Sum of torque

[EOW]

I move in translation all the device. Like that I can have -X and +X for the velocity on the green part. The green part receive a torque. The arm don't receive a torque and the disk accelerates the rotation around itself.

[EOW]

With the pressure with the law 1/d². The torque from the SideA is not the same than the torque from the SideB. I drawn only 2 springs but there are in all the blue surface.

[EOW]

In the last image, all the black lines are link together. All the device I drawn turn around the red axis. Black lines "separate" the pressure from the top.

On the SideB the mean pressure is 0 if the number of black line is infinite. But with 100 lines for example the force is decrease by 100.

On the SideA, the torque is integrale(x²(x-R)dx) from 0 to R = R^4/4-R^3*R/3

With R=5, the torque on the SideB is hum maybe 5 and it is 52.0833 for the SideA.

There is no torque from the curved shape because the device turns around the red center and this is the center of the half circle.

There is no torque from straight horizontal lines and springs because they cancel themselves all torques IF the number of horizontal lines inside the half disk is high.

The attraction can be in another law than 1/d²

I added an image, N is the number of black lines. Here N=9 for example. At top in each slice the pressure is 0, at bottom the pressure is R/N.

I drawn the device after 30°, to look how the device rotates, all the device rotates.

[EOW]

Image 1/ With half disk the sum of torque is 0

Image 2/ With a full disk. Springs are only in the right half disk. Springs push balls. The law of repulsion for springs is like d². The force is higher when the cos(angle) is lower so there is a difference.

Image 3/ With balls inside in PartB and balls outside in partA1/PartA2

Image4/ With water. F1>F2 so the disk want to turn clockwise. F3 gives a clockwise torque to the device around the red center. All the device is always in that position because I synchronized 2 angular velocities.

[EOW]

I came back with this device. If I take a half disk, the centrifugal forces turns the half disk and the arm receives a torque. The arm turns at w (labo reference) and the disk turn around itself at w.cos(a) (labo reference) like that the disk is always in the position I draww.