A new math puzzle, yay!

[Philippe Hurbain]

I'm like 80% complete with this model and this is just the beginning. If you manage to tackle all of this my hat's off to you.

I'm afraid that like all these kind of assemblies this one relies on some part flexibility...
I used a pragmatic approach to build the assembly attached:

• Since the system is symmetric, I rotated the beam assemblies on the angle connector so that the balls are 200 ldu apart. I didn't even tried to get an exact solution (that should be feasible) but used trial and error.
  
• I then snapped the assembly in the ball joints, and rotated the whole assembly at ball point around Z direction till the yellow axle joiner sits in the top fork.
  

As you can see the joiner lies pretty far away from fork rear wall...