exactly what aerodynamically happens to a flat plate section when it’s “broken in half”, and how the airflow behaves at that sharp corner with the hinge gap. Les McDonald has tried a tailplane section which looked more like an airfoil with about 15% thickness. He reported that this airplane flew somewhat softer in corners. As far as I know he returned to the flat plate section later.
A typical tailplane in the 35 engine size airplane is built up from ¼ inch balsa ( for the more simple designs). For more demanding tasks spars and ribs of 3/8 thickness are used; that makes up for a thickness of 8% which is generally accepted as an appropriate figure for flat plate sections. For bigger airplanes in modern fashion size the wood dimension will have to be increased accordingly.
There have always been flyers who have tried to build tailplanes with a smoother contour at the hinge point. The obvious solution is to use a hollow chamfer type construction. This needs some additional work; you have to carefully work out the shape, make sure that the hinge point is located ‘within’ the leading edge of the elevator, and insist on a very narrow gap. Otherwise all this elaborate work is in vain.
      In full size as well as in model applications it has been tried to build V-tails. The reason was to save drag by eliminating the fin. It has been found that (in full size just as in RC application) it’s not the perfect solution it was thought to be. In our circles it’s just a matter of aesthetics whether somebody would like to choose this feature. Theory says that the V-tail must have the same horizontal AND vertical projected area as the conventional tailplane. Usually this results in a very large tailplane which is a) heavier, b) more difficult to build, c) will need a more complicated horn and pushrod system.
I have chosen such appearance myself many years ago. My PALATIN design was strongly inspired by Bob Hunt’s Genesis design. The V-angle was very shallow. I think this solution was possible only because this model had enough side area aft of the CG.

PALATIN by Claus Maikis, ST 46. Shallow angle V-tail.
Efficiency of tailplane would again call for a high aspect ratio. But again we have to consider pilot’s reaction, skills, and desire. Ted Fancher recommends a rather low AR. The drag of low AR plus that of the deflected elevator creates a stabilizing effect which can produce a dampening effect. This might help many a pilot; for instance when coming out of the corner - it may eliminate overshooting or at least reduce those damned wobbles.
There’s also the method of keeping the elevator thinner than the stabilizer. The idea is to let the elevator move in the turbulent airflow behind the stab so that at very small deflections there will be no reaction. The airplane should fly very stable when no control input is given, which should help for horizontal flight and for the straight legs in the manoeuvres. Only when bigger deflections are given should the elevator begin to work. Another idea is to slightly ream out the hole in the elevator horn. The resulting small play in the control system should cause similar results. Both systems seem to work.

Personally I prefer to solve problems by using aerodynamic measures (for instance trim). This helps me to understand what I’ve done and what has happened.
Over the years tailplanes have developed to a size of about 20% of the wing area or even more. The most common layout is to divide the area into roughly 50% stab and 50% elevator.

 
                                                                         
 
Basically a fuselage is only a means of holding engine, wing, and tailplane together. Fuselage dimensions, and thus moment arms, have been developed empirically to the best compromise, to give comfort in handling, sufficient stability, and agility in corners. The nose moment is generally given as the distance from spinner backplate to wing leading edge at the root. Tail moment arm from flap hinge point to elevator hinge point. This has become common habit and it’s easy to work with. Nevertheless it’s wrong. Aerodynamically it doesn’t mean anything. Also it doesn’t take into account the geometry of wing and tailplane.
    For example imagine a swept wing design where the wing is positioned much more forward than on a traditional layout. Where will you measure the point ‘flap hinge line’ ?! These definitions just won’t work here. A correct system can only use centers of lift of wing and tailplane as reference points. For practical purposes we might use the distance ‘engine CG to airplane CG’. Not very scientific, but easy to calculate when it comes to transferring data from one design to another; for instance when our next design must have a much heavier engine. We just multiply engine weight times (let’s call it) nose length; then divide the product by the new engine weight and we roughly have the new nose length to use for our drawing.
 
       
Moment Arms
The airplane is “balanced” when the CG is placed at exactly the correct point; for our symmetrical airfoil that is at about 25% MAC ( = Mean Aerodynamic Chord). On rectangular wings that’s easy to locate. On tapered wings we have to find out. There’s a graphic method which can be used easily on a reduced scale drawing. The drawing will explain how it’s done.

Side view of G. Beringer's (France) first design. Interesting dimensions of moment arms.
     
 

How to determine CG location on tapered and/or swept wing (sketch at top):

Draw an accurate planform (reduced scale is sufficient). Draw Root Chord and Tip Chord as shown. Connect ends as shown, draw 50% line on wing. At crossing of both lines, this is location of Mean Aerodynamic Chord (= MAC). At 25% of MAC draw line (to wing center). This is CG location of wing (and possibly of whole airplane).

