I retested by trying to separate the rotation of the Sharp rotor from that of the VAWT as much as possible. The wind speed was at least 10 m/s during the test. The Sharp rotor retained its diminished rpm (orange adhesive light) but, as can be seen by looking at idling at 0.25, it appears that the VAWT had a faster rpm (look at the brown adhesive light).
As in the previous test, and despite its relative independence of rotation speed from that of the VAWT which surrounded it, the rotation of the Sharp rotor was considerably slowed down, and its lift reduced in addition.
Then I tested the same VAWT but alone, with a wind that became weaker, perhaps 8 m/s maximum: the rpm was significantly higher, around 500, which gave a TSR of a little more than 1, which remains low but understandable given the condition of the blades.
This would confirm that it is better to put the VAWT side by side with the Sharp rotor. And it is possible that it is preferable for each of the rotors to have their freedom of rotation, even if it is tempting to tune the TSR of each of the rotors by fixing them side by side. All you need is an axis around which the rotors rotate freely.
However, the simplest solution is to attach an inflatable Flettner rotor (like Omnidea balloon) side by side to a VAWT which will “decide” the TSR of said rotor (and also its own TSR with the Flettner rotor) according to its diameter, of the order of 1/3 of the diameter of the rotor of the VAWT, for a TSR of approximately 1 for the Flettner rotor, and 3 for the VAWT. The energy consumption of the Flettner rotor remains minimal compared to the power of the VAWT.
I remember to have tested a foam cylinder of 0.54 m span and 0.27 m diameter glued side by side to a foam Savonius rotor of the same dimensions. The whole thing flew for a short time at a correct elevation angle of 40 degrees.
The Darrieus type VAWT that I experimented with does not generate a Magnus effect but is more powerful than a Savonius rotor and can sweep much more.
If we can use the shapewave® technique for the blades and their profiled uprights (which moreover can perhaps also be stabilizers of the whole), and as the Flettner balloons are easy to make, we would have a relatively lightweight, with a rope drive transmission around the Flettner cylinder(s).
