Over the past few months, we @ Social Bicycles have been blown away by the quality, speed, and reliability of our Ultimakers. The parts coming off of the machines have changed the way we make products - not only how we develop and test them, but even how we mass produce them. PLA is a great material, too, and with the right settings and a high quality filament (Printbl FTW!) we've been getting beautiful, precision prints with ample strength for both prototyping and real-world requirements.
Nonetheless, our application (bike-sharing systems) frequently demands something higher performance than what PLA is capable of delivering. I've experimented with other materials, like ABS (smelly, tough to get good prints from a stock UM) and Taulman Nylon (awesome material, strong, but not as rigid as PLA, and a lot slower to print) but what I knew I really wanted to print, from the beginning, was a material we already make many of our production parts in - Acetal. (As a quick overview, Acetal (or Delrin, or POM) is a strong, rigid thermoplastic with excellent wear resistance and very low friction. It is frequently used for gears, bearings, and moving parts.)
After several weeks of testing, I am happy to say that we are 3D printing in acetal. We are getting amazingly strong, slippery, rigid, complex parts with only very minor modifications to our stock machine. It is frequently said that 3D printing is limited in its ability to replace traditional mass-production because of the limited materials it can utilize, but Acetal is one plastic for which there is generally no superior, and the parts we have been able to produce are certainly as strong as their injection-molded counterparts.
Here are a few photos of one part I printed. More details about the process below.
Print in progress
A buttery-smooth helical focusing mechanism for a camera lens I have (Printbl Grape PLA, White Acetal) Will post thingiverse model soon.
The "filament". ACETRON. Lawl.
After searching around for prior examples of acetal being used in 3D printers - or a source for acetal filament - and finding nothing, I turned to McMaster-Carr to see if they had what I needed. Of course, they did. Mcmaster stocks 4' lengths of 0.125" (~3.2mm) white acetal rods - and while that diameter is too large to fit in a stock Ultimaker without clogging... Mcmaster also has 3.4mm drill bits for opening up that pesky hot-end's ID, and thin-walled teflon tubes for replacing the bowden.
Bowden Replacement: 5239K12
3.4mm Drill: 30565A254
Acetal Rod: 8497K11
Bowden Stiffener: 5239K13
( ^ Sleeve this over your new, thinner bowden in order to increase rigidity and improve retraction when printing in floppy filaments like ABS, Nylon, PLA)
The hardest part about printing acetal is getting it to stick. A heated bed would probably help here. My best results come from printing on a wooden platform (birch ply) and a large raft with very thick lines. Temp of 258° seems good, although thin sections (<1mm) can get layer adhesion problems, so higher temps might work better. No fan, 30mm/s speed (much higher is probably possible). 0.1mm layers gives amazing detail. Acetal seems to hold heat very very well, which means that thick sections (>4mm) get messy, quickly. Parts that are designed as if they were to be injection molded (ie with uniform wall thicknesses) seem to print great.
Apart from first layer adhesion challenges, the only real problem with Acetal is that it (currently) only comes in 4' rods. The problem here isn't running out of filament - it's easy enough to feed them into the machine as necessary. The problem is that unless you fuse the rods together, it is impossible to use retraction once there is more than one filament section inside the machine. This makes prints with fine detail and many jumps more challenging.
My next goal is to try to find a source for proper acetal filament - there are probably manufacturers that will do custom lengths. There are also many different Acetal copolymer mixtures with very different melting points, so it is entirely possible that a different blend of Acetal could be even more 3D printing-friendly. I'm also excited about setting up a dual-extrusion machine so that first-layer adhesion can be solved by printing a few layers of PLA first. A heated build environment is something I would also like to try. Would love to hear about the results from other people who try printing in Acetal and have a heated build environment.