For swept constant chord wings see sketch at left. It is this point which is used as reference for calculating Moment Arms.

 

A “moment” is the product of “force (or weight) multiplied by distance (=arm). For instance: an airplane can have the same moment with a light engine on a long nose - or a heavy engine on a short nose. As long as the moment (working around CG )is kept the same, the airplane is balanced just the same. As long as CG is placed within a reasonable range around this 25% figure, the airplane will fly well in perfect horizontal flight (for one moment let’s forget all other aspects). Static balance is okay.
But now we want to do some stunts. Changing flight direction means we have to apply forces and we are entering the area of dynamic flight. Now some other laws apply. The inertia of the nose mass (mainly engine weight) is responsible that our airplane doesn’t want to rotate as quickly as we want it to do. To change direction we have to apply a big force. Then to return to straight flight the same force is needed again. It’s very difficult to do this quickly and precisely. The reason is that “inertia” is the product of ‘weight times moment arm SQUARED’ ! This means: with twice the moment arm and half the weight our airplane would balance at the same CG location - but inertia would be much higher and turning ability would be much worse (because the moment arm dimension is squared).
I remember when I changed from ST 46 models to ST 60 designs. Because of the heavier engine I had roughly calculated the nose length for the ST 60. Instead of the long nose for the relatively light 46 I now had a very short nose for the ST 60. Immediately my corners were much better. I really attribute this to the short nose moment arm. As a consequence it appears logical to keep the nose very short (as far as engine and tank lengh allow.
We would like to apply the same principle to the rear fuselage. However stability in level flight, manoeuvrability , and CG location have to be considered. Modern aerobatic designs sport rather long rear fuselages. With the latest trend for rearward located CG the task is a little easier to solve. Anyway it pays to build the tailplane very light.

Side area
One basic method to maintain line tension and to control line tension loads is to vary fuselage side area. The trick is to shift side area to the exact place where we need it. A lot of side area aft of the CG has some nice benefits: at Upwind the wind pushes the rear part of the fuselage into the circle, thus pointing the nose outward and thus keeping line tension. Downwind the rear fuselage is pushed outward thereby pointing the nose inward and help to keep line tension “bearable”. Of course this effect depends on wind speed, is not controllable, and can have negative effects if the designers have gone too far. Modelers have tried to find a neutral solution and have used big canopies as a large side area close to the CG (later known as the “Russian style” layout). According to special requirements and preferences suitable fuselage shapes can be found.

     
Two versions of placing the center of side area close to but slightly aft of the CG. COMMODORE (front, 35 engine) and CORONADO (46 size) are my own designs.
   
                                            Sundries
Much thought has been spent on component arrangement. Because our airplanes should perform equally well in upright and inverted flight, a fully symmetrical layout should be the best solution. This would require engine thrust line, wing center line, and tailplane to lie on one common straight line. However there are a few factors which may stifle designing a totally symmetrical airplane. Namely this is engine weight, and landing gear weight and drag. A wing center line slightly below the thrust line (usually about 1 inch) is a popular solution, as is the position of the tailplane: on or slightly above thrust line. At least this partly takes the tailplane out of the wing vortices. Another benefit is easier tailplane pushrod installation. There’s another thought: even if it was theoretically possible to build a true symmetrical craft without any disturbing moments - would this be a good design? Just maybe the perfect solution is the one where different moments balance each other? And where - if one moment increases - a second moment compensates for this, accordingly and automatically? This has to do with “dampening”, a tool which is widely used in controlling systems.

     
Luciano Compostella, Italy, with his famous "Tango" design.Upright engine installation (ST 46), almost "inline" configuration, top European flyer.  
 
Designers can have their fling when it comes to fuselage shape. We can assume that the airflow around the fuselage is not particularly clean. So why not try some fancy shape? It has been argued that thick fuselages with large cross section (like circular cowls) create too much drag. I don’t think that this is a big disadvantage and the drag can be neglected. Al Rabe has figured out that the very roomy fuselage of his Sea Fury increased overall drag by just 3%. So nobody should feel handicapped by trying an unusual fuselage shape, as long as the propeller’s efficiency is not impeded.  
                                                                 
Twins are not a problem aerodynamically. It has been tried to use engines with different rotation directions in order to reduce torque forces, however these have never caused any trouble. It goes without saying that engines should be mounted as close to the airplane center as possible for rigidity and weight reasons of the wing. Then it appears logical to use two tip fins on the tailplane in order to bring the fins right into the prop blast of the two engines to make the fins more efficient..
Undercarriage is not an aerodynamic topic. However it is responsible for the way the airplane behaves during launching and landing. Basically there are two types of gear: the so called “tail dragger” and the tricycle gear. With the exception of the “Shark” and a very few other designs the latter has not really caught on in control line circles. It seems obvious that with a tricycle gear you can “bang” down the airplane hard and fast on a hard surface without caring much about precision, but reality is different. It’s the location of the main gear which makes the difference. It’s evident that the main wheels have to be located aft of the CG. If they are too far aft, during launch on grass fields our airplane will have some problems to rotate, build up lift, and take off. On landings our airplanes usually come in in a nose high attitude and will touch down with the main gear first. If gear position is too far aft the sudden push on the wheels will rotate our airplane nose down quickly and the airplane may even bounce.
       
