Saturday, 27 April 2013

3D Printing Nylon

Oozy, webbed nylon printing
Nylons can be exceptionally strong, much more so than ABS or PLA, especially for tensile stresses. That makes them a tempting materials for 3D printing mechanical parts, so I grabbed some Taulman Nylon 618 to check it out.

At first my prints were unbelievably oozy - the nylon just flowed from the nozzle and leaked all over the print. I also had problems getting it to adhere to standard blue tape. Apparently both of these are pretty common.

Damaged bed surface after print removal
I addressed the bed-sticking problem with a cheap sheet of MDF board; 618 nylon sticks to cellulose really well. In fact, it works so well that the board's surface gets damaged each time I remove a print; ideally I'll be able to find some tape-mesh to decrease the contact area, or maybe just something less sticky, I suppose I could also just make it easy to remove the wooden bed and keep switching them out (given how cheap the stuff is).

The oozing problem was far more insidious. Initially, I tried to use retraction, but it had no effect (actually it made things worse by causing the print head to sit still momentarily, melting the plastic beneath it and letting it flow). I then happened upon a great idea; get rid of retraction entirely. Get the print head moving so quickly that not much plastic has time to flow while it's moving. In combination with the 'avoid crossing perimeters' flag in slic3r (which forces the printer to complete each part separately, getting rid of that spiderweb of inter-part trails), and the switch to a 0.2mm nozzle produced magnificent results:
A lego block being printed with very clean, ooze-free walls
Notice that only a single line of oozed filament conencts each component thanks to the 'avoid crossing perimeters' flag
I also had to significantly lower the extrusion multiplier to aid in ooze-prevention, though this may just be the result of thicker-than specified filament. I'm still tweaking the extrusion rate settings to get perfect, smooth surfaces, but the intermediate results are totally functional:
InMoov index finger - the surface is still a bit rough, but the sanded joints are slippery-smooth and its many times stronger than the ABS versions I've managed to make (which suffer from cracking and shattering near the weak, stress-bearing pivot points).
Printing at the full 245°C significantly improved the strength of printed parts and didn't seem to have any apparent adverse effects on the teflon of the Printrbot Ubis hotend (though be aware that every thermistor will have a slightly different calibration). Fortunately the Ubis hotend only includes teflon on the outside of the PEEK barrel, so it should be a few degrees cooler than the nylon inside. I didn't use any bed heating at all. Use of a cooling fan while printing turned out to have a slight detrimental effect.

Nylon loves to absorb water, water that boils and bubbles out of the extruder during printing; it's often recommended to dry it out in an oven at low heat before printing, but I haven't found that to be necessary. There's a small trail of steam from the print nozzle, but there appear to be no adverse effects, though this may not be true if you live in a high-humidity area.

I could comfortably hammer the printed lego parts without damage; a massive improvement over ABS. Despite some reports that gluing/bonding 618 is ineffective, I found that superglue worked perfectly on sanded nylon parts (see the index-finger above).

Apparently, natural nylon is very easy to colour with fabric dyes too; this allows each printed part to be coloured separately, removing the need for multiple rolls in different pre-dyed colours (though I have yet to test this).

All in all I'm quite impressed; nylon 618 is a fiddly material for printing, but the results are certainly worth it for any part that needs significant mechanical strength.

No comments:

Post a Comment

About Me

Melbourne, Victoria, Australia
I'm hard at work on an MSc in Physics at the University of Melbourne, my research topic is 'building nano-scale neural networks'. In my (limited) spare time I tinker with 3D printing, electronics and programming.