I originally wrote this article in 2003, back when we used to ‘build’ our models!
Some of the adjustments described in the first few paragraphs are therefore difficult to do with the more common ARTF’s, but I have left them in as the comments are still applicable and definitely relevant if you do actually build one yourself.
Trimming an aerobatic model can be quite a long process, but is time very well spent, the payback is in the pleasure of flying a well set up model which will do what you want when you want it to and will improve your flying standard into the bargain.
TRIMMING FOR AEROBATICS
Building and initial setting up
The process starts on the building board. Time must be spent in checking to ensure that the aeroplane is symmetrical in all respects. It seems obvious but how many modellers actually check for instance that both wings are exactly equal in span when measured from the fuselage sides?
Checks should be made of the following,
Check the triangulation i.e. check that the dimensions from the fin post to each wing tip are the same and that the wingspan is equal on both sides. Adjust the wing mounting position until correct. When this is done, check from wing tip to tail tip and tail tips to nose on both sides.
If the fuselage has not been built straight then it will be impossible to equalise all dimensions and consequently it will also be impossible to achieve a straight flying model. Always use a fuselage jig or at least build over a straight line drawn on the building board.
Check that the wing and horizontal tail are exactly parallel with each other by sighting from the rear of the aeroplane. Now check that the fin is exactly perpendicular to the horizontal tail.
When joining wing panels do not line up the wing sections at the root. It is more important that the wing tips are at the same angle of incidence. In other words the centre lines of the wing tip chords are parallel. This can be checked with an incidence meter or alternatively, attach two parallel sided sheets of balsa to each wing tip along the centre chord line and sight from tip to tip to line up the top edges of each sheet. The same process is then followed when joining the tailplane halves.
Time spent checking and then checking again before getting the glue out will pay dividends when flying the model and will keep the need for trimming to a minimum.
Before flying your model the centre of gravity and lateral balance (i.e. from wing tip to wing tip) should be checked and weight added if necessary.
Check control surface movements. Elevator halves should move exactly the same amount either side, likewise for ailerons. Aileron differential will always be required to make the model roll along its axis. I always start by adjusting the throws to give equal movement up and down on both sides and then dialling in 20% differential (more up than down) at the transmitter. This will usually be somewhere near right in my experience but will probably need fine adjustment after flying. Of course a computer radio and separate aileron servos will be required for this (and are highly recommended for aerobatic flying).
Finally, on the subject of building, lightness is the key. Not only for low wing loading and power to weight ratio, but also for crispness of control response. Light wings start and stop in rolls, snaps and spins much more quickly than heavy wings.
Flying
You have now flown your model and hopefully it flies well. It is now time to start trimming it to make it
even better. There is more to trimming a model than adjusting the transmitter trims until the model flies straight and level hands off. Unfortunately this only works when the model is straight and level, not very often in aerobatic flying.
What we are looking for is an aeroplane that tracks in a straight line, whatever direction it is pointing. It will loop inside and outside with no tendency it corkscrew out. It will enter and exit snap rolls and spins easily and has controls that do not interact with each other (rudder only yaws, elevator only pitches and ailerons only roll).
To achieve this you must start with a sound design and build it straight (see above). Trimming will improve any model but will not make a bad model good.
Firstly set the control responses to your taste, my experience of flying other people’s models is that the
controls tend to be unbalanced, for example sensitive on aileron and soft on elevator. This is a personal taste thing but I would recommend not having the controls too sensitive. I always set up so that at maximum stick deflection I have the maximum response I need, then use exponential to reduce the sensitivity around centre so that it is reasonably soft. Trying to make small adjustments while flying manoeuvres is impossible with over sensitive controls.
When you are happy with the feel of the controls, the first thing to check is whether the centre of gravity needs adjusting. Take the model to a safe height and check the stall/ spin characteristics. The elevator must be powerful enough to raise the nose at slow speeds to induce the stall and you are after a clean but not vicious stall. Once established in a spin, on releasing the sticks the model should recover within a half a turn. If the elevator seems dead and the model is reluctant to stall the centre of gravity is too far forward (nose heavy). However if the elevator is over sensitive and the model reluctant to exit the spin easily then it is tail heavy and weight is required in the nose. When the C of G is correct the model will snap roll and spin nicely and will not require very much down elevator when inverted. Ideally when the cg is near right, when flying a 45 degree upline, roll inverted, the model should continue on the 45 degree line with only a slight tendency for the nose to drop without the use of down elevator.
The above is perhaps over simplistic and in reality it involves a lot of trial end error and a fair bit of experience to know when you have got it right. However is a reasonable guide to the way forward.
The next stage is to set the engine thrust line. If the model climbs under power and dives when power is reduced then down thrust is required and vice versa. To set side thrust, establish the model in a true vertical climb and then centralise the sticks. If the model pulls to the left then add right thrust. If it pulls to the right, left thrust is required. Usually a few washers behind the engine mount will do the trick. You may find that a few attempts will be required until the right balance is found. It is possible to cheat if a small adjustment is required by enabling a transmitter mix, with throttle as master and rudder as slave.
