Molten Glass

Has anyone tried to print using molten glass as a material?

Hi, I'm new to reprap. I'm a software developer, and I've been seeing reprap around the net for years, but I only recently became interested in parts fabrication. I'll be building a prusa next month. I've done a lot of research. ABS, PLA, PVA, Polycarbonate, Frostruder, Polymer clays, glass powder techniques, etc. I'll of course be using plastic for my first reprap.

However, my girlfriend and I do a fair bit of lampwork. I have a couple of stores, but this one will probably amuse this crowd the most (it's the one I enjoy most too): [www.etsy.com]

For those of you unfamiliar with lampwork (as I was until about a year ago), it starts with a torch. I started with a cheap MAP gas torch, but I now have a couple of nice propane/oxygen torches. Glass is a bit of an insulator, so you grab these thin (think, pencil diameter) glass rods in your bare hand and you slowly heat up one end. Once it is molten, you can do whatever you want with it. Traditional lampwork is mostly about bead making. In recent years, many artists have gotten into glass pipe making for the underground marijuana smoking market. The possibilities are endless. These days, "glass blowing" and "lampwork" cross over so much that the line between the two disciplines is mostly a question of scale and equipment. For example, most glass pipes are "blown" using a torch, but many people still "blow" cups and lamps etc etc using larger glass blowing furnaces. Ok, if that sounds interesting, watch some YouTube videos.

Back to reprap. Has anyone considered using molten glass as a material for the additive process? It's certainly more expensive than plastic, but I'm certain it would be an ideal material in many cases. For example, food containers, chemical containers, jewelry, science, art, etc.

Glass gives off zero fumes when burned. The big safety concern is heat and eye protection.

I'm not sure if electrical system are practical for heating glass rod, but there are always torches. There are many different kinds of glass, mostly identified by their COE (coefficient of expansion). Probably borosilicate is the best choice for additive manufacturing, because it is much less likely to crack when heated and cooled. You guys are already using boro for your build plates, so you probably know that.

Any thoughts?

It is an interesting idea, my partner makes fused glass jewellery and the concept of using 3D printing would interest her.

However, as glass melts at way above 600 degrees you may have problems building a machine that extrudes at these temperatures. Check out some of the 'molten metal' threads for some ideas.

G

---

---

How often I found where I should be going only by setting out for somewhere else. - R. Buckminster Fuller

[apapageek.com]

Right. It's definitely an engineering problem, but it's not impossible. The kiln your friend uses to fuse glass is almost certainly electric.

Also, I've read quite a bit about welding. One of the problems with molten metal is that it reacts with oxygen, ruining the weld. This is why the TIG process uses Argon as a shielding gas to keep the molten metal away from our atmosphere until it cools sufficiently. All welding processes do this in some way.

Glass has no such problem. It does not require a shielding gas. Also, the melting point of glass is significantly lower than that of metal, and it is not a conductor, so it requires significantly less energy to melt.

It'd be very cool, but there are an awful lot of problems to overcome:

You can't spool glass, so you need rods and a loading mechanism.
It's difficult to grip, a hobbed bolt would slide along it.
It's very sticky, you'd have to try and scrape it off the hotend to clean it.
It's incredibly viscous, so forcing out of a nozzle would be tough
You couldn't use 3d printed plastic parts anywhere near it, or they'd melt.

These are probably all possible to overcome, but I suspect a problem is that you need to overcome all of these things to be at all useful - they're not really applicable to plastic extruding. Even the multi colour print heads are taking a long time to be adopted, and one of their main problems (nozzles dribbling) is an annoying problem with normal printing and that hasn't been fully sorted. Casting into an investment casted mould might be an easier way to go.

Well, in the short term, yes, casting is easier. However, in the long term, consider the different kinds of glass you could use in one of those clever loading mechanisms:

• dichroic (glass infused with atomized flecks of metal: think, space helmet glass)
• all possible colors (glass is very colorful)
• all possible opacities (again, as varied as the imagination)

Now, not only can you print with two colors, but you can print with a rainbow of colors (insert picture of a unicorn throwing up), and you can cover them in transparency. It's a little mind blowing what you can accomplish, visually.

Of course, now that I think about it, we can do all of this today with glass fusing techniques. You need a kiln, a good mould, a little glue, and a pair of tweezers.

We would mostly be buying ourselves reproducibility, accuracy, and a significant reduction in labor time as compared to the complexity of the piece. Right now labor is at least O(N) for the complexity of an engineered fused glass piece. It would be more of a constant with an automated system.

I'm pretty sure the piece would still need to be annealed in a kiln after printing, but that's probably acceptable for the time savings in assembly.

There are already CNC machine that will cut glass for you, so you can get the accuracy, it's just down to how you wield the tweezers

And of course you can't extrude Dichroic because it a flat piece of glass with a pattern attached, if you extruded it you would probably loose the effect.

