matlabsimscape

Automatically computed torque actuation of a revolute joint block error?

Im simulating a swing door with generator using Simscape/Simulink. I imagine there is only one input for system which is the force at the knob needed to open the door, so all revolute joints have their torque actuation set to automatically computed. However, im getting an error saying:

"In the dynamically coupled component containing Revolute Joint 'SDL/SwingMotion', there are more joint primitive degrees of freedom with automatically computed force or torque (4) than with motion from inputs (0). In general, the equations of motion do not have a unique solution. Solve this problem by increasing the number of joint primitives with motion from inputs or reducing the number of joint primitives with automatically computed force or torque. Resolve this issue in order to simulate the model."

The animation works fine if i set the torque actuation of all 4 revolute joints to none, but the torque produced by the force wont be transfered this way and the generator shaft wont have any torque. However, I am able to measure the output RPM (angular velocity of generator shaft). I dont quite understand the error. Why do the revolute block treats the degree of motion of other revolute blocks as its own? How do i resolve this? Block Diagram can be seen here.

Block Diagram

The Assembly looks like this:

Assembly

Any help is much appreciated!

Solution

  • For a revolute joint there are three options:

    So you have to select either one of the option. If you select torque automatically computed, then you need to provide the motion which it has to follow. If you don't need to follow a provided motion, than no torque is needed. If you set torque computed to 'none', then there is no torque that can be exerted from the base to the follower and vice versa, that is the idea of a Degree of Freedom.

    It seems as if you want the Base and the Follower to be somewhat rigidly connected and follow the same motion. Than you can consider using the Rigid Transform block, which is just a rigid link in which you can define a translation or rotation offset.

    EDIT

    What you were effectively doing was combining forward and inverse dynamics. You put a force on the doorknob, let Simulink calculate the motion (up till now forward dynamics) then you desire the torque on the GenShaft from the motion it is doing, which is inverse dynamics. That does not work.

    A better way to check the torque on the GenShaft is to, for instance, desire a certain door hinge angular velocity, put a proportional controller on it, and check the required torque. Notice that with no friction modeled in the hinge the required torque will go to zero.

    So if you set all torque computed to 'none' except for the GenShaft, set this to provided by input. Then put a proportional controller on the angular velocity of the door hinge. You can then check the torque needed.

    Updated model

    Updated model