After the constructive criticism of the major disadvantages of a vertical trajectory with a power kite, I return to the Magnus effect balloon for which such a trajectory is planned, by presenting a method that I had outlined (see the three comments from the link): transmission not by two axes at the two ends, but by belts placed along the balloon, in order to limit the stresses on the structure to allow greater scaling.
I put again the video, after having reviewed some results. I measured (with a rpm-check) rpm 3200 with the 8 cm diameter wheel, by using a drill given to 500 W and rpm 3000, so 13.4 m/s tangential (wheel) velocity without the balloon, then rpm 2800-2900 with the balloon (about a max of rpm 240 for the balloon of 3 meters in circumference), so about 12 m/s tangential (wheel and balloon) velocity.
The decrease in speed with the balloon is therefore not enormous and corresponds to the margin of inaccuracy and losses due to oscillations (balloon not precisely centred), the belt more or less well tensioned, the shape of the balloon… This gave me (counting my own errors) 141 W instead of the theoretical 41 W according to the formula below which seems reliable to me. This is not so bad and it can be much improved with clearer material.
To evaluate the power consumption I use the formula “Power = Cf ρ U³ tan / 2 Area” indicated in the last page of the document below, where the tangential velocity is cubed, and Cf = 0.007:
The video (not new but deleted then put again):
A sketch for a larger balloon with several motors-belts and a long ground station:
The same with a small ground station:
Drawbacks: without guides, the belts can go, and the balloon can escape. In this case a huge mass of air would travel around destroying everything in its path, unless one end could open and let the air out; slowness and time lag between stop (reel-in yoyo phase) and start (reel-out power yoyo phase); failure on one of the motors…