Project Description: This five-person project was the final project for my Additive Manufacturing course in Fall 2022. We were tasked with iterating on a prior graduate student’s research for SLA printing onto cylindrical surfaces. The goal of the research was to print a film that could be collected from the cylinder and respooled. The intended application is to eventually use this film printing technique as a substrate for cell growth for artificial meats, where the printing process could help recreate meaty textures. DRUM stands for Digital Rotational Ultraviolet Manufacturing (because 3D printing processes always need a new acronym).
Software Used: Fusion 360
My Role: I led our team since I had prior experience with SLA 3D printing. I worked closely with a friend on the team to collectively design our mechanical system to retrofit a Photon Mono X. I tackled the print surface drum, while he focused more on spooling the film we’d print. I also worked on figuring out a hacky way of using the printer’s slicer to expose useful patterns for our system.
Other Roles: The other teammates took on the electronics for our printer consisting of controlling a stepper motor for our rotation, including calculating the required speeds of rotation for our curing times.
Big Wins: I was able to develop a sound prototype that allowed us to print onto the cylinder. My mechanism for leveling the cylinder worked well enough. Another group of students from the class opted to continue this project for their senior design, where they were able to learn from our prototype to refine this approach aimed at resin printing long continuous films with complex patterns.
Challenges: The time allotted for this project was merely the final five weeks of the semester, which included Thanksgiving break, so we were only able to take the design so for with minor iterations. Additionally, we were provided with a very minimal budget, so we had to resort to resources available in our shop’s scrap room outside of ordering belts. Due to these limited resources and issues with our electronics, we never printed a successful continuous film.
Key Takeaways: I learned the difference between steel and hardened steel shafts when going to the machine shop to trim our axles down and needing to use a grinder instead of the band saw. The final design we went with wasn’t our favorite in terms of capabilities, but it was the one we had time and resources to make a prototype for, so we made a good decision there, even then, every design needs more time to iterate than you anticipate.
Here are three concepts we had to construct a printer capable of printing resin films onto cylindrical surfaces.
The first idea consisted of shearing a layer of resin onto a conveyor and then curing it via DLP.
The second idea was putting a UV lamp and mask setup on the inside of a cylinder that would be submerged in resin to cure local films on the surface.
The third idea was to try and use the optic system from an existing MSLA or DLP printer, and retrofit a cylindrical print surface onto the machine. We ended up choosing this option due to time constraints for the project.
Here is our final CAD, with a spotlight given to the leveling mechanism for the cylinder’s shaft, controlled by screws on either end of the shaft. This retrofit addition had to mount to the printer without drilling any holes or using adhesives, which I achieved by incorporating our design around the mounting screws that held the vat in place.
With this design, we needed to accurately figure out which portion of the bed we wanted to expose, so we made a calculator in Excel that allowed us to tune parameters such as exposure area and then determine our rotation speed to achieve enough curing of the resin.
Here’s our assembled prototype. We also added a layer of thick paper over areas we didn’t want exposed to prevent UV leakage through the MSLA mask resulting in resin waste.
Here is the extent of film we were able to cure onto our cylinder. We ran into issues with resin sticking to both the FEP film and our cylinder, plus motor controls that didn’t work reliably. Tuning the cure time or more motor torque could have potentially peeled the film up as a continuous process.