Hello,

I would like to model an ideal clutch. The only clutch that is avaible from the standard library is based on coloumb friction.

Have a look at my very simple model (without annotation):



It is a bit confusing to me that the idea of this approach works fine if I leave out the boolean input. So the equation phi_rel=0 leads to a rigid connection between th 2 flanges as expected when locked is true. If I use the codeline "0 = if locked then phi_rel else tau;" instead to include the outside boolean control this leads to a singularity. Setting locked=false (the clutch is opened) the model works fine. What is changing when introducing the boolean input? The resulting equations should be the same.

I´ve added a test environment for the clutch "IdealClutch_Test.mo".

OK, the longer I think about this model the less I understand it . I´m an electrical engineer and so I tried to compare an ideal switch with a mechanical clutch ...maybe that´s not the right way. Can anyone help me?

This is my next try and it´s getting better

The idea of considering an electric switch isn´t that bad!

I attach the new model using a parametric variable s. According to that I do not include an equation for the angle of the flanges anymore but for the derivative of the relative angle (the velocity).

It seems to be working proberly...

As far as I know the problem in your case is, that your function is not continuous, since you have a sudden change between total torque transmission an no torque transmission. If you have any derivation in your modell, Modelica can not derivate that function.

A switch is not a god replacement for a clutch since the sudden stroke, which is created if the clutch closes "digitally", would destroy mechanical parts. Therefore the transmission function must be a continous function.

You can compare that to a current, which can not change from 0 to 1 as a step.

Why not? In an electrical circuit in the case of a switch-on the current changes vom 0 to 1 and the voltage from 1 to 0. There is a small resistance and conductance included in the model of an ideal switch so it is not "ideal" at all but that is what I tried to translate to the rotational problem.

By the way, the model I´ve attached before is NOT working as it should. At the moment I´m working with the clutch from the standard library and I set the force to a great value in order to avoid friction and make it more ideal. I would prefer a simple model like I tried to develope and I´m sure there is a solution.