Building Large(r) Format Printer For Ceramics

Powder and inkjet printing
Post Reply
Posts: 1
Joined: Wed Jan 20, 2021 4:33 pm

Building Large(r) Format Printer For Ceramics

Post by BlackAndChrome »

So I have a been entertaining the idea of building a 3D printer exclusively for ceramics as I have my own kiln. You may be familiar with the paste extrusion SCARA based machines sold by 3DPotter and some of the similar Open Source paste extrusion products, but to me dealing with powder sounds substantially easier than getting tubes filled with slip/slurry of an exact consistency and completely void of air. Not to mention self support limitations of paste, extruder mass when loaded with sufficient clay for large parts, etc.

Anyway, I'd like to build something based on your Oasis system with the largest deviation being build volume and an emphasis on build speed over maximum accuracy. We are talking mugs, vases, sculptures etc nothing intricate like small lattices or figurines. Initially I would like to aim for a build area around 100mm^3 (cubic or cylindrical) but with the ability to go with detachable feed canisters to allow for a 200mm or even 300m swap out. The way my brain works I'm much better working on software when I have hardware in hand to put it on so hoping you could help me make sure I'm on the right path before I leap in feet first building the hardware.

I noticed the build log says there is room for 5 print heads, were you intending this for parallel production or staggered to increase the swath size? Is there much room to expand on layer height before the print head has to slow down in order to dispense enough fluid making? Do you have some typical layer height and axis speeds I can use to estimate print times? Any general words of wisdom for someone headed in this direction appreciated.

For applications like this I was also curious if you had ever heard of anyone using a single "jet" with the typical profile printing of FDM, Something like a tiny mist emitter or scooter fuel injector? Just a shower thought I had hah.

User avatar
Site Admin
Posts: 503
Joined: Fri Aug 14, 2015 4:48 pm
Location: The Nethelands

Re: Building Large(r) Format Printer For Ceramics

Post by dragonator »

Hello and welcome Nick

Powder printing has different drawbacks. Cleaning parts is inherently messy and powdered ceramics require respirators. Also parts are quite weak before firing, so cleaning them is challenging. I have yet to have made parts that stand cleaning as well as I would want to. If you can get the formulation right, it should be doable though. The advantages are indeed a higher resolution, no layers, and fewer limitations regarding supports. The not filling of the slurry into canisters is also a big plus.

When you mention larger and say 100mm^3, I will assume you mean a 100x100x100mm cube, and not a 100 cubic millimeter build volume. This size should already be reasonably doable on bare Oasis hardware. 200 or 300mm is a bit more challenging, but that can be fixed, either in software by making the software send while printing (something I have not yet spent much time on fixing) or with a newer controller (I am working on V4.0, which will should have 4x the RAM and at 300DPI can store a full 12 inch sweep in memory). The current V3.2 controller will need to run at 100DPI to fit your max size. All of these, future improvements and current hardware should work.

The logs do indeed say 5 printheads. Staggered is an option, though getting staggered printheads aligned sucks. The intended use was for full color printing. Clear binder, and CMYK will allow for full color printing. I never got around to making that though, so I have a massive printer with a tiny buildplate.

For speed you will always be limited by a maximum layer height and the time it takes to deposit a new layer. The thicker you make the layers, the more ragged the shell becomes. Right now I print at 0,1mm to 0,2mm. Depositing a layer takes some time. You can't do it too fast or you risk damaging the layer below. I'd have to look at the videos, but I think depositing a new layer takes me around 10-15s. Every second you can take from this speeds up your print time significantly. The maximum print speed I have achieved is around 115mm/s (4,5"/s). This is limited by the max number of droplets ejected from the printhead per second. Each nozzle currently only manages around 2700 droplets per second. If you print 600 dots per inch, you will have 4.5"/s. My experience is that powder and inkjet prints fairly wet, so 150% to 200% is not out of the question. At 200% (or 1200 droplets per inch), your speed is halved. I print only in the positive direction, returning to the starting point each time, but printing should be possible to do in both directions. There are some delays to send the data, but that is mostly limited by python for me. It simply takes time to send data with python. A full sweep on a fast computer might take 1-2s. If you can send while printing, this limit fades.

The printhead has a swatch of 12,7mm (0,5") so printing at 4.5"/s both ways with 1s between direction changes, 200x200mm and 0,1mm layers (@10s) will take you 200mm/12.7mm=16 sweep, 200mm/115mm/s=1.75s, (1.75s+1s)*16sw + 10s nl = 54 seconds per layer, and 9 minutes per millimeter of part and an hour and a half per centimeter. Smaller printed areas might speed it up somewhat, but that gives you a rough idea for the values I presented.

I am familiar with single jets (this is how Focus worked with the C6602A) but it kind of defeats the benefits of binder jetting. Most systems either eject droplets, of can at least turn sprayers on and off fast enough that you can make sweeping movements. Simply adding more nozzles makes it a lot faster.

Hope this helps you get an idea of what to expect with the hardware I have experience with, and what to improve to get to the speeds you need. I do advice to have the powder you want to print tested before committing to a large printbed. Powder testing sucks if each batch is 4kg.
Post Reply