How to sense the torque with the revolute joint?

Hello there,

based on the following simscape multibody example, I would like to measure the torque at the joint (see attached file). Is there a way to sense the values automatically? When setting "automatically computed" in "Actuation" there is an error message appearing.

How can i sense the torque instead?

Thanks in advance :)

[Nathan Hardenberg]

I don't think the way you want to do this is possible. But I'm not 100% sure, thats why I'm only commenting. You can only sense the torqure resulting from movement. The internal mechanics probably just calculate the torque internally. You could manually calculate the torque by measuring Position and Velocity and knowing the internal mechanics, but thats not really the point of simulating.

What you can do, is providing a desired (position) trajecory and measure the needed torque. This has the disadvantage of not knowing the "natural" movement. But for a steady state case this can be useful, since it provides the torque at the current position (disregarding any velocity)

Example for steady-state:

[Nayla]

hi @Nathan Hardenberg, thanks for the comment. What's the name of the block you applied for the steady state, I mean for the constand "deg2rad(0)" ?

[Nayla]

By sensing the velocity instead and then multiply it by the damping coefficient I can also get the Torque. Is this also a correct way to determine the torque?

[Nathan Hardenberg]

The name of the Block ist "Constant". In the Constant I use the deg2rad()-function to indicate that the input of the joint is in radians.

Regarding your other comment: What you describe does not calculate the torque. One example to prove my point is when your damping coefficient is 0. Your toque then should not be 0. I'm not into this topic at the moment that's why I can not tell/know how to calculate it.

Answer 1

[Nathan Hardenberg]

I tought about it a bit more: The torque is just dependent on the deviation to the natual position. The damper creates a torque in the opposite direction which slows down the movement, but the current toqure on the joint is just depentant on spring stiffness. Since the movement is not the problem here (it gets simulated with the multibody) the spring-torque T can easily be calculated:

with

• K : spring constant
• : diffence to equilibrium

Obvioulsly the values in the Revolute-Joint and the values in the calculation have to be the same