Two-wheelers work differently. The preferred method is to touch down not too slowly with the airplane still in a horizontal attitude. This seems to be quite difficult. However if the wheels are in the correct position it’s really not that hard. Because our airplane is still faster than in a three point landing it’s easier to control and less sensitive to gusts. After the speed has decreased the model’s tail will descend slowly, giving a very realistic appearance and a wonderful sight! A perfect landing is easier to do this way. Perfect three point landings are very difficult to achieve.
Modern stunt designs have a long tail wheel strut. This makes the airplane sit “tail high” on the ground with the thrust line almost (!) horizontal. On launching the airplane cannot rotate early and quickly. It needs to build up some speed in order to produce enough lift, so the whole process happens smoothly. On landing the same principle works again. If the airplane is put down too hard, the tail cannot drop down suddenly because the long strut will prevent this.
 
A problem may arise with some oldtimer designs or semiscale models (like, say, a Messerschmitt 109). Many of these airplanes have a long landing gear. On the ground they are standing in noticeably nose high position. Right from the start they tend to lift off very early. When a low pitch prop is used, thrust is very high at the very first moments. The airplane will accelerate and rotate very quickly, especially on grass fields (because the wheels are held back by drag), and the propeller may hit the ground.
There’s one more effect which sometimes deserves attention. It’s called “P-factor” ( I will not delve deeper into this topic HERE, but in another place). In short: at very high AofA (when thrust line is at a big angle to flight direction) the prop disc creates forces which may turn the airplane away from the flight direction. In case of our long legged airplane standing nose high at the very moment of release , the big AofA causes a force which turns the airplane to the left, means inside! Especially on bad grass fields where drag is higher and acceleration lower the take off roll takes longer ; ergo P-factor effect can be noticeably stronger (it has more time to act). In fact I’ve got the impression that when flying my Me 109 on a grass field, during launch I have to be careful to not let the airplane come in on me. And I’ve made it a habit to make a few steps backwards during launch roll, at least a little more than when flying on concrete.
                                                                                     
Finding the correct wheel location is easy. On our drawing we run a vertical line through the CG. We draw a second line through CG but at 20 degrees forward . Depending on the length of the landing gear we find the correct point for mounting the wheels. For uneven surface we can use a slightly higher figure; for beginner models even more in order to prevent nose over and prop damage. Remember: a location too far forward and the airplane will bounce on concrete, too far aft and it will nose over on rough surface.
For me the ideal configuration is an airplane with a slightly aft wheel location (maybe 15 degrees) and long tailwheel leg, sitting very low and flat (very small AofA) on smooth asphalt ground. Only a tiny little bit of up elevator is needed - then the airplane is released and left on its own. You just CANNOT make better launches!
  A "minimum" aerobatic design: box fuselage, rectangular wing, sheet tailplane, upright engine installation, yet good performance    
Of course there are many more and other features which must be considered when designing an aerobatic airplane. I just wanted to mention some basic knowledge and rules in order to provide a level which allows to take a first design step. You may think that I didn’t give much concrete advice. In this case - you’re right! I cannot. Nobody can. Remember - the Nobler, the Stiletto, the Genesis have all been World Champion airplanes. - yet how different they are.      
                 
It’s not easy - and very often not clever - to change or improve on these designs. We may alter their flight characteristics to meet our desires, preferences, skills, or the lack of them. Craft which are destined to provide top performance and have been developed over a long time period, are refined to a very high level. There’s not much freedom left to improve their performance. Errors and extremes have been ironed out. It is my conviction that extremes are an indication of bad design. Just as in trimming, any measure should be used as sparingly as possible. One extreme may gain a small advantage in one direction, but it will certainly create a disadvantage elsewhere - maybe in many other areas. Also there is one important point: our activities, even in competition, cannot be measured in seconds or in meters/inches. The same goes for the airplanes. What we are looking for is flight characteristics, and there are different needs for different pilots. What may be heaven for one pilot might be hell for another one. This is what makes our event so colorful. I hope that with this story I can help to keep it that way.