Trimming

What's Trimming?

You've finished building your new model, test flown it and it goes great. Got the control movements sorted out and it loops and rolls. Right, on to the next model! Yeah, right!!!

A test flight should be just that - a check that the model goes up and comes down and that it behaves predictably. The first few flights should be used to get your engine and propellor combination sorted out. If you've already flown with the engine in another model, this will be a fairly short process - you know the propellor size and type that it flies best on although this may vary slightly with your new model. If the engine is new, make sure that the engine is well run in and that you're happy with the propellor before you move on to trimming.

Trimming the model has nothing whatsoever to do with the trimming levers or, on more expensive radios, the digital trims! A well designed and built model shouldn't need these moved anyway (and I believe in Santa Clause, too). Trimming the model is a process which is designed to improve the flying characteristics and ease of contol when manoeuvring your model. Although all the diagrams in these pages are very idealistic, it may not always be possible to get an "absolute" solution to any particular characteristic. In these cases, you have to make a judgement call and balance the behaviour in one mode against that in another. All models will benefit from trimming - even a trainer - although pattern and aerobatic models will benefit the most. There are four fundamental stages in properly trimming a model - the engine, the airframe, the controls and mixing. It is essential that the trimming process is carried out in this order because the number of iterations to get a correctly trimmed model will be greatly reduced.

The level of trimming that you can achieve is also limited by your transmitter. This is where computerised radios really come into their own with channel mixing, travel adjust, offsets and so on. Ever wonder what they did? Me, too! The problem with the (thick) manuals that come with computerised gear is that they only explain how to programme the radio. They are extemely poor at telling you why you would want to! Hopefully, these pages will help you to understand why you would want to read the instruction manual!

Where flying tests are required, choose a day with little or no wind and turbulence.

When we come to describing programming computerised radios, I will use the terminology adopted by JR - because that's what I fly with - but if you're a Futaba or Sanwa owner, I'm sure that you're well capable of identifying the appropriate functions.

If you think I'm real smart for knowing all this stuff, I'm not. The majority of this series of web pages is based on the expert knowledge and skills shared with me by Ian Beveridge and Doug Thornton over several beers! Thanks for sharing your knowledge.

Return to top of page

Setting The Thrust Line

Most models require the engine to be set at an angle relative to the fuselage centreline and is typically 2½° right thrust and 1° down thrust. The engine mounting will also be offset to the right when viewed from the front. For a clearer understanding of thrust lines, see our online Thrust Angles section. Setting the thrust line correctly is probably the single most effective thing you can do to improve the handling of your model. It is also easy to test fly for and easy to correct - a real "no-brainer".

Simply put, right thrust is required to compensate for the spiralling airflow from the propellor hitting the tailfin while down thrust stops the model climbing when the throttle is increased. (Unlike a trainer, where more throttle equals more power equals more height, the aim now is for more throttle to equal more power. With today's high drag or "constant speed" models, a considerable increase in power only gives a modest increase in airspeed.)

To get the thrust line set properly, you need to eliminate pitching and yawing with throttle use.

It is essential that you have decided on the engine/propellor combination that you will use and that the engine is "on song" and well run in. It is also essential that the model is trimmed to neutral. If this isn't the case, you'll be wasting your time.

Return to top of page

Establishing The Side Thrust.

Fly your model straight into wind at full throttle and pull up into a vertical climb. The theory here is that as the airspeed bleeds off, a continued vertical climb is now fully dependent on engine power alone. Do not use rudder (or rudder trim) to correct for any deviation from the vertical. Your model will do one of three things.

1. If the model continues to climb vertically, the side thrust is correct.
2. If the model gradually, then more noticeably, veers to the right, the model has too much right thrust.
3. If the model gradually, then more noticeably, veers to the left, the model has too little right thrust.

This test should then be repeated but this time start by heading downwind. If you're happy that you know what you have to do for side thrust, you can go straight on to check the down thrust in the same flight. Repeat this process a few times to ensure consistent results.

Return to top of page

Establishing The Down Thrust.

