Hi.  I'm learning about collision detection in Solidworks for the first time, and I wanted to ask about using collision detection to define an assembly's motion as opposed to using mates to do so.

For example, say you have a basic hinge design, with interlocking concentric rings attached to each of two parts, and a pin that goes though them.  It seems to me that, once the parts are properly assembled, instead of using mates to define the parts' motion, you could enable "Collision Detection with Physical Dynamics" between the parts, suppress all the mates (with the possible exception of restricting the pin's movement along the axis of the hinge), and the hinge would behave as intended.

While it seems clear to me that you CAN do this, my question is, is it good practice?  On the one hand, it creates realistic motion (i.e. the parts are restricted primarily by their inability to overlap, which is how it works in real life), which can be useful; but on the other hand, I expect it would be more computationally intensive with dynamic motion compared to just using mates, especially with more complex assemblies.

And the project I'm thinking of doing this with does indeed have a lot of moving parts.  Imagine simulating a bike chain to get an idea of the sheer quantity of individual part interactions that I would be keeping track of in this manner.  I want to do it for the realism, to verify that the mechanism's design works the way I intend it to; but it it impractical (or just considered not good practice) because of the computational intensity?  Thanks.