From January to April of 2019 a small team of engineering students and I investigated a dual nozzle approach to dual color clay 3D printing. We were able to generate some very promising results, and also encountered a number of complications.
In order to make this happen we adapted a clay printer design I have been consistently tinkering with since 2015. At the outset we confronted a question: “One nozzle for multiple materials, or one nozzle for each material?”
In the realm of consumer plastic printing both approaches have been pursued, but in recent years it seems that the approach of using one nozzle through which multiple filaments can move has become the preferred mode. This approach can be seen in printers such as the Prusa MK3S MMU2S, or achieved another way in the Mosaic Palette 2. In the realm of clay printing the only dual color printer we were aware of was the Lutum Dual, whose press materials suggest they were the first to attempt dual color clay printing. We didn’t consider using a Lutum machine for the project as our printer development has steered us away from air pressure driven systems and toward an approach that utilizes a mechanical feed, as we believe this allows more thorough integration between software/firmware/hardware.
We did not choose a single nozzle approach for two reasons. First, we found that when “purging” one clay color for another, clay had a tendency to mix and linger much more so than plastic filaments do. Second, the amount of material that needed to be purged from the system was much larger with clay than it was with plastic, due to our extruder design.
So we built a machine with two independent extruders.
Our design was a mechanical feed with an auger in each print head. More information about this type of system can be found here. The design is ever-evolving, but has a build volume of approximately 700mm x 450mm x 810mm, not that we ever made anything close to that size. The printer has been constructed in an ad-hoc fashion over a period of years and has approximately $3000 worth of materials in it. It will continue to be revised going forward.
Once we had developed a tool, we needed to take a look at software with which to prepare models and generate G-cod instructions for the printer to execute. We looked at Cura and at Slic3r PE and ultimately went with Slic3r PE as our initial results were more promising. Having gotten excellent results with both programs when printing in plastic with commercial machines we viewed the choice as not particularly important as both could be good solutions and switching between them would be possible if need be. Late in the project Pathio emerged as an entirely new program with great potential. While we didn’t use Pathio in testing, we did find the Pathio forum moderators willing to help us work with the program. We’ll keep an eye on it going forward as its association with the E3D toolchanger and integration with the Duet 3D control board we use suggests it will grow in ways that would be beneficial to our specific needs, even if it is not an open source project.
Printing was less complicated than we might have imagined. Once the various objects in a design were assigned to a particular extruder the print process unfolded in a very predictable and familiar way:
We tried several different clays in the process and found that any of them would work as long as they were compatible with each other. Our primary clay was a simple white earthenware clay to which we added mason stains to make other colors such as blue and black.
However, throughout all of our dual color printing we kept running into one persistent problem, which was predictable from the outset. We continually struggled with the nozzle not presently in use dragging across and disrupting the surface being laid down by the other nozzle.
This is a well established problem in plastic printing. So much so that novel solutions have appeared in that space, such as nozzles that tilt when not in use, lift and lower when not in use, or which move independently of each other and get out of each other’s way. These approaches can be seen in numerous commercial printers such as the BCN3D Sigma, or the Lulzbot Taz Pro.
For now, we will continue exploring dual color printing with the limitations imposed by two fixed nozzles and see how we can make the most of this system. In the longer term, other solutions to the issue may be explored - including the possibility of a solution similar to the the toolchanger mechanisms we’ve seen recently.
Along the way, our team made numerous significant improvements to the printer, process, and materials. Many of these improvements were outlined by the team in the posters they developed to showcase the project:
As April 2019 comes to close our project ends and the team of students graduate and embark on their careers elsewhere. We will to continue to develop and build upon the ideas explored in this work.