Double-Ended Ball Joint/Socket Issue
@StellarFusion The "squirrely behavior" in your design and the reason you are receiving error messages about joint conflicts is that after your Ball37 joint your design is simply over constrained.
When trying to apply just one additional ball joint between the correct upper part and one of the free truss ends The joint solver simply cannot connect the two joint origins, because they don't precisely align. Precisely in this case far exceeds what you can measure with a caliper. You'd have to have a rester electron microscope to measure that. Or better. The precision needed is the internal precision of Fusion 360.
The root cause for that lack of precision are the fully undefined and imprecise sketches in the design.
That is also the one and the only reason why this works in the last two designs I posted. If you start with constrained and dimensions sketches it will work, otherwise it will not. There is really no other magic at work here.
Another thing that no one has mentioned here is that this design is a fully constrained, rigid design. As it pertains to Fusion 360 the ball joints are completely unnecessary to create this design. It can in fact be created with a single rigid group joint.
@jeff_strater I also have seen "squirrely behavior" particularly when using ball joints.
The reason for some of that behavior can probably be explained by ball joints not being implemented using quaternions as one of the developers has explained in depth in this thread. She also provided excellent reasons why that choice was made.
As everything in life one has to make compromises along the way and if its known that there are somewhat mathematical reasons for this then maybe precautions can be taken to help the joint system a little along when creating such an assembly. So I did use a little trick .
First I used a slider joint between the upper and the lower frame, with some sensible joint limits. Setting sensible joint limits is always a good idea. That however did not work and I also got a joint conflict. So I opened up one more degree of freedoms and changed the slider joint into a cylindrical joint. Voilá. Problem solved 😉
However as I've said before and as you have confirmed, there is absolutely no way to get this to work when one does not work precise. That is not just true for Fusion 360 but for any other CAD system. One would run into these same problems using Solid Works or any other CAD system.
@chrisplyler wrote:
Dude, I'm getting awfully tired of seeing you help people with examples that aren't rendered and self-illuminating. What gives?
Uhh, yeah, sorry for slacking off.
Would this work? It did not quite turn out the way I had hoped. I've yet to produce a single rendering with good caustics in Fusion 360
That is actually embarrassingly simple. I got lucky 😉
I created a large flat block and applied the carbon fiber twill texture. Thats it!
The nice anisotropic "streaky" effect comes from the small light sources on the underside of the upper hex frame.
Nope. I use the "cool light" environment in Fusion 360 and turned it down to 1000 lux so the incredibly weak emitters in Fusion 360 can take effect.