Fly your model straight into wind at full throttle and horizontally. Abruptly close the throttle and keep it closed for 2 or 3 seconds. Open the throttle fully again. Do not use the elevator (or elevator trim) to correct for any variation in height - unless of course it's screaming straight for the ground! The model will do one of three things.

1. When the throttle is closed, the model slowly arcs nose down (as the airspeed drops) and when the throttle is re-opened, the model should continue towards the ground in a straight line. Recover (Duh!) and congratulate yourself that the down thrust is correct.
2. If when the throttle is closed, the model arcs nose up and when the throttle is re-opened, the model goes nose down, you have too much down thrust.
3. If when the throttle is closed, the model arcs nose down and when the throttle is re-opened, the model goes nose up, you have too little down thrust.

Repeat this test a few times to make sure you're getting consistent results.

Return to top of page

Changing The Thrustline.

Whoopee! My favourite - an engine out job... This can usually be done at the flying field.
Do not attempt to change the thrustline by sticking things under the engine mounting lugs. The correct way is to alter the angle between the rear face of the engine mount and the front face of the firewall. There are several ways of doing this.
Perhaps the easiest way is to place a washer or washers on the bolts that hold the engine mount to the firewall between the mount and the firewall. Most mounts have four bolts holding them to the firewall so a few minutes thought and you should be able to work out what goes where.

1. To increase down thrust, add washers to the top of the engine mount.
2. To decrease down thrust, add washers to the bottom of the engine mount.
3. To increase right thrust, add washers to the left of the engine mount (when viewed from above).
4. To decrease right thrust, add washers to the right of the engine mount (when viewed from above).

Down thrust and side thrust can be changed together.

Check that there is still adequate clearance around the engine and exhaust then ensure everything has been re-tightened correctly and try re-testing. There should be a marked improvement. If it's not perfect, it's engine out time again! Keep repeating the whole process until the model successfully and consistently "passes" the thrust line tests.

Once you are happy with the thrust line settings, you can make the installation permanent by machining the rear of the engine mount to the appropriate angle. If you have had to add a considerable amount of side or down thrust, you will probably gave to move the position of the engine mount to bring the propellor/spinner back onto the fuselage centreline

Remember, if you change the propellor to a diffent type or size or change the engine, you will almost certainly have to alter the thrust line again.

Return to top of page

Checking The Wings.

Once you have set the thrust line up, the next stage is to check the model for balance and wing warp. Hopefully, you'll have built your model straight and true. These tests should prove it. Any deviation from straight loops and rolls is generally not excessive although we have witnessed an ARTF "pattern ship" which was so bad that it was in knife edge flight after the first quarter of a loop! (I guess you shouldn't try to make a decent model that includes the words "cheap", "ARTF", "pattern" and "ship".)

This is a two stage test and you have to be able to execute several continuous loops followed by several continuous bunts. The easiest (and most effective) way of doing the bunts is to "cheat" by rolling inverted and bunt upwards into wind. Since you are trying to establish whether the model rolls (or "screws") out of a loop and what's causing the rolling, you must not use aileron or rudder to correct tracking through the manouevres. It should also be fairly obvious that incorrect side thrust will also cause the model to screw out.

If you're new to trimming a model properly, it's a good idea to get a friend to help you with this by taking a note of what the model's doing as you carry out the manouevres.

The first step is to execute several loops so head straight into wind and start the loops. Take great care not to input rudder or aileron to correct any deviation from the loop. Ensure that you have enough airspeed going over the top of the loops - dropping a wing because it has stalled will only confuse the issue. As usual, one of three things can happen.

1. The model will fly superimposed loops.
2. The model will screw out (in a spiral) to the right.
3. The model will screw out (in a spiral) to the left.

Note what happened in this first part of the test.

Now for the part requiring "bottle"!

Execute several bunts. Unusually do this into wind as it makes it much easier to work out what's happening. As with the loops, ensure that rudder and aileron are not used while maintaining sufficient airspeed over the top of the bunts. As with the loops, the model will do one of three things can happen.

1. The model will fly superimposed bunts.
2. The model will screw out (in a spiral) to the right.
3. The model will screw out (in a spiral) to the left.

