3D Printer Vacuum Bed (2/17/2017)

One of the biggest problems people run into when 3D printing ABS or exotic materials is warping caused by uneven cooling while the part is being printed. To solve this, Stratasys has patented solutions like heated build chambers, and reprap machines have implemented heated beds with ABS slurries or gluestick/hairspray to help parts stick to the bed. My solution is to pull a vacuum under the print to counteract the forces that cause warping.

By having a vacuum under the part, you would have nearly 15psi holding the part to the build platform as it’s being printed. That’s under perfect conditions, but by using a relatively affordable vacuum pump it would be possible to have 10psi holding the part down. Additionally, after the part is done being printed, air could be pumped in to have a positive pressure under the part to automatically push it off the bed.

With the basic principle in mind of a build platform with an air chamber that has a controllable pressure built in, the next step was to start sketching out concepts. I found that the easiest thing to do would be to use a perforated sheet of aluminum as the print surface, and then machine a pressure chamber for it to mate with. Then by adding hose barbs and a vacuum pump, a basic prototype could be built easily enough to test the functionality. With a prototype in hand, I would then be able to make any necessary tweaks. To design a prototype, I opened up my CAD software and got to work.

I decided to make the build platform to fit a Makerbot Replicator 2X that my school had but wasn’t using, and got permission to use it to test this out. I measured the build platform that was already on the printer, as well as the bolt pattern so I could make sure my platform’s XY space would be the same and to make sure I’d be able to mount it. I also decided to make it as close to the same thickness as possible to minimize any loss in build volume., and so that I wouldn’t have to move the Z-axis homing switch. With these constraints, I sketched out a solid block that was the same size as the original build platform so I could take material away from there.

I decided to make the perforated plate from 1/8” thick aluminum since it was readily available, easy to machine, and, in my experience with Makerbot Replicators, a decent build surface. I then extruded a pattern of 1/8” holes across the build area, mostly because they would be easy to drill, although smaller holes would have been better since they would have less unsupported area to bridge. I also thought it would be a good idea to put a thin layer of silicone rubber on the bottom of the build plate to form a better seal with the base plate.

The next step was removing the material from the base plate. I decided to make it from 5/8” aluminum, again because it was readily available and easy to machine. I extruded a cavity in it for the air to be pulled out of. I made this cavity larger than needed to prevent any large pressure changes from small leaks between the build plate and base plate. To prevent the build plate from bowing under the vacuum though, I extruded posts to support it in places that I felt wouldn’t majorly impact printed parts, since the posts would block holes, preventing air from being evacuated. To change the air pressure in the base plate, I placed four holes at the bottom for hose barbs. The barbs and hoses needed to be out of the way of the build plate support arms, so I took a measurement of the maximum dimension between the arms. The plan was to use two hoses to evacuate air, and two to pump air in and eject the part from the build plate, with the hoses going into valves controlled by the printer’s control board and GCode commands.

I made a first prototype from plastic to make sure that the build platform would fit on the printer, and to make sure I wouldn’t need to make any design changes. The base plate was machined from a piece of ABS plastic, and the build plate was made from a piece of laser cut acrylic. This prototype fit well enough to test the printer, but the mounting bolts were a bit too far apart, so I moved them closer together in the design. While leveling the bed, I found that the build platform was a bit cumbersome to level, and the springs were a bit too weak because this platform was heavier than the original, so stronger springs would be needed for a metal prototype. Regardless of these issues, the build platform worked well, and I started making plans to build a metal prototype.

Unfortunately, at this point I got busy with school and some things in my personal life, so I wasn’t able to build a metal version, or test the vacuum pump. Some things I did note though while designing this is that the entire build platform would need to be covered in order to pull an adequate vacuum. Some ways that this could be solved are by printing one or two solid layers that cover the whole platform, and then printing the part on top of them, by covering the area not needed for the printed part with some kind of tape (maybe combining this with some laser cutting process and a taping/de-taping machine for automated production), or by making custom perforated build plates to only have holes where they’re needed (also better for automated production). I plan to attempt this again at some point on my own printer when I have more time.