10 times faster 3D printing

Julius Huijnk

5 min readOct 9, 2020

I’d like to share an observation and idea on the speed of 3D printing. Disclaimer: I don’t own own a 3D printer.

The observation

3D printing is like very slow magic. Out of nothing, there is something. Eventually.

For each tiny layer, the printer does something like this:

The idea

It seems like printing can go 10 times as fast if it could have 2 or more different size ‘brushes to paint’ with. Like this:

First print the outline, then the big strokes, and fill in the small details with smaller strokes.

A square would be perfect, but with other shapes it should also save heaps of time.

The max gain is simple to calculate; A 10 times thicker stroke, is 10 times as fast.

Implications

This approach does require new filament that looks more like a ribbon:

A new nozzle. Not just bigger, but wide and flat:

New software:

To me it seems like an iterative improvement on the concept of the 3D printer and very doable.

So, does this makes any sense?

Update

Comments on Reddit over here, all quotes below from these comments.

The main feedback:

It’s difficult

It’s an interesting idea. The software side wouldn’t be a big deal. I do thing getting consistent extrusion from a flat nozzle, regardless of filament shape, would be difficult. Molten plastic is very viscous and doesn’t self level.
Ok, I get what you are saying, but yeah. Not that easy. Once plastic melts it’s pretty hard to control.

I’m reading this as ‘difficult, but not impossible’. Once you figure out how to do it, it’s seems to be not more expensive than the current extrusion solution.

It assumes a (more) solid inside

This also kind of assumes that you want solid prints. 1cm wide nozzle like you propose would be fast when doing the top and bottom layers, but worthless when doing infill, which is still like 90% of the print.

Yes. It could be that I lack knowledge on infill printing, but I’m suggesting the inside would be fully solid, so it would be used for 100% of the print.

That would make it use more filament and heavier. That could make sense for a lot, but not all objects. For some objects you would be able to trade weight and money for speed.

Cooling is the hard part

When it comes to speed, the melting and cooling of plastic is really the limiting factor in desktop printers today. No problem making a print head move around super fast, but cooling the plastic will still take time. That’s why we have “minimum layer time” on small details already. And plastic needs to cool down, no matter what size or shape nozzle you have.

Not sure that changes the fact that you need 10 times less printer movements, so it would still increase speed by as much. This article was written to inspire and learn. If it’s true that (losing) is the real bottleneck, then that’s a great lesson, and brings about the next point made:

Extrusion is a dead end

FDM is most likely a dead end for you, that’s just the physics of plastics.

Did anyone write an article or create a video on a first principles approach to the max theoretical speed of different 3D printing technologies?

One neat trick is to rotate the vat of resin and project a movie, so the full object can be solidify ‘at once’ and not layer by layer. If you project from multiple sides and sync the projections, I would assume you could speed it up even more.