Trim the rudder so that the model flies a straight vertical line, land and set the mixer so that as the throttle stick is moved from full to idle, the rudder trim reduces to zero. This will work well if only a small amount of trim is required, but will never be as good as actually adjusting the side thrust angle.
It is important that C of G and engine thrust lines are adjusted before anything else as they can effect other things. You could have the model looping as straight as a die, but change the side thrust and you will be back to square one.
The next step is to fly a few consecutive inside loops. You will probably find that you have to make aileron corrections one way to keep the model tracking straight. Make a mental note which way you are correcting.
Now fly a series of outside loops and do the same. If you find that you are correcting with aileron the same way then I'm afraid that you have a warped wing. If you are correcting the opposite way then one wing panel is heavier than the other. The latter is easy to correct, the warped wing is not so good. If the warp is bad the only solution is to saw the wing in half at the root and reset it or possibly even throw it away and start again. However if the correction required is small, set up a mixer with your elevator as the master and aileron as the slave. You can then mix small amounts of aileron with elevator input in the required direction to correct the fault. Although this will help with a bad warp, it is only really a solution with minor warps.
Major warps require major surgery.
If a heavy wing is your problem trim the ailerons with the model upright then roll to inverted, weight will need to be added to the wing that rises. The reason being that when upright, aileron trim is used to hold the heavy wing up. When inverted the trim acts in the same direction as the weight and the model rolls towards the heavy wing. If in doubt add weight to one wing, fly again and see if it is better or worse.
Of course reality strikes as what you will probably experience is a combination of weight and warp. Trial and error time again. Try one change at a time as described above and eventually you will find the right combination.
If you find that in a vertical climb the model pitches towards its canopy (assuming your cg is correct) then you have too much wing incidence, towards its undercarriage then not enough incidence. Aerobatic models typically require between a quarter and a half a degree of positive wing incidence.
We now move on to look at control interaction. Starting with ailerons. When rolling, the model should roll along its axis, with no yawing motion. This is easiest to observe in a vertical climb. If it doesn't roll straight then you must adjust the differential setting mentioned earlier. You will need to determine whether to increase or decrease the amount of differential movement.
The way I do this is to establish the model in a true vertical climb, of course by now your model will hold that vertical line with no correction! Now perform a half roll. You will probably find that the model has yawed off line. Use the rudder to correct the yaw back to the vertical. Which way did you move the rudder? If you moved it in the same direction as you rolled, then more differential is required i.e. more travel on the up going aileron. If you moved it the opposite way then less is required. So if you roll to the right and correct with right rudder then increase differential, roll to the right and correct with left rudder, less differential. You will probably find that a few adjustments are necessary before you find the correct setting. Again please note that a computer radio makes this task a simple one. If you don't have that luxury, to do this by altering your control linkage geometry will be time consuming.
There should be no interaction from the elevator as we have already dealt with correcting deviations during loops. However if you find that you still have minor problems/ they can usually be removed using an elevator to aileron mixer as described earlier.
Finally we must deal with the rudder. Unless you have a real thoroughbred airframe you will find that when you use the rudder, in addition to yaw you will also have pitch and roll effects. Rolling with rudder is a result of the wing dihedral angle. If the model rolls in the same direction that rudder is applied then there is excessive dihedral. If it rolls in the opposite direction then you do not have enough. Without the benefit of electronic mixers, the only way to correct this problem is to cut the wing at the centre and reset the dihedral angle. It is likely that this would have to be done more than once to find the correct angle, not good. With a computer radio, all that is required is to set another free mixer, this time with rudder as the master channel and aileron as the slave. Adjust the amount and direction of the mix until there is no roll present when the rudder is used.
Exactly the same process is followed to correct unwanted pitch with rudder, although this can be more of a compromise, as the amount of pitch present will vary with the attitude of the model. With a well designed airframe there should be very little pitch with rudder and it will be easily corrected with a mixer. As mentioned before a bad design can be improved with trimming and electronic mixing but will never be good.
One final mix I like to use (assuming you haven’t run out of available mixers) is an offset throttle to elevator mix. Even the best aerobatic models, even after all trimming has been done, have a tendency to gradually pull out of a vertical dive. To correct this, if your radio has the facility, set up a mix with throttle as master and elevator as slave and offset the mix point to around 2 clicks above idle, above this point there is no mix and below it a small percentage of down elevator is programmed (around 2%). This will have the effect of holding the vertical line when the throttle is closed. The mix percentage is so small that you should not notice the mix at any other time.
This may all sound like a lot of hard work, but take the time to do it and your effort will be paid back many times over. You will have a model which is a real pleasure to fly and is predicable through all manoeuvres.
Although the down side is that now when a duff manoeuvre is flown you won't be able to blame the model.
This was published in the club newsletter some time ago, I also did a couple of club meeting talks on the same subject if anyone remembers.
Hope someone finds it useful.
Jon Tappin was display and synchronised aerobatics pilot for JR-MacGregor Team and was also part of the team GB F3A in 1999 in Florida.
