Getting the most from your model: Centre of gravity or CG for short
So you just bought a plane or glider and put it together, the manual gives you a range of where the CG should be. You balance it on your fingers or use one of the many balancing tools out there, you put some weight in the nose or move the battery around to get it within the range that’s in the instructions.
That’s it? Done, right? Let’s go fly!
Sure, it will fly, assuming you measured it correctly and got it within the range. However, if you didn’t measure it properly or measured it wrong, you might be in trouble.
A phrase that was passed down to be by somebody many many years ago, the name lost to the mists of time:
A forward CG will fly badly, a rearward CG will fly once.
The meaning behind this is that if you have the CG too far forward, it will fly, but it will just not fly very well, it will require lots of up trim and feel soggy on the elevator. If you have the CG too far back, it will be unstable and not want to fly at all, as soon as you throw it or throttle up it will be uncontrollable on the elevator.
So for the first flight it’s best to be sure you have a forward CG, some designers will factor in some safety margin in their designs so their range of acceptable CG will be a little forward so don’t get too carried away, just make sure it’s within what is suggested in the manual.
The point is to get it in a safe location, which you can then run a few tests when you fly it and decide what to do from there.
Last point which has caught me out before: make sure that the control linkage for the elevator is slop free and free to move and that the hinge line for the elevator is solid, any problems in this region will give a false reaction to the tests we will run when flying.
CG position false information
Over the years I have heard a bunch of fallacies about CG position.
• “I’m going to move the CG forward so the nose sits a bit lower”
➔ CG does not affect the attitude of the model, the model maintains the same angle of attack no matter the CG position, moving the CG forward requires up trim to balance it out.
• “It keeps climbing, I’m going to move the cg forward to keep the nose down”
➔ For reasons I will explain further down, the opposite is actually true. Moving the CG forward will actually make it climb more when you increase the speed of the aircraft.
What’s the theory?
Before the aerodynamic pros shoot me, this is a simplified explanation, if you want to go in depth, there are other sources!
On a conventional aircraft with a normal horizontal tail and elevator, the centre of lift on a stable aircraft will be slightly behind the CG. The Lift pulls upwards, the CG pulls down, resulting in the nose trying to go down, the tail then counters this with a slight down force.
The key fact to remember is that the downwards force from the tail is dependent on air speed, and whilst the lift from the wing does increase as airspeed increases, its leverage on the CG is much lower. So as the speed increases, the tail creates more downward pressure which pulls the nose up, assuming the thrust from the motor (if you have one) stays the same. The speed will then drop off, the downward force from the CG overcomes the force from the tail and the nose drops.
If you move the CG forward, you need more up trim at a certain speed to counteract the force from gravity to keep the nose up. This means that the tail produces more downward force faster as the speed increases, which pulls the nose up earlier.
The model will just want to travel at one speed for a set elevator trim, increasing throttle will make it climb. If you drop the nose but keep the throttle the same, it will also speed up, but then the nose will come up and it will return to its previous speed.
For learning, a reasonably forward CG is preferable. It makes the model more docile, though on models which have a wide speed range, you will find yourself messing with the elevator trim a lot or having to hold a lot of down elevator on those joyful, fast low passes. The downward force from the tail also produces drag, so a forward CG will be less efficient.
Conversely, if you move the CG back, then the lift from the wing will be in front of the CG. This will mean you need lift from the tail rather than downforce, which will pitch the nose down as the speed increases, i know from experience that this is far from fun!
Ok enough theory, how can I tell if my CG is right?
My favourite test, which works extremely well on gliders, is the dive test.
First, trim the model out so that hands off the Elevator, it’s going at a fairly steady speed, not too fast, a little above minimum sink. This is extremely important as if you get that wrong, you will get a false result.
Fly from left to right or right to left, push the nose down to a 45 degree dive and let go of the elevator.
Depending on the CG position you will get a different result, a very forward CG will result in the model pulling the nose up very quickly, maybe even to vertical
As you move the CG back (don’t forget to retrim), the model will get closer to continuing straight on and not pulling out, go too far and it will start to tuck under.
If you are a beginner, you want it to pull out to level and maybe climb slightly. As you get more advanced, especially on a slippery carbon fibre moulded glider, you want it to be closer to neutral.
The fact that powered models tend to be a lot draggier than gliders means that a dive test can be inconclusive, simply because they do not gain enough speed in the dive. It also adds a complication of the thrust line. If the thrust line is wrong and pointing upwards, then when you throttle up, the thrust will raise the nose which might hide a rearward CG.
A few clues can get a general feel for CG position:
How did the elevator feel? If it was twitchy and, during the landing, did it feel like it was trying to raise the nose and you were having to hold it down? These will suggest the CG is rearward.
If the elevator felt soggy and, during the landing, it needed a lot of up elevator to slow it down, they are clues that the CG is too far forward.
3D Aerobatic Models
These are a little different. If you are just sport flying, then the previous segment is enough. However if you are getting into prophanging (hovering), knife edge, flicks spins etc., Then you want the CG as far back as you can bear, holding down elevator as the model slows for landing is not uncommon. Prophanging a model with a forward CG is basically impossible, the further back you can get it, the easier it will get. Do not just throw the battery back as far back as it will go and place some lead on the tail, work back slowly. The point behind which the model becomes unflyable is very sharp so approach with care.
One good test is to climb at 45 degrees on full throttle inverted; this is similar to the dive test on gliders. The model, if trimmed correctly with a nice rearward CG, should carry on that 45 degree climb.
A final point about elevator travel
Outside of 3D aerobatic models (where more is basically better!) elevator travel is important. I often see people setting their models up with way too much travel. This is a path to trouble, as at some point you will get it wrong and haul on the elevator, at that point the model will give you a hard pass and screw itself into the ground. This is more important with a rearward CG, as most models are more likely to tip stall more aggressively.
You will see this occasionally at the field, somebody takes off, the motor quits, they panic, haul on the elevator and the model just falls out of the sky. Or, at the other end of the spectrum, they come through, fast pull on the elevator to do a square loop, but instead of pulling up, the model rolls on its back and, still with a load of elevator in the model, spears the ground.
There are 2 ways to test for elevator travel, both equally important. If this gives 2 different values go with whichever is lowest.
In both cases make sure you have plenty of height spare!
High airspeed full elevator loop. Be careful with this one, only do this if your model is structurally able to endure this.
Get the model moving quickly, lots of throttle or a dive (if a glider) then pull the elevator, it should do at least one loop without screwing out or flick rolling. Reduce the elevator until it goes straight around the loop.
Slow speed, drop the throttle back to around ¼, then slowly feed the elevator in. It might try to screw on you a little, but you should be able to pick the wing up with a combination of aileron and rudder (if you have one). Reduce the elevator travel until you can just about fly it around on full elevator, it should be hard work, but not impossible.