G

---

---

How often I found where I should be going only by setting out for somewhere else. - R. Buckminster Fuller

[apapageek.com]

Well, I guess you could come at it from another angle: Cut via CNC, assemble via suction grippers.

I don't think that's how dichroic works, btw: [www.sundanceglass.com]

The problem with the CNC // Suction technique is that there is probably a lower limit to resolution. I'm sure CNC machines can only cut a cube so small.

The interesting thing about glass is that you can make "stringers" or "cane" very, very fine in diameter, with great ease. You just heat it up in the middle and pull from opposite ends. The viscosity of the glass does the rest. Check out this video for a quick example. You really have to see it to understand: [www.youtube.com]

This stringer making process could probably be automated and made more scientific/accurate with linear servos and grippers.

... think about glass-fibers, feeded into the spot of a laser, that melts the tip of the fiber, so it's adding to the surface/object ... and then move the spot with synchronous feeding the fiber

---

Viktor
--------
Aufruf zum Projekt "Müll-freie Meere" - [reprap.org] -- Deutsche Facebook-Gruppe - [www.facebook.com]

Call for the project "garbage-free seas" - [reprap.org]

Have you seen the solar sinter? Sort of similar: [www.markuskayser.com]

A really big problem with glass is its expansion and contraction rate. If you think ABS has warping problems, you're not ready to even think about glass. If it doesn't cool evenly, it cracks, and if you don't anneal it properly it can shatter unexpectedly at a later time even if it doesn't crack right away.

Well, borosilicate is much more friendly in that regard than soda lime. Anytime you see a glass artist making a fairy by crafting a limb, then "tack welding" the limb to the body, you're seeing boro. Soda lime cracks if you look at it wrong.

Of course, a heated build chamber would probably help (for ABS too). I've read there are patents on this idea, which is rather ridiculous, IMO. Kilns are nothing more than über powerful/accurate ovens. And a reprap with a heated build chamber is nothing more than a kiln with a robot arm inside.

Having said all of that, I'd love to try it. I'm not sure if boro would explode when used additively or not. It would be a fascinating experiment.

I don't think building something like this will ever be practical.

I would attempt SLS with glass powder before trying FDM. Both processes will trap a lot of air so your transparent sections won't really be that great.

You will need to work above the strain point which means that the heated chamber must be 800f+.

If you really wanted FDM then the nozzle would have to be graphite (just like the extrusion dies they use to make the rod to begin with). Regardless, such a machine would be horrifically expensive to build and require frequent replacement parts due to the heat.

If you really want to "print" in glass, just get the parts in plastic and cast using pate-a-vere. There are literally dozens of methods to work glass that are cheaper, easier, faster and will produce higher quality parts.

"Hemp necklace is knotted by small children in a southern Tennessee sweat shop"

Pure awesome!!

Yes, a heated build chamber at that temperature would certainly be a challenge.

Maybe a graphite nozzle would be practical. I know a few torches use a water cooled torch nozzle, so that might be another option. Cold metal won't stick to hot glass. Only hot metal sticks to hot glass.

Also, I'm not sure if glass rods are made using extrusion nozzles. I always thought they were still made the old fashioned way, out of cane: [www.youtube.com]

But maybe the world has moved on.

What if you just heated the build platform? Would thermal conduction work, at least for thin objects, as it does for ABS plastic? If this is the case, some people might find this useful. You could always take wafers of glass made in this manner and stack them in a kiln for fusing. If we had multiple colors available and high resolution, this would be an appealing way to build pieces. I'm thinking of those crazy paperweights people make with crazy details on the inside.

Oh, air bubbles... yes, that might be a problem. Not sure. I think it depends on whether the droplets are concave or convex.

... I used glass powder for sealing sensors and 'glueing' glas windows in kovar-housings - this powder melts first time at around 350 degC, but when solidified to glass, it needs around 750degC for remelting.

The normal use is in a kiln heated to 400degC, but I've used it too melting it with a H2O2-torch or with a IR-laser.

What's possible with powder - mix it with e.g. Dexpanthenol and Aqua dest. to a paste, dispense it in tracks as normal, let it settle, then heat the topmost surface with an IR-heater, so the Dexpanthenol evaporates and the remaing dust tracks melts/fuses to solid ...

---

Viktor
--------
Aufruf zum Projekt "Müll-freie Meere" - [reprap.org] -- Deutsche Facebook-Gruppe - [www.facebook.com]

Call for the project "garbage-free seas" - [reprap.org]

It's been a while since I started this thread. I check in on things like this every now and then. Looks like someone at MIT made a rudimentary version of this 3d glass printing idea: [www.ll.mit.edu]

[www.ll.mit.edu]
[www.ll.mit.edu]

It looks like the extrusion is too fluid to be accurate, but it does a good job proving the concept, while illustrating the technical challenges to overcome.

