It is late spring in the DC metro area, which means increasing humidity. Despite my best efforts, PLA filament spools used by one of our 3D printers seem to absorb moisture, although our warehouse is not exactly a controlled environment. The effect, at least on our machine, is a lot of bad prints. Well, to be fair, I am linking the increase in humidity to bad prints, as despite cleaning and other adjustments, no other cause is apparent. For those of you in the know, the filament brand, lot, and color (in this case no dye) has remained constant, so there has not been some recent change in filament for which we have not calibrated our printers.
The prints fail as the filament stops feeding through the hot end / print nozzle. At times it appears the filament slips against the gripping gear attached to the feed stepper motor, and so it is simply not fed through. With other failed prints, the PLA appears to have ‘exploded’ or bubbled on the hot end, such that the feeding plastic rolls over on itself to eventually clog the nozzle. Adjusting the print height doesn’t appear to affect this either. The tension on the gripping gear has been adjusted, and actually the whole print head has been completely disassembled and cleaned multiple times. Doing so does improve the situation so that we can get a full print. But at other times, we still can only get a partial print. Any benefit from complete cleaning and adjustment is fleeting.
PLA is a hygroscopic thermoplastic, and according to RepRap wiki entry on PLA, a test sample absorbed enough moisture from the air to increase its mass by 0.5%. A handy table comparing the properties of PLA to other materials, in the chapter Polylactic Acid Technology of Natural Fibers, Biopolymers, and Biocomposites (2005), lists the moisture regain as 0.4-0.6%. It is amazing that such a small amount of absorption can so completely ruin a print, as in comparison to other materials, PLA is actually quite hydrophobic.
Of course in the environments where the IEPAS MMA project plans to operate, the negative effect of temperature and moisture on PLA filament is a cause of real concern. The countries the project is meant to assist are often tropical; typified by high humidity, heat, and sometimes saline air. Moreover, the buildings of many meteorological services in developing countries are not fully climate controlled; instead relying on natural flow of air, fans, and the like. Air conditioning is often reserved for operations centers, server rooms, and sometimes a few executive offices. A construction shop is unlikely to be provisioned with air conditioning and dehumidifiers, and nor should it, given the relative cost and reliability of electricity.
Fortunately, there does seem to be a simple fix that does not involve locking the filament in a special dehumidifier or baking the filament at low temperatures for long stretches of time. The default temperature setting for PLA seems to be 230 degrees Celsius, and on the RepRap wiki entry mentioned above, 230 degrees is also referenced as the upper limit. We reasoned that a higher print temperature might make the PLA less viscous, thereby helping to minimize the bubbling or other apparent explosive effects of the escaping water vapor.
The Makerbot support guide on PLA also notes that the plastic is ‘heat hungry’, and sometimes a low print temperature can cause the print head to clog.
On the other hand, you might have more extrusion problems if you’ve set your extruder temperature too low. That’s because PLA uses a lot of heat to change from its solid to its liquid state, and it’s actually pulling that heat out of the nozzle and cooling the nozzle down. If the nozzle isn’t hot enough to keep melting PLA, the filament inside it can harden. This shouldn’t be an issue unless you’re printing at very low temperatures or at very high speeds.
The sweet spot for our machine is 238 degrees Celsius. We also started running on the fast setting, rather than standard. Surprisingly, the prints are at least visually better than even when printing on the standard setting with new (‘dry’) PLA filament.
Running at a higher temperature does not quite explain why the feeder kept losing grip of the filament, although I hypothesize that there may have been greater back pressure on the incoming filament at the lower temperature, which eventually caused the gripping gear to shave bits of the incoming filament until enough build up left it without any friction.