Hi, I'm working on a mechanism that is supposed to simulate a transmission composed by a chain drive and a gearbox. The input torque is generated by a DC motor. Each reduction stage has a specific inertia, given by the pulleys (for the chain drive) and the wheels (for the gearbox). The bodies can rotate along their principal axes, but they have all the same base.
I was wondering if it was possible to model the pulleys and the toothed wheels with Multibody revolute joints, therefore adding to the driveline network multiple "Rotational multibody interface" blocks. If I try that, the velocities of the bodies do not match the expected ones. "Pulley1" should be the fastest, while "wheel", should be the slowest, due to the reduction ratio of 116. But by measuring the angular velocities on the joints we can see that the wheel is actually the fastest, while the pulleys are much slower, by 10 times.
If I measure the angular velocities through the motion sensors, however, they match the expectations.
Is there an answer to this mismatch? Is it actually possible to use the interface blocks in this way?
This is the model:
The subsystems contain the interface block, the revolute joint and the solid block:
This is the mismatch in question, where "enc" represents the motion sensors, while "w" represents the joint velocities: