Cineversity

I’m attempting to rig a fairly complex type of locking pliers using Dynamic Connectors for the first time.  I’m having mixed success.

After watching some of the tutorials and playing around with their project files I was able to successfully rig a slightly simplified version of the pliers.  A distilled version of that project is in the zipped file link below (WorkingConnectors.c4d). 

http://marshall-arts.net/Support/DynamicConnectors.zip

In this version I treated the two linkages that go from the top handle to the lower jaw as one locked piece.  They are supposed to hinge at their connection point and straighten out when the pliers lock onto an object.  Since most of the rig is working I could probably live with this version by not explicitly showing that locking feature.

That said, I wanted to make a completely accurate version of the rig, so I added a hinge between the front and back linkage (Not_Working_Connectors.c4d).  Everything works from the top handle through the jaw when I run the animation.  Things go wrong when I try to add the Moving Handle to the mix by connecting it with a hinge to the Front Linkage.  When I run the animation the Back Linkage and Moving Handle immediately skew off of their pivot points, and everything goes out of alignment.  If you deactivate the dynamics of the Moving Handle the simulation runs correctly.  That would lead me to believe that I’ve done something wrong attaching the Moving Handle to the Front Linkage, but I don’t know what the problem is.  Obviously.

I don’t have collisions enabled on any of the connectors.

Any thoughts as to what the problem is?

Also, I’m not married to this solution of using connectors.  If there’s a way to do it using constraints I’d prefer that, but I didn’t think that was viable given all of the interconnected parts.

Thanks.

Shawn Marshall
Marshall Arts Motion Graphics

Hi Shawn,

I’m not certain if the amount of inner mechanics and dependencies might be the problem.

Having worked a lot with similar tools in the past (Motor-cycles) , I know that they don’t work very well if they miss out the spring. In fact it is a lost tool that slows down the work rather than it supports it if they miss something. Adding a spring to support its natural movement would add even more complexity to it.

Has that original tool a spring?

I have no idea where to start looking for an alternative, as I have a) not the tool here, and b) if I would like to—I would like to build it from the scratch. Perhaps with the plan of the animation and cache steps to simplify the calculation—just thinking out loud.

My gut feeling tells me always, don’t think about reality, think about the visualization that “you” are after. It has shown over the time, that 3D works better if the end product is in mind and not to replicate reality.

Having said that, I would try to do it with joints and IK, and perhaps would animate the “Moving Head” and go from there. The “Bar Connect” 1-2 would be as well just animated and then the “Moving Handle” would follow the Bar Connect 2 (pair) perhaps in a dynamic on its own.

Do I avoid here to solve the problem? Perhaps—you might think. But why solving something that could be done simpler and stable otherwise. I remember when Dynamic2 came out, how many people tried to create a engine simulation (piston, crankshaft, rod,  etc). I have yet to see a full dynamic solution with that. Mostly a simple XPresso did the trick back then. I shoot for things that give me my images, I’m not so much in the “turn C4D into a simulation game”. Perhaps someone else might try it, or a one-on-one session. :o)

All in all, it is certainly a nice test and a great puzzle to see where and how the power moves. Having studied those stuff here and there a little bit, I might see too many problems sometimes, and work so on getting simpler versions to meet my visual goal.

All the best

Sassi

Thanks for the quick reply.

Just as I was about to post this I figured out what was making the simulation glitch.  I had the Moving Handle’s Shape set to Automatic in the Dynamics settings.  Changing it to Moving Mesh stopped the simulation from immediately jumping out of alignment.

There is a spring on the tool, but its main purpose is to hold the handle and jaws apart when they’re not locked onto something.  In this simulation gravity is performing a similar function.

The WorkingConnectors version of the rig is pretty close to how it’s supposed to move.  The only difference is that the two linkages are being treated as one object in that simulation.

I would rather do this WITHOUT a dynamic simulation.  I have no objection to “faking” the movement as long as it’s correct.  My problem is I have no idea how to approach something like this that has multiple pivots points and linkages all interacting with each other.  Maybe I need to study some IK tutorials.

Cheers.

Shawn

Hi Shawn,

Nice find. Thanks for sharing. I switched the visibility of the objects off, as well as the collision, but I saw some jittering in the Joint/Slide combinations; I thought that was the problem. Which normally shows a conflict in the forces, or most of the time a scale/contact problem. Like putting a force on a slide that can’t be easily solved. When it is working now, great. Perhaps for the dynamic use simple polygon object, that can’t collide, but have the same distance in the relevant points. The plier parts could be then just a child of each, with no dynamic relevance. I hope that makes sense.

Yes, doing it as simple as possible is my philosophy. I tried to explain the steps above, and rethinking the change of dependencies, once the direction of events is centered or “on the head”. With centered (which is mostly where I start first, is to check if the element in the middle can move all parts, which splits the “chain-reaction” in half. With on-the-head, I thought about the front element, as that is what is normally targeted as movement. With its two freedom movements (sliding up and down as well rotating around a sliding point) you take already two difficult movements out of the dynamic. The pars directly connected follow then like a IK. The handle comes then and when all is done. Yes, it is not that simple, as the handle has also some movement options, but once anything else is a given, the handle follows. Perhaps there another option to simplify the chain and group the decencies. Trial and error—no doubt.

All the best

Sassi