Tuesday, July 29, 2008


Making CAPA Filament

I've been having trouble finding a CAPA or PLA filament supply, so today I did some experiments today trying to make CAPA filament from granules.

I created this:

It's basically the high pressure screw driven syringe with changeable nozzles. A nut is soldered into one end of a piece of plumbing pipe, and a pipe connector to the other. The wood contains a captive nut, a washer is soldered to the threaded rod to form a plunger, and the handle at the top is turned to force the plunger down.

The pipe is then dipped in boiling water for a few minuites so the CAPA melts, then the handle can be turned and this happens:

The nozzle I'm using has a 3mm hole in it, which created a filament of 3.4-3.5mm (pretty good consistency). I'm going to try making a nozzle between 2mm and 2.5mm tomorrow and see if I can get a 3mm filament.

See more pics on flickr


Keep in mind that extruded diameter is probably a function of extrusion speed as well, so keeping that constant may give you more consistent results.
I had thought that could be quite a problem, but the i have to turn the handle quite a bit to get it to extrude so keeping the pressure constant shouldn't be too hard by hand, but i'm thinking about motor possibilities, connecting to a cordless drill is a possibility.
Nice job. How much capacity is there in that extruder pump?
Probably around 8ml, but only about a 1/4 of this much plastic can fit in as granules.

I'm considering:

1. a longer barrel.

May have problem with friction

2. motorised piston going back and forth, with a hold in the side to feed in plastic from the hopper.

Getting a bit complex

Switching to an electric heater could work too, as my hot water method only gets the plastic to < 100^C, at which point the CAPA is still very thick
Very nice!!! :)

I wonder if we could make a continuous process out of this somehow...

Perhaps some sort of centrifugal pump (http://en.wikipedia.org/wiki/Centrifugal_pump) with a hot impeller? The granules would be slowly droped (manually or automatically) in the center of the impeller, would slowly melt, and then be pumped out through a hole. We'd have to find a way to keep friction down between the molten plastic and any stationary sides of the pump, but maybe it could be done?
Would a motorized piston going back and forth trap air in the extrusion?
I'd imagine it already does trap air. We really do need something like this for the extruder, though; going through multiple stages like this is redundant and, at this scale, hugely inefficient. The centrifugal pump idea is neat, but I don't think we could easily keep the friction between the outer walls and half-melted plastic down low enough, and you'd probably get granules mixed in with your output, which would clog the extruder. Instead, I think we could do something with a screw inside a heated barrel leading directly to an extrusion head, somewhat thermally isolated from its hopper to prevent half-melted plastic from clogging the pellet intake. A similar design has been tested without the thermal isolation, and worked until it clogged.
Great to see some 'real' work being done on a filament extruder

One issue that is raised whenever someone has ideas for an extruder is air bubbles becoming trapped, how 'bad' (or good) has this been with your design?
Post a Comment

Links to this post:

Create a Link

<< Home

This page is powered by Blogger. Isn't yours?

Subscribe to
Posts [Atom]