True3DGlass: A Revolution in Glass Processing | Hacker News Reader

True3DGlass: A Revolution in Glass Processing

pkm.kit.edu

53 pointswallflower14 comments

The process depicted in the video is a variant of Stereolithography a process by which photopolymer is solidified by exposure to directed UV light. The video depicts the authors using a commercial 3D printer created by Digital Wax Systems. Reading the paper, it appears that they mixed in amorphous silica to the resin and ran the printer successfully. Few details are given on their test rig.

This approach has been applied successfully to create ceramics in the past, and doesn't seem particularly novel to me. Companies such as ADMATEK are already commercializing this technology.

Weaknesses of this approach are that the end part needs high temperature sintering to remove the photopolymer from the part, and will not yield the same material properties as monolithic glass. Significant shrinkage can be expected and the sintering process means that you are limited in the thickness of the parts that can be designed.

All in all, this is a nice note, but isn't fundamentally introducing a revolution in the 3D printing space.

I disagree with your conclusion.

Post-process sintering is showing advantages across several materials. You'll see that Markforged and Desktop metal have introduced integrated processes to control shrinkage in the final parts. You mentioned ADMATEK as well.

The ability to prototype/produce glass parts within an office environment without needing safety equipment is a major improvement!

It's not the first.. . Neri Oxman's group at MIT did it in 2015: https://www.dezeen.com/2015/08/26/neri-oxman-3d-printing-tra...

This paper cites that one in its first paragraph:

"Until now, only two ways to shape glass in a 3D printing process have been demonstrated: a fused deposition moulding approach in which soda lime glass is heated to around 1,000 °C and a manual wire feeding approach in which a glass filament is melted using a laser beam [6,7,8]. In both cases, coarse structures with high surface roughness are obtained."

The paper you mention is reference number 6. This new paper apparently significantly improves upon resolution and quality. The press release is a bit confusingly worded, referring to "extremely small and complex structures" made of glass. Those haven't been 3D-printed before.

The paper (probably requires a Nature subscription): http://www.nature.com/nature/journal/v544/n7650/abs/nature22...

From their press release:

Three-dimensional printing allows extremely small and complex structures to be made even in small series. A method developed at the KIT for the first time allows also glass to be used for this technique.

Looks like their PR isn't in sync :)

This looks like 2.5D actually.

The base material is a _liquid_ mix of plastic and glass that is first shaped and then, separately, baked. So I can't see how this can be used to print, for example, a hollow sphere.

It looks like it's a powdered mix, though.

Can you bake two hollow hemispheres together?

It says it can be pre-cured under light. I was imagining that a uv lamp on the printer would harden it like dry clay prior to firing.

Just because you can find a specific example where a solution doesn't work doesn't mean the solution has no merit.

A fully hollow sphere may not be possible, but support structures can enable even more complex parts than hollow spheres.

It's like the trick for 3D printing bronze - print with a powder containing the wanted material, then burn out the carrier material. That's how Shapeways prints metals and porcelain.

They're also calling it "Liquid Glass".

Reminds me of Liquid Metal:

• http://liquidmetal.com/properties/the-science/

• http://liquidmetal.com/properties/properties/

Wonder if they'd be useful in combination with each other?

extruding glass as in the MIT project is very different then from printing material that is then thrown into an oven. the resulting piece has really different properties I imgagine

I wonder if this process is prohibitively expensive for just using moulds.

Reply on news.ycombinator.com