Melting stuff with a light bulb

[Peepsalot]

I'm wondering if some knowledgeable flashaholics would have some input on this idea...

First some background info:
There's a project I'm interested in called candyfab (http://www.candyfab.org/).
It's basically a DIY 3D printer / rapid prototyping machine that uses sugar as a medium. It builds 3D objects one layer at a time, by melting areas of granulated sugar on a flat surface, then depositing a fresh layer of sugar on top and melting parts of that, etc. With enough layers, you can build a full 3d object.
The reason sugar is used is because it is cheap and has a low melting point. However the concepts developed for this machine may be applied to other materials too, such as plastic or possibly even some metals. Already they have shown it can be effective with some plastics.

Currently the printing method involves blowing a gentle, thin stream of hot air directly at the sugar to melt it.
I have a feeling that using a focused beam of light would be more effective, allowing better resolution and precision when printing these 3d objects.

So, I thought I'd brainstorm on here for how one would go about designing a replacement "print head" that uses a light bulb as a heat source instead.
I think the main goals would be something that is:
-inexpensive
-relatively safe
-ability to focus the beam tightly (~1mm maybe?)
-operating on 120V would probably be best, but not necessarily mandatory
-smaller overall size is probably better
-variable power

I guess my main questions would be:
1) What type of bulb would be good for such an application?
-probably the smaller the better, as long as it has enough power to do the job
-finding the right power level may require some experimentation, hence variable power would be a plus
2) What could be used for a reflector?
-plastic reflectors might melt depending on design
-i'm thinking something widely available and cheap like an aluminum can bottom or similar.
3) What sort of lens(es?) would be needed for the focusing
-need to withstand the heat, and hopefully not too costly

Please don't suggest a laser as that would be too costly and also dangerous at the power levels required for this project.

One other thing which might be worth consideration is if the bulb can handle switching on and off thousands or even millions of times. The existing hot air design is always on, and the printer simply moves quickly enough over non-printing areas that it doesn't melt. So it' not mandatory, but having that extra ability to switch the heat on and off might also give some improvement in output quality.

 

[Burgess]

Quite off-topic, but . . . .


I well remember, in the early 1960's,

making lots of Toy Soldiers by using my new (toy) device which used a regular light bulb to melt plastic.


Then you'd pour the liquid plastic into "molds", and let it cool.

Pop the mold "halfs" apart, and trim the "sprue".


Instant toy soldier ! :twothumbs


Believe it was a 75-watt bulb.


Generated a LOT of heat ! :devil:

_

 

[TigerhawkT3]

If you want a tight hotspot, try an ellipsoidal reflector. I've got one for my HMM, and it works very well. Look for a 120V medical bulb on bulbconnection.com or something (50W and up should work), then find something to mount it in. I'd recommend a Kiu socket. You'd also need some way to hold the reflector at the proper focus point. If you go this route, you might get away with underdriving a bulb to the 0L/W point, so that it creates just heat instead of light and heat.

Alternatively, you could use a laser - maybe one that could be focused to a pinpoint spot, like this one.

 

[James Jackson]

Why not use the heating elements from a toaster? They get plenty hot. You don't need the light - you just need the heat.

The only drawback is that the heating elements are exposed - which can cause some issues if you get your fingers on them.

Or... soldering irons - they have 'focused heat' using their iron tips. They even have the models where you can 'dial' in the heat - more - or less - depending on the application.

Regards,

James Jackson

 

[LightInTheWallet]

Please don't suggest a laser as that would be too costly and also dangerous at the power levels required for this project.

Sounds like a lack of experience/knowledge seems to be a factor in dismissing the PERFECT source you are seeking. The extensive use of reflectors and optics needed to approximate a laser with general lighting techniques would be PROHIBATIVE, IMHO.

 

[TigerhawkT3]

Please don't suggest a laser as that would be too costly and also dangerous at the power levels required for this project.

Sounds like a lack of experience/knowledge seems to be a factor in dismissing the PERFECT source you are seeking. The extensive use of reflectors and optics needed to approximate a laser with general lighting techniques would be PROHIBATIVE, IMHO.

It sounds like you've done this before. Would you care to share exactly what power levels would be required for this project? Have you yourself acquired the necessary experience/knowledge to find this "PERFECT" source? Are you saying that ordinary light sources and lasers would both be too difficult to implement and that the OP should stick to the standard hot air method?

 

[LukeA]

If you want a tight hotspot, try an ellipsoidal reflector. I've got one for my HMM, and it works very well. Look for a 120V medical bulb on bulbconnection.com or something (50W and up should work), then find something to mount it in. I'd recommend a Kiu socket. You'd also need some way to hold the reflector at the proper focus point. If you go this route, you might get away with underdriving a bulb to the 0L/W point, so that it creates just heat instead of light and heat.

The ellipsoidal reflector suggestion is a good one. Then there's a single hot point in 3D space, not a beam of dangerous-to-the-eyes IR radiation.

 

[yuandrew]

Wonder if a 250 watt halogen bulb will work

http://www.lampsplus.com/Products/Light-Bulbs/Type_Halogen/Wattage_175w-@-500w/16366/