Sum of torque

[EOW]

Maybe if I transmit the torque from one arm to another.

The yellow arm can turn around the pink axis. The black arm can turn around the white axis (center of the circle). Each arm has a different torque, higher to the yellow.

[EOW]

I drawn all forces and details. I supposed the torque can be at 2/3 for the yellow arm and 1/3 for the black arm, it's an example. Like I can adjust the law of attraction I can have this example. But maybe my forces are wrong.

With gears it's logical.

I can rotates the device around the pink axis to have a continuous movement. Look at the epicycloïde. But I need to have N device with a phase angle at 360/N.

If the wheel moves to the right the energy is destroyed, if it moves to the left the energy is created.

[EOW]

I had a problem with the force from the gears, like the radius of the gear is 2 times lower, the force is 2 times higher so the sum of energy is at 0.

Now, I used a torus (like a wheel of a bike that moves in translation and rotates). There are small balls in a part of the torus.

[dieter]

I am currently using the open dynamics engine ODE with an Api wrapper for Blitz3D.

So now I can do this in a BASIC programming language / compiler. This is good not only due to simplicity, but also due to extremly fast compilation/run times, compared to bulky, complicatified Cpp paths.

But the accuracy / precision of ODE is low, so maybe it's too low for you. Anyway, I could send you the wrapper. ODE is open source, Blitz3D too, on github I think.

[EOW]

Thanks for the name of the software  do you use it to simulate some devices ?

I can give the torque from the ground and give the force at the bottom from the ground in the same time.