image

No Single Predicable Connector

The image above represents a mechanism that the Linkage program cannot simulate. The simulator code only looks for predictable connector positions and there are none in this mechanism. Let me explain: The simulator will look for two links that each have a connector with a known position and with another connector in common. It’s quite simple to then determine two circles centered on the known connectors and of the appropriate radii for the common connector. The intersection of those circles is where the two links are connected.

A quick look at the picture tells me that there are no links that have one connector at a known position with another connector in common.

Simulating this mechanism will require an extra step to find groups of links that can be simulated within the group. Or even simpler when the group contains connectors of all fixed lengths. here’s sort of how it would work:

  1. Find that GEJ is a triangle made from multiple connectors.
  2. Simulate links 1,6,8 to deal with the actuator.
  3. Take the new triangle GEJ and create a dummy link that has the right proportions. it would include connector A, which is very important.
  4. Simulate the entire mechanism including the new dummy link.

After that, the dummy link would qualify for the basic simulation code because the dummy link would have a connector at a known position, A, and would also share a connector with another link that also has a connector at a known position. the dummy and link 3 get simulated and the dummy link is moved, along with all of the connectors GEJ and the rest would work from there.

The trickiest part will be finding these triangle pieces of the mechanism and simulating them to create the dummy link. I think that if I simply pretend like all of the connectors are at a known position except for one connector of the actuator, it will work fine. The code is probably not modular enough to accept a second mechanism to simulate, so that will also be a bit of a coding burden. But the concept seems simple enough.

And I’m psyched to get this mechanism simulating because the actuators are then much more powerful.