3 Fundamentals of High-Speed, High-Quality 3D Printing

If you’ve ever used a glue gun you know the importance of applying just the right amount of force to the glue stick — too much and the glue comes out too cold or starts to ooze out of the back; too little and the glue overheats and loses its binding properties. As you can see, there is a very subtle balance of force and heat that determines the efficacy of your last line of glue. Much like the old English tale ‘Goldilocks and the three bears’ — finding the correct balance of temperature requires some trial and error.

Extrusion-based 3D printers work on the same principle and suffer from the same problems. However, a fully autonomous 3D printer does not have the luxury of trial and error — the balance of force and temperature must be precisely achieved first time for a range of materials, environmental conditions and speeds.

A significant amount of ongoing analysis and design has been put in to ensuring the AiMaker achieves the ‘Goldilocks zone’ of force and temperature for the full range of operating parameters. Our analysis has ranged from basic heat transfer theory to advanced computational fluid dynamics (CFD) software — all looking to achieve the perfect hot-end design.

In practice, achieving the Goldilocks zone for set operating parameters is relatively easy. The tricky part is creating a design which performs under a wide range of operating conditions. Our approach focuses on three main ideas.

The temperature in the AiMaker hot-end depends on the heat output from the heater cartridges, however the relationship between temperature and heat output is heavily influenced by extrusion speed, extrusion material and cooler fan speed. The perfect temperature is achieved in the AiMaker through an AI-powered feedback loop. Information on the temperature and other influencing factors is fed in to the AI algorithm, which in turn subtly controls the heater cartridge output to maintain the ideal temperature. The graph below shows the reduction in large temperature deviations that the temperature stabilising AI algorithm has helped us achieve, giving greater stability than conventional deterministic PID algorithms.

Additionally, multiple sets of heaters and thermistors down the length of the hot-end allow for control of the temperature profile as the filament moves through the printer, adding another layer of control.

Solid material is pushed in to the AiMaker hot-end and quickly melts within the internal channels. Minimising the natural resistance to flow of these internal channels brings a marked improvement in printing performance. Flow resistance in the channels is caused by three main factors which can be best understood by envisaging a tightly packed corridor of people.

The first factor is friction — as people bump in to the walls of the corridor and each other, their progress is slowed. The friction in our hot-end channels is minimised by smoothing the walls and minimising the amount of wall the fluid (or people) is in contact with.

Resistance to flow is also caused by the narrowing of the channels towards the nozzle tip. As the corridor narrows, people bump in to one another more frequently and also have to file behind each other, significantly slowing progress. Our hot-end channels are balanced to find the optimum width in order to minimise this effect, while still narrowing to the desired nozzle tip diameter.

Finally, the profile of the corridor plays a role. A sharp 90 degree bend in the corridor will result in further congestion as people slow down as they approach the bend and file behind each other to take the inside route. Again, the profile of our hot-end channels is optimised to minimise this effect.

In a seriously slick hot-end, the Goldilocks zone spans a greater range — more force can be applied at a given temperature before material forces its way out of the back. CFD proves the ultimate tool for this task, allowing simulation of flow resistance for a range of speeds, temperatures and channel profiles.

With temperatures of 200°C and more, the AiMaker hot-end is no pleasant place for filament to be — keeping the hot-end as short as possible allows for a swift entrance and exit. Material degrades at higher temperatures so reducing the time spent in the hot-end dramatically improves the print quality. Additionally, reducing length minimises the volume of entrapped material, leading to a reduction in oozing from the nozzle tip caused by thermal expansion as well as gravity. And finally, a shorter hot-end results in lower resistance to flow — less wall for people to bump in to.

And there you have it — our three fundamentals of hot-end design.

These ideas have driven the design iterations of the AiMaker’s hot-end and have helped the Ai Build team in unlocking the potential of extrusion-based 3D printing for large scale manufacture.