A couple of little problems are coming to mind, though the concept looks interesting.
The first thing that comes to mind, looking over your screenshots is the design, the way it is shown, has the magnetic cylinders positioned vertically. This would mean that the steel balls fitting inside the cups would be relying on the magnetic forces from the sides, between the poles where they are weaker, and not from the poles themselves where the field is it's strongest. The second thing I see after downloading your model and applying transparency to it is there looks to be space (around 2mm?) between the magnet and the cup for the ball. The pull of the magnetic field falls off almost exponentially with increasing distance. Turning the magnet on its side and allowing it to directly contact the ball would make a lot of difference in how secure the connection would be. I would think it would cut down on the mass a bit. Same number of balls, half the magnets.
Mark
The first thing that comes to mind, looking over your screenshots is the design, the way it is shown, has the magnetic cylinders positioned vertically. This would mean that the steel balls fitting inside the cups would be relying on the magnetic forces from the sides, between the poles where they are weaker, and not from the poles themselves where the field is it's strongest. The second thing I see after downloading your model and applying transparency to it is there looks to be space (around 2mm?) between the magnet and the cup for the ball. The pull of the magnetic field falls off almost exponentially with increasing distance. Turning the magnet on its side and allowing it to directly contact the ball would make a lot of difference in how secure the connection would be. I would think it would cut down on the mass a bit. Same number of balls, half the magnets.
Mark