I'm not sure that this assembly is physically possible: since axes of the 2 rollers are fixed, the distance between the 2 rollers is fixed also, so the profile of the cam should be such that the 2 rollers always touch both sides of the cam. If the cam is a centered cylinder then it's trivial — and not very useful — but if it's some free shape for some fancy movement then it might not be possible with a single cam. Ducati has developed a push-pull cam system they called desmodromic valves, but they need 2 cams, one for each movement. Alternatively, you need a single roller and a spring.realthunder wrote: ↑Tue Oct 29, 2019 11:30 pmfreecad-heini-1 wrote: ↑Tue Oct 29, 2019 9:17 am Pistons with cylindrical cam rollers should be mounted on the cam discs.
...be handled by a physics simulator.
Doesn't help you with the constraints in the assembly though. If your cam has a mathematical expression — and I bet it has — you should use a spreadsheet, expressing the height of the roller -vs- cam rotation angle. That would be your poor man's physical simulator.