Repeat this sequence several times to make sure you are getting consistent results. That's the test flying done - now for the thinking part! You need to combine the results of the two tests to find out what the solution is. The best way to analyse the results is to use a series of tables.

If you have to add weight to either wingtip, do so in small increments then repeat the flying tests.

If you have a warped wing, there is no good solution. That is why it is absolutely essential to check the accuracy of wing panels (particularly foam or ARTF models) before you start to build and reject them if they are not up to scratch. There should be at the absolute maximum 2mm (0.080") mismatch at the trailing edge of the inner rib (or root) when the wingtip ribs (or tips) are parallel - anything else is junk. Adding a trim tab may help but the amount of screwing out that will occur will depend on airspeed, something that an accurate wing avoids. Start with a trim tab of approximately 25mm (1") long by 19mm (3/4") wide made from 1.6mm (1/16") ply as near the wingtip end of the aileron as posiible. You are trying to increase the lift on the "dropping" wing. It might, just might, be possible to do some very nifty mixing programs on a computerised transmitter but I wouldn't bet on it!

It is possible that you may not be able to entirely eliminate screwing out of both loops and bunts due to minor wing warp. In this case, the best option is to set the model up so that it tracks correctly when executing loops and use rudder to correct when executing bunts. The reason for this is that you will inevitably pull more looping or positive 'g' manouevres and so you will have to correct tracking less often.

Return to top of page

Checking The Centre Of Gravity (CoG) And Reflex.

You should run through the Thrust Line and Wing Balance tests to ensure that your set up is still valid - in fact you should continue to run all previous tests as you proceed through the trimming procedure to verify that any modifications do not invalidate any previous work. Establishing the correct position for the CoG involves a number of checks. Assuming that you have set the CoG in accordance with the instructions that came with the model or plan, you should be in the correct ball park. Generally speaking, the position of the CoG shown on models is conservative - it's too far forward for most aerobatics. The reason for this is obvious - test flying a model is exciting enough without the CoG being too far to the rear. Once the model has been successfully flown and "sorted" the CoG can then be moved progressively to the rear as required.

If you've never come across reflex before, reflexing the wing is done by setting both ailerons slightly up from the neutral position. The effect of this is to decrease the lift generated by the wing. What??? Sounds stupid but....

If you reflex a wing, you are altering the fundamental aerodynamic properties of the wing section, particularly with a symmetric one. Flaps are an example of this but in reverse- when the flaps go down, there is increased lift on the wing. Reflex is a permanent negative flap so the lift on the wing is reduced (when the model is right way up) but more importantly, it is increased when the model is inverted. The trick is to set the ailerons to the correct reflex angle to minimise the difference in elevator input required for normal and inverted flight. What you are trying to achieve is the best balance between the lift generated in normal flight with that generated in inverted flight. At the end of the day, this will be a compromise for the simple reason that a wing requires to fly at an angle of attack to generate lift.

Take the model up to a safe height, fly straight into wind and execute a spin. Make sure that you use only elevator and rudder to initiate and stay in the spin. If the model goes into a slow spiral dive with the rudder and elevator held on, there are two possibilities. The first probable cause is the CoG is too far forward, the second is that there is insufficient up elevator and/or rudder to keep the model in a spin.

Obviously, the quick solution is to try increasing the elevator and/or rudder throw. Do this by using one of the positions on the elevator and rudder rate switches - you can then switch the increased throw off if the model becomes too "twitchy" for you. You may have to resort to permanently switching to the high rates for spinning (and perhaps landing) and the low rates for normal flying if you find the required high movement too high. If this doesn't do the business for spinning then the chances are that the CoG is too far forward.

Before you alter the position of the CoG there are a couple of other checks to be carried out.

Take your model to a good height (sufficient for a two or three second dive!) and head straight into wind. Cut the throttle, push in elevator to go into a vertical dive then release all the controls. The model will do one of three things. Remember that to avoid wasting all your building and trimming efforts, pull out of the dive in plenty of time!

1. The model will continue in a vertical dive.
2. The model will pull out towards the cockpit side.
3. The model will push out to the wheels side.

