New extruding idea

Im really looking forward to attaching an extruder up to CopperRap, and have looked at the many and varied ways that some of you folks have invented for dealing with the problem.

As I understand it, the fundamental problem is that the plastic changes its properties as it gets hot but before its fluid enough to extrude and sticks to the walls of the extruder which means you have to overcome this resistance by increasing the driving force.

Has anyone considered an alternative approach? Could you have a very much larger chamber that is heated up to a temperature that will melt the extruding plastic and then, when you need to extrude, release some of this plastic from the "tank" of molten 3Dink in a similar way to how an inkjet printer works? You could have some form of valve at the extruding head that opens and closes to release the molten plastic, then use a sensor of some kind to decide that the "tank" of molten plastic needs topping up.

Im sure there are lots of reasons why this wouldnt work, just wanted to throw a new idea into the arena!

Phil

Hi Phil,

... i had to realize a similar setup with hot glue (around 80°C, controlled +/- 0.5°C) and a pneumatic dispenser for a project in my daywork - here the biggest problems were changing viscosities by varying nozzle-temperatures when pausing or dispensing and avoiding cob-webs when cutting off the tray ...

With a volume-dispenser (e.g. stepper-driven syringe) the dispensing is more uniform over time, but you have some problems with ambient temperature-changings too.

I applied three separate heaters along the syringe and nozzle for a better controlable temperature profile and a heated base-plate for defining the ambient temperature for the objects.

For a single object with sheetwise fabbing it's easier to solve than for my problem, where i had to output and handle the same volume/shape of glue for 20 to 60 objects, but it's a bit tricky too

Viktor

Actually, that's kind of what happens in the nophead's extruder (take a look at his blog, [hydraraptor.blogspot.com]), with its relatively wide (5-6mm dia) melt chamber and tapered feed channel - my extruder is a blatant copy of his design, so I was able to evaluate it myself, with very good results.

First of all, a tapered channel effectively eliminates the problem with softened plastic sticking to the extruder walls. Even if it sticks, the downward force exerted by the feeder pulls the material away from the wall, not into it. Secondly, you don't really need a valve - liquids are not compressible. Doctors don't use a valve on a syringe, do they? With some molten plastic in the melt chamber and the filament acting as a piston, you get a very precise control over what happens at the business end of the extruder. It's just the control over the filament itself that is a problem, there's a lot of inertia in most feed systems out there. AFAIK, the most recent of nophead's desings reduced this to a minimum, too.

Enleth - most fluids aren't compressible, but molten plastic actually is to a certain degree (http://en.wikipedia.org/wiki/Non-Newtonian_fluid). It isn't a huge deal at the usual scale of extrusion for a reprap, but if you're looking at pushing out from a large tank/pool of plastic, it can become more of an issue.

by the way,
what about shooting very very little platic balls through a very hot plasma
so they hit the target in molten state :-)

as these little plastic drops would immediatley glue to the target,
one might more easily build overhanging structures.

roland,
physicist ;-)

Hi Roland,

... you can buy a wax-ink-jet-head which is capable of 'shooting' 30 microns big droplets of molten wax that solidify on the target surface.

The price of the complete system should be something around 20.000 Euros.

With plasma-melting plastic-beads you should pay something more, as you have to implement much more really sophisticated stuff (e.g. plasma-heater, 'machine-gun'-feeder, ...).

Better look ahead for solid-wax-printers which should arrive the market next year - this would be much cheaper

Viktor