Sum of torque

[EOW]

The calculations for the torques in the first image. Eq1 is the torque from the centrifugal force from the half disk itself around itself to the arm. Eq2 is the torque on the arm from the centrifugal force around the support. Eq3 is the torque on the disk. With:

d: the distance of the arm
alpha: the angle
R: the radius of the disk
w: the angular velocity of the arm

I can take 2 half disks like the image shows, like that Eq1=0 for 2 half disks. And like d1 different of d2 the torque on the arm + torque on the half disks are not at 0.

[EOW]

The sum of torques is not 0.

[EOW]

Better, if the center of the quarter circle of the white shape is the red center not the black center.

All the last device I explained are mounted on a device with an arm with an angle and a disk, I drawn only the disk, look at the third image.

[EOW]

With the spring. The arm lost the energy FRwtsin(a)cos(a) and spring win the energy FRsin(wt.cos(a))sin(a), the sum of energy is not constant. It's like to compare:

b.cos(a) and sin(b*cos(a))

and the result is not the same, an example:

a=2.09
b=0.1

The result is -0.05 and -0.049979

Another example:

a=0.001
b=60*pi/180

The result is 1.74e-5 and 3.04e.7, the relation is 57.29 = 180/pi, it's the maximum of the relation

[EOW]

At start, the disks don't turn around themselve in the lab reference. The arm turns at w. The spring gives the forces F1 and F2. Like the disks don't turn around themselves the lenght to the spring changes very few. But the forces F1b and F2b give a torque to the arm, this torque depends of w.