A filament extruder for 3D printing
Filament 3D printers depend on a supply that in 2015 was imported. This project designed and built the machine that produces it: a single-screw extruder that takes polymer pellets, melts them and pushes them through a nozzle. The whole machine was made for 2.1 million Colombian pesos and in its first test it extruded filament continuously. The test was stopped early because of a mechanical problem, and the thesis reports no measurements of the filament.
Context
Three processes take place in series inside an extruder. In the feed zone the solid polymer advances by friction against the barrel. In the compression zone it melts, from the heat of the heaters and from shear work. In the metering zone, the molten polymer is pumped toward the nozzle at stable pressure. Screw geometry determines how each one happens.
Design
The 175-page thesis develops the calculation of each zone from classical extrusion theory.
Solids conveying. For each channel depth and pressure, the advance angle of the material and the mass rate are computed. Channel depth in the feed zone was chosen from those curves and from the measured pellet size.
Solids conveying mass rate as a function of channel depth, for different pressure ratios
Melting and pumping. Tadmor's model was used for melting length and the flat-plate flow model for the metering zone.
Structure. Stresses at the screw root and in the barrel, deflection of the cantilevered screw, and bearing selection for the axial load were checked.
Screw model with its three zones
Manufacturing
The screw was machined on a CNC lathe and given a surface heat treatment. The barrel has grooves for the thermistors and a feed throat with the geometry recommended in the literature. The hopper was 3D printed.
Extrusion screw before and after heat treatment
Heating is done with band heaters along the barrel, and a profile with fans cools the feed zone.
Electric band heaters
Operating test
The machine was set up with a gearmotor borrowed from the fluids laboratory.
Pellets in the hopper at the start of the test
Die head and nozzle with material reaching the extrusion point
The material advanced, melted and came out of the nozzle continuously. A few seconds after extrusion it could already be handled. The test was stopped for safety because the barrel began to rotate on its support. On disassembly, the thermistor wires were also found to have burned.
Against the design constraints, cost was met, weight was 19.6 kg against a maximum of 22.5 kg, and length and width exceeded the plan because of the size of the available gearmotor.
What is missing
Filament diameter and its variation were not measured, which is what determines whether it is usable for printing. There is no winding system or puller control either, which is the mechanism by which a filament extruder regulates diameter. The calculations use friction coefficients and viscosities taken from the literature, and the author warns that they may differ from actual values.
The support that failed was redesigned and made within the same project, and glass-coated thermistors were proposed. The corrected machine was not tested again.
How it fits in Robiolab
This is a machine-design project from the group's early stage, within the digital-fabrication line that includes the stereolithography printer. Its relation to robotics is indirect: controlling the printing material opens the possibility of making filaments with tailored properties, for example flexible or conductive, which are of interest for soft robots and sensors.