Looks like he just took the top off of a kiln and added an X/Y table up there and let gravity do the rest. It's a very smart way to demo the concept. The results are low quality, but it side steps the engineering challenges associated with the viscosity of fluid glass. All he needs is to figure out is a more accurate nozzle and he's got viable 3d glass printer.

I think he just needs a heated graphite nozzle and a z axis.

Here's a DIY inexpensive kiln: [www.instructables.com]
Use that design for the heated build chamber. Fairly inexpensive. The benefit is that you can make a very small build chamber if you're only interested in small pieces. You can build a small inspection/build removal door into the front and have an open top for the insulated X/Y platform. Otherwise, you could buy a commercial kiln and use that for the build chamber.

Edited 2 time(s). Last edit at 06/08/2015 10:03AM by Trevarthan.

Are optical fibers for communications made using extrusion? It might be worth a look at those machines to get an idea about how to handle the input material.

Quote
the_digital_dentist
Are optical fibers for communications made using extrusion? It might be worth a look at those machines to get an idea about how to handle the input material.

Good question. Watching the video from How It's Made now to find the answer: [youtu.be]

Quote
Trevarthan

Quote
the_digital_dentist
Are optical fibers for communications made using extrusion? It might be worth a look at those machines to get an idea about how to handle the input material.

Good question. Watching the video from How It's Made now to find the answer: [youtu.be]

Looks like the answer is "no, they're gravity extruded using temperature to control the diameter". But that's how plastic filament is made too, so that doesn't necessarily mean it CAN'T be forced through an extrusion nozzle.

I'm super curious about this idea, so I did a bit of research through my lampworking books. Here's what I found:

• Soda lime glass is relatively unmoving at 850 - 1050F (454 - 565C) and doesn't have a glow. This would be the ideal build chamber temperature, as the glass does not flow and is not at risk to thermal shock.
• The exterior of soda lime can be manipulated at 1200 - 1400F (648 - 760C) and the glass glows a dull red. This would be my first guess at an ideal target temperature to bond glass from the nozzle to glass on the build plate.
• Soda lime glass really starts flowing at 1500 - 1700F (815 - 926C) and the glass has a cherry red color. This is the temp most lampworkers use to shape beads, because while the glass moves, it moves slowly. This is also probably the ideal temperature range for the nozzle.
• At 1700 - 1900F (926 - 1037C) soda lime has a yellow to yellow-white color and flows more quickly. This is usually uncontrollable for lampworkers.

The problem I see for graphite as a nozzle material is that, according to wikipedia, it begins to oxidize in oxygen rich environments at about 1292F ( 700C ) and up. I assume this means some sort of inert shielding gas, like argon, would be necessary to avoid oxidation. I'm not aware of any problems with that. Just pump argon into the build chamber. It doesn't even have to be pressurized or air tight. It just has to push the oxygen out. That's how my TIG welding rig works. It creates a little bubble of argon around the nozzle.

So, does graphite offer any real advantages here? I don't know. I can't think of any. I'd rather use stainless steel. According to Google:

Quote
Stainless steels have good strength and good resistance to corrosion and oxidation at elevated temperatures. Stainless steels are used at temperatures up to 1700° F for 304 and 316 and up to 2000 F for the high temperature stainless grade 309(S) and up to 2100° F for 310(S).

With a 316 or 309 stainless, we might be able to avoid a shielding gas altogether.

Ehh... this is do-able, but maybe not as cheaply as plastic.
I've been looking at doing some glass work for a custom hotend that uses glass in place of a fluoroplastic liner, and the working temperatures of glass can be greatly manipulated.

Technically, the glass transition temperature can be manipulated to the point that it flows at room temperature, but the glass is extremely brittle, reacts with water in the open air, and probably not something that people looking for 'glass' think about when they think 'glass'.

As far as the nozzle, the best material would probably be iron with a coating of waterglass dabbed onto the surface.

The water glass (whether it's potassium silicate or sodium silicate is un-important in this application) has an extremely low melt temperature and forms a protective layer of glass over the iron, so long as the nozzle is heated up semi-regularly to prevent water from reacting with the cation in the silicate and causing oxidation.

If you want somewhere to draw inspiration from, glass is cast in steel molds.

Just some notes: borosilicates are probably a good idea to start because they have very low coefficients of thermal expansion. In theory, borosilicate should never need a heated bed, but having an excessive heat difference can cause thermal shock still. Some amount of pre-heating with a blow torch, prior to material deposition would be necessary.

As far as what it could be used for, glass is largely superior to plastic in just about every way except cost and weight... and even cost is iffy considering the true easy with which it can be made.