Powder support

There are no problems with suport material in powder fabbers - unused powder works like a support. Why we can't use this? We don't really need solid support, just a box to prevent powder escape.

Each layer is printed in plastic and then all holes are filled by powder (salt for examplel). Batcher for powder is simple, as far as I understand: three disks at one axis, first and third fixed, middle rotating. All three has same-diameter holes, middle disk hole between first and last disks work like a box for one portion of powder. Maybe some vibrator needed to prevent powder stack, maybe not.

I don't think powder can work as a support material for plastic.
At the start of a new layer, when the new thread of plastic from the extruder comes out, it needs to stick to the old layer. Since the plastic probably won't stick to the powder, it will just trail around behind the extruder and form a cobweb.

Powder would probably work as a support material for slurries.

I believe he's proposing a powder/sintering system, as a counter suggestion to the current method of using extruded plastic.

Actually, the suggestion is to use plastic and fill in with a layer of powder or crystals. Sebastien does have a point, unless the plastic touches other plastic, it will slide as the filler will act like a greased surface. You'd have to almost use some fast/safe solvent-solute base system to melt away. Dissolved sugar is an example, but it'd take too long to resolidify and would leave a mess. Same for common table salt, plus the corrosion. Maybe something that dissolves in alcohol?

It IS an idea. We've hashed over this somewhat before. Mechanically, the biggest problem is getting a good seal or mechanically sound motion platform in the powder to raise and lower it. Heck, I could almost see using glass flock and 2cps cyanoacrylate glue (this is about 2.5x thicker than water). This thickness is supposed to dry within seconds, so it may work well to make a decent build material. Shear strength of 3200psi isn't too bad, either.

[www.devcon.com]

... acetone is very low viscose and is drying very quickly, so plastics solved in acetone could go (maybe with an extra drying idle time per slice).

Maybe sand as body-powder and ABS-acetone-solvent would be a good try ...

Another idea: maybe it's possible to 'print' droplets with a dispenser instead of an ink-jet-head, when the dispensed fluid is outputted above the powder-surface in discrete steps for a spherical droplet and then lowered to the powder to wet it in a defined half-sphere.

The speed wouldn't be very fast and the size of the droplets shouldnt be much smaller then 150 microns in diameter, so the wetted area should be 300 microns or a bit more in diameter, but for a test it's easy to setup ...

Another (more complex) try would be an 'pulsing' dispenser, which shoot small droplets onto the surface - maybe with a heater in the capillare for a kind of 'bubble-jet', or some other approach.

One of my more sophisticated ideas is to mix a thermo-curing epoxy with carbonyl-ferrum-microspheres (i have grades from 10 microns down to 1 micron in size from BASF and a house-neighbour, which is the leader of a group developing nanoparticles at Degussa).

Then i have only to apply a hot-spot with an inductive hf-microantenna on a discrete point to hardening the epoxy (similar as curing with an uv-laser).

Then a setup would work with a liquid-base as in UV-curing-Epoxy-STL-printers, but instead of an expensive UV-laser i can use a 'printhead' with one or an array of needle-pairs of hf-antennas, which heat the micron-sice area between them through oszillating of the ferrum-microspheres for instant-hardening the epoxy ...

Viktor

PS: it should be possible to use this hf-induced 'instant-hardening' of filled epoxy with a normal fluid-dipensing syringe and micro-antennas at the tip too!

So i would use a filled epoxy at room-temp, and with dispensing and heating at the output the reprap would lay a hardened/hardening tray ...

Edited 2 time(s). Last edit at 10/02/2007 02:23AM by Viktor Dirks.

Another approach I have seen kicked around on various forums is to lay down succesive layers of a low melting point plastic (usually powder coat material(a fine grain polyester)) across the whole print table. The write head would be a laser scavenged from a DVD burner. Obviously. as the head moves between succesive the laser fires selectively to fuse the powder. The layers would be thin, but the laser could probably move pretty quickly. Resolution would be limited by the size of the powder grains rather than the viscosity of the melted plastic, and multiple colors are readily available. Thank heavens for the quick change tool holder.

