# Is there a benefit to having the motors for a multi-rotor copter on the top vs the bottom of the frame?

There is a new quadcopter being launched on Indiegogo that has the quadcopter motors positioned on the top of the propellers as opposed to being positioned on the bottom.

Is there any advantage to this or is it just for looks?

• Welcome to Aviation SE! I suppose if flying indoors you don't risk hitting the rotors on the ceiling. Other than that, it's just for looks. – Keegan Nov 19 '14 at 5:40
• I would suspect something with compression versus tension load on the motor shafts, but I don't think you can get definitive answer from anybody except the actual designer. – Jan Hudec Nov 19 '14 at 14:24

It could be just for looks, but rotor placement has a noticeable effect on the behaviour of the quadcopter.

The higher the rotors are on the frame the higher the center of lift will be (the center of lift for quadcopters is the center of the plane created by the rotors.) A high center of lift means that the frame will be more stable. A low center of lift means the opposite. The reason for that is that the quadcopter can better naturally deal with lateral rotational forces on its frame. The center of lift being above the center of mass means that the frame is "hanging" like a pendulum from it. If you apply a lateral force to it, it will naturally swing back to place and the more you push the "harder" it becomes to move the frame.

A center of lift that is in the same spot as the center of mass of the frame means that there is no pendulum effect. This is like balancing a ball on top of another ball. All lateral forces, either internal like variations on each rotors' force or external like wind perturbations have maximum effect in moving the frame.

Contrary to what you might think, stability is not something you want in a quadcopter. The more unstable the frame, the more agile it becomes. Small variations in rotor force will create large rotational movements. The penalty you pay for that (small, unpredictable external forces also create large rotational movements) you make up with processing power. Processors are now so fast and sensors so sensitive that they easily can compute the external forces and create a response fast enough to keep the frame flying.

• Actually moving the centre of lift up or down does absolutely nothing to stability of quadcopter. There is no mechanism to shift or bank the lift vector when the aircraft is banked, so the lift vector passes through centre of gravity making the craft neutrally stable in either case. I also thought it affects stability before I was corrected by pdel here. – Jan Hudec Nov 19 '14 at 14:21
• I think that what you are saying is correct only if the lift vector does not change direction (pdel comments on that as well.) This is not true for a quadcopter. – ravingraven Nov 19 '14 at 16:38
• Sorry for the double comment but, take a look at this as well: eprints.qut.edu.au/33767/1/33767.pdf under "4.4 Design for Optimal Sensitivity" – ravingraven Nov 19 '14 at 16:43
• I will believe you there is an effect that can tilt or displace the total lift vector when banking, but I don't see it and I didn't find any conclusion in that article (and most quads don't have flapping rotors, so the discussion from that point may not even apply). – Jan Hudec Nov 19 '14 at 18:16
• I am not sure that I follow you to be honest. A quadcopter can rotate its total lift vector by rotating one or two adjacent rotors faster than the others. The lift vector will rotate towards the rotors that are spinning faster in relation to the weight vector that is always pointing down. This is how quadcopters move laterally... – ravingraven Nov 19 '14 at 18:46

The main reason is related to how the arms impact the thrusters' aerodynamic efficiency. The impact is lower when the arms are on the air-intake side of the propeller rather than in the downwash. It's actually quite surprising that most multi-rotors do not have downward-pointing props. One reason might be that it becomes more difficult to mechanically implement the landing skids with downward pointing props.

@abey is right, but his answer is a little short. Let me give some background.

Propellers work by accelerating air. If you think of a stream tube through which all air flows that will pass through the propeller disc, the diameter of this tube is bigger than the propeller diameter itself ahead of the propeller and smaller behind it. Since the air flows faster behind the propeller, the same mass flow needs less stream tube cross section. The propeller sucks air in, so the speed increase starts at some distance ahead, and raises pressure in the disc, so more speed increase follows behind the propeller when the pressure $p$ in the streamtube falls to the level of ambient air pressure $p_{\infty}$.

A strut placed ahead of the propeller will see less dynamic pressure than one placed behind it. Turn the picture above by 90° anticlockwise, and you have a quadcopter propeller arrangement. Make the initial speed $v_{\infty}$ zero and you have the situation for a quadcopter in hovering flight. Consequence: Propellers below the strut will have a higher efficiency.

Full disclosure: The picture above is idealized - in reality, the air behind the propeller will mix with air surrounding it, which will be dragged along. The streamtube idealization gives an easy way to calculate propeller thrust when combined with the assumption of identical air speed over the cross section of the stream tube, but in reality the speed distribution is uneven, with a gradual decrease with bigger distance from the axis of symmetry.