Plastic-on-plastic non-adhesion for pre-assembly

Another line of thought that seems worth exploring is the idea of 3D creation of assembled plastic parts that are not bonded together.

Since it is all made with hot-melt plastic, the question is how to separate layers so that they don't bond. Two options come to mind:
- powdered graphite
- powdered talc

The process would be to apply a coating of powder between surfaces that should not bond. The powder could be mixed with a solvent that assists in application, and rapidly evaporates after deposition.

The coating does not need to be thick, just enough to coat the surfaces of layers already applied. The coating would need to apply to side surfaces as well as top surfaces, so probably a tiny 0.5 mm conformal foam brush would be the way to go.

,

Heh, imagine a tiny bucket on an X-Y carriage, plumbed to a larger tank so that the liquid in the bucket is constantly recirculated as the solvent evaporates.

A lever and cam periodically dips the tiny brush tip and wipes off excess, then the X-Y carriage applies the coating to the part for a short span, and re-dips the brush again.

When not applying the coating, the brush stays in the bucket, so it does not dry out, and could descend slightly further with a gasket around the upper part of the brush, that seals the bucket opening to prevent unnecessary loss of solvent when not needed.

,

So, this would allow for gears and levers pre-formed within a flat assembly, but doesn't provide for a way to temporarily fill voids for bridging or covering, unless the plastic itself were deposited in a way that void-fillings can be removed after the part-creation is completed..

Edited 1 time(s). Last edit at 02/17/2011 02:29AM by Dale Mahalko.

without any adhesion, i think it would be hard to actually trace out a shape. straight lines might be ok, but the corners will probably need some anchor.

I'm thinking that it might be possible, but you would need to alter they way gcode is produced. the filament could be laid down as if it were radiating from a single point, which would be the anchor point. the first string would be the trickiest, but subsequent strings could at least adhere to the previous string. The perimeter would have to be traced last.

The problem is that if something goes wrong, it will go wrong halfway through the build (or at the worst possible moment, according to Murphy's law :-)

it might be better to paint on something that will just reduce adhesion, and not prevent it. maybe elmers glue or a paint, adulterated so it dries quickly, like maybe with alcohol?