Brian D

i like that approach. i think i'm going to have to try it once i get my cartesian bot up and running well.

One problem that you are going to have trying to use powder as a support material with an FDM extruder is that if the paths for extrusion over the powder aren't straight then the viscosity of the extruding plastic is going to mean that the thread of extruded polymer is going to get dragged around over the powder as the extruder orifice changes direction. This will happen unless you design your object to make sure that powder is always traversed by a straight-line extrusion. I just don't see that as being practical.

If you use something like a laser, however, to melt a plastic powder in situ, however, it should work perfectly.

Laser still sounds awesome. The first time I ran into RP was on the history channel - modern marvels, a few years back, with a STL system. It looked like a laser flashing on a vat of liquid, and thirty seconds later a smooth, clear plastic model of a space shuttle ready for wind tunnel tests popped out. Of course, it was probably time lapse, but I will never forget that film clip. Unless I eventually get a reprap that does one better: make something I can actually touch!

The only problem with powder as support material that I can see is: if it's fast, it won't be precise, and you'll fill holes you didn't want filled. Hollow parts, mainly. Of course, you may end up wanting support material for those holes too, but then your once-hollow honeycomb style model is now filled with powder. Maybe we just need to work on getting the HDPE to hold itself up. Wasn't there a photo on one of the blogs somewhere of a plastic extrusion going nearly vertical, with loops?

-Samuel

Hi Samuel,

... if you want a high accuracy, e.g. 0,1 mm per slice, then you have to process 10 slices per millimeter or 100 slices per centimeter.

With an even as big accuracy and activating/melting every pixel of the body it can last many minutes per slice, so a 'normal' part like a gear with some centimeters in square can consume hours for building, dependant of the 'pixel-processing' speed ...

For this reaseon i'm trying a LOM-approach, where i would layer sheet over sheet of material (for example a 0,1mm thick self-adhesive plastic-sheet from a roll) on a bed (or a roll, as in lathe) and draw only the contour-lines and disassembly-squares with a diode-laser, cutting through the sheet, so the actual slice would be ready much faster.

A very nice method is the selective-inhibition-printing, as mentioned in a previous post, where an ink-jet prints a liquid on the powder-surface drawing only the contours and separating lines, and after every slice an ir-heater sinter/melts the complete powder-slice at once ...

Actually i'm waiting for some hardware to proceed, but in the mean-time i'm playing around with a 1-Watt-diodelaser and with dispensing of fast hardening pastes at room temp (maybe with forced heat-hardening or sintering with the diode-laser too).

When in the future the ink-jet-tests would work, then the SIS-method (Selective Inhibition ...) would be worth a try too

Viktor

There was a company which marketed such a 3D printer over ten years ago. They went out of business.

Hi Forrest,

... yes, some companies use paper-LOM until now for lost-casting-forms or large-scaling prototypes of combustion-engines, very big hardware or so ...

In 1993-94 i designed and build two 2D-CO2-Laserplotters (with a 5-Watt-RF-CO2-laser) for marking and cutting of plastic-sheets -- for the possibility of 3D-processing or marking of very high parts i made it so, that the base-plate could be separated and then i have a 'bottom-less' working area, for inserting an elevator under the plotter-basis ...

When we first presented our concept with an estimated price of 7500 to 10000 (with the Z-elevator) Deutschmarks on a fair, then a better designed 2D-laser-engraver from an US-company on a neighbour-booth, with a price of $30000 (then maybe 70000 Deutschmarks or so), forced my boss to stop the further developing, because he thought, the 'better-shaped' and older/riper competitor 'wouldn' give him a good start' ...

The Idea of 3D-LOM-printing with the same apparatus wasn't so spectacular for him anymore ...

Two or three years later i saw the first 3D-LOM-Printers with 50W-CO2-lasers on fairs in action, but then we were working in a complete different area.

Viktor