Idea for DIY PolyJet printing

Hello forum!

I recently came up with a crazy idea for a DIY form of PolyJet printing, and I hope that we could all collaborate on making it a reality.

As you know, PolyJet printing uses an inkjet printhead to lay down the layers, and then cures them with UV light, in a similar fashion to our SLA machines. However, the "ink" they use is very viscous, as to allow it to pass through the small channels on the print head. In addition, DIY software for controlling inkjet printheads is still in its infancy, and only one project as far as I know, has done it with success, and that is the InkShield. However, this can only control one printhead (color), and only one type of printhead. In addition, it does not take into account motion control, therefore, a motion control system must be developed for each application, and there seems to be more tweaking to be done to get it to the same resolution as a store bought inkjet printer.

So that being said, here is my idea, bear with me, as this may be confusingly written mumbo jumbo.

1. The idea is to take an inkjet printer (probably an epson, as they use the piezoelectric transducer type printhead, which seems to be able to manage droplet size better than the other brands).
2.Gut it, but keep all of the control electronics, and the x axis intact.
3. Take the paper feeder motor, and convert it into a y axis on a reprap style, moving bed system.
4. add a z axis
5. write code to execute a "carriage return" function on the y axis before printing the next layer.

I know that there would be some difficulty implementing this system, but I think the possibilities are very much there. For instance, you can buy a large format Epson printer, that has a print size of 13x19, and the Z is only limited by the shrinkage value of your resin. This could make making large polyjet machines a relatively attainable goal.

Some inherent problems in this idea are as follows

1. making the resin viscous enough to pass through the print heads. I assume a DIY formula would need to be made, as I am pretty sure all currently commercially available resins for SLA printing are not viscous enough. Potentially adding a thinner would allow the resin to be viscous enough, or alternatively, I know of a resin that is 50cps, that could potentially be converted into a UV curable resin.
2. writing the software for the carriage return. In my head, I see it as a traditional stepper motor configuration that runs alongside the encoded DC motor from the paper feed mechanism. After the print job finishes, there would be a a pause, that allows the Y axis to return to the starting position for the next layer, and allows the z axis to raise.


There already is a great piece of software, that I'm sure many of you are familiar with, called Mandelbulb 3d, that can take a 3d model, and convert it into voxel slices. These voxel slices are outputted to a folder containing many images. The entire folder of images could be sent to the printer at once (and I'm sure someone other than me could write a script to accomplish this). In addition, breakaway supports could be generated first in mesh mixer, and then taken into Mandelbulb before slicing. However, another cool application could be dying the resin in a CMYK arrangement, and using a continuous feed mechanism to allow printing in full color. This would take special slicing however, as you would need to print the inside in white, and then the outside 1mm or so in color. I see another application in slicing a model and a support structure separately, and outputting the support structure to say, the cyan cartridge, but using a slower cure resin, which would allow for soluble support structures. In that system, the model would be outputted as a full black image, and the support structure would be outputted as a pure cyan image, thereby allowing the printing of two separate materials.

The UV light would probably be 2 100w UV LEDs available on eBay. One on each side of the print head, similar to the Objet machines. Obviously heat control would be an issue, but a water cooling system could be worked out relatively easily.

In addition, you could possibly, through the built in printer electronics, upload the images via USB or SD card, and only have the ramps board controlling the stepper motors.

This is definitely not a finished idea at the moment. However, I do plan on purchasing a large format Epson printer sometime in the future to begin piddling around with the idea.

Let me know what you guys/gals think!

-d

First off, this is stuff I'm very interested in: BrundleFab

For viscous materials, you want to use piezo heads, which means you need to start collecting Epson printers.

One issue you will have is that a good 80% (not all, but 8 out of 10 printers I disassembled from various brands) have non-optional mechanics underneath the print carriage/paper feed area. Usually a rod transferring power from the paper feed motor to the ink purge unit. These machines are non-usable for powder applications. So expect to tear apart a number of printers before you find one that will mechanically work.

The other issue you will have is a Continuous Ink System - you will need one compatible to your printer to supply the UV resin, and you must ensure all the piping is UV opaque, or you will get clogs.

See the 'layer head' I have developed for BrundleFab - it's sheet metal bent to support a halogen light and thermal monitor for my fusing pass.

You should be able to adapt the concept (and the layer head sheet metal CAD is OpenSCAD, fully parameterizable) to hold a metal halide buld, which should give you a ton of UV for curing.

Edited 1 time(s). Last edit at 09/11/2015 11:15AM by Ezrec.

.. also, you may want to look at BrundleFab's firmware.

It has support for stepper and (in the printhead) DC Motor + Encoder control, and for driving printheads (currently InkShield).

Should be easier to adapt to your system than the 'classic' stepper-only firmwares that have a lot of FFM assumptions.

If it is too difficult to get commercial printheads to work in this application, it might be worth trying to build your own. There are several projects on the wiki dealing with homebuilt piezo-driven inkjets:

Scratchbuilt Piezo Printhead
Reprappable-inkjet
Heated Piezo for Jetting Wax

Thanks for the input guys!

I've had some experience with some of the nicer large format Epson printers. And I have a model in mind that I may buy to disassemble. I am doing more research on resins, and am going to email MakerJuice about a thinner for their resins, as they are located in my state.

Thanks!

Ahoy gentlemen,
I am planning to work on this project,
Any updates?
I know there are uv sensitive resin printers in the market for printing on the metal plates etc.
Are they of any help to us?

Also if you need any assistance in hardware I am there.

Edited 1 time(s). Last edit at 04/16/2017 02:02AM by sudarshan.

Check this out
[www.kickstarter.com]

It seems the interest in inkjet resin printers is a bit low, since so much faster LCD resin printers are coming up.
Both are messy to work with ( in my imagination at least ) and that would keep me from building any of these.

Try to make a prototype....you will be a star overnight....
Jokes apart...have you tried the xaar 1001 printhead...they are versatile and can even print ceramics...they are meant for flatbed printing.
i guess there is a ytec3d forum in it.

No joke sudarshan, prototype forthcoming. As for the Xaar it is way more print-head than what I'm working towards. Xaar 1001-GS6 122/grams 1000/nozzle 70.5uM/pitch and the real kicker $1700usd. Head out of Epson XP430 35/grams 180/nozzle 217uM/pitch and the whole printer is $50usd. The older project inkshield uses a HP C6602 print-head and you only get 96dpi and with the Epson, I expect a very reasonable 300dpi. All of that is contingent on the ability to produce the drive waveform and bit-stream for the head in a timely fashion and causal (clock/sync) of which the Arduino style controller is not capable, that is why I took the development to FPGA for the DDS/AWG and the serial bit index stream for nozzle selection. As I mentioned in my pm to you, my biggest problem right now is resins with a low enough viscosity to jet properly. Do you know of any?

Sudarshan, the NexD1 was suspended by Kickstarter because people thought it fraudulent.

Edited 1 time(s). Last edit at 09/14/2017 09:46AM by myearwood.