Make a note of what the model does then check the wing reflex. Fly the model horizontally at about half throttle then roll it inverted. Release the elevator and watch what happens.

1. The model will continue horizontally.
2. The model will gain height quickly.
3. The model will lose height quickly.

Make a note of what happens and land.

If your model spins easily, continues in a vertical dive without pulling out and flies inverted with no elevator control - perfect CoG position and reflex!

If all this good stuff isn't happening, then either the CoG position or the reflex (or both) are incorrect. Since setting the CoG and reflex are going to be a compromise, change either individually, re-fly the tests and re-evaluate what happens.

To increase the reflex, disconnect each aileron in turn. Depending on how the aileron servo/s are mounted and how the linkage to the aileron is set up on your specific model will dictate whether you have to increase or decrease the length of the mechanical connection from the servo to the aileron. Whatever the set up, turn the clevis at the aileron two full turns so that the aileron sits slightly up from the top wing surface when the clevis is reconnected. Ensure that the aileron is securely re-attached.

To decrease the reflex, adjust the mechanical connection from the servo to the aileron so that the aileron sits slightly down from the top wing surface.

Altering the CoG should be carried out in small increments - no more than 6mm (¼") at a time. Commercially available self-adhesive weights are ideal TEMPORARILY. When you finally achieve the correct CoG position for flight, make a note of it's position. Remove the self adhesive weights and do something usefull inside the fuselage to change the CoG position - move the battery pack or servos around to suit. You may need to have the battery pack in the rear fuselage. (I've sat here for ages trying to make a sentence using "add", "lead" and "weight" but there's no way I can fit these words together!)

If everything has been sorted out to your satisfaction, the next step you need to take is to sort out all the servo movements and control throws.

Return to top of page

Setting The Servos And Control Throws.

If you have got as far as this, you have gone a lot further than most in setting up your model properly!

The next step is to make the most of your radio gear's capability, whether or not it is computerised.

By the time you have got your model to this stage, you'll have messed around with the trims, the dual rates and the servo travel. You almost certainly will have reduced them from the original set-up. Computerised gear is great but it makes you lazy! Rolls too fast? - set the aileron servo travel to 50%, fly on rates 60%, need a bit of trim? - use the trim offset memory. Sound familiar? Guilty!

We've all done this, the $64,000 question is "what next?".

A servo is designed for operating over a typical arc of ±30°. In the (not untypical) case above, the aileron servo is actually operating over 50%x60%x±30° which is....±9°! What's wrong with that then? If you look at the power output from a servo, generally speaking there is very little power output over the ±6° in the centre of the servo's travel - in short you have programmed away most of the servo's power that you'll be using. The "blow-back" or air resistance will greatly reduce the actual movement of the control surfaces.

The first thing to note is the physical movements that you currently have set and any offsets for ailerons, elevator and rudder. What you are trying to do now is to maximise the servo travel for the maximum required control surface movements so you reset the dual rates, trims and servo travel to maximum - if you have computerised gear, these can be set to above 100%, provided that you don't make the servo "over-travel".

The next step is to get back to your original movements using mechanical adjustments only. To reduce the control surface movement, you will have to move the clevis at the control surface as far away from the hinge line as possible. If the control surface movement is still too great, move the clevis at the servo end closer to the center of the servo. Make sure however that the clevis doesn't bind up with the servo arm, particularly in the "pull" direction. (A good way round this is to use ball joints fixed to the top of the servo arm rather than clevises). By the time you've done this, the neutral position will probably have moved so disconnect one end of the control rod and screw/unscrew the clevis until you return to your original neutral position.

Carefully check that all the control surfaces and servo arms are free moving over their entire range of travel and that the neutral positions and travel ranges are as before. Now reset the low rates to about 75% of maximum travel.

The model is now ready for another test flight. My experience is that after you have carried out this process, the model will be "twitchier"! For "twitchier", read "more responsive". This is entirely due to a decrease in "blow-back" and effective use of the servo's power range. You may have to reduce the control throws again but by now, you know that this should only be done temporarily on the transmitter. Establish the new required movements then alter the mechanical linkages to suit. Something for nothing? Sure is!