Despite the reputation of polymers used with FDM 3D printing like nylon, ABS, and PLA as being generally indestructible, they do come with a whole range of moisture-related issues that can affect both the printing process as well as the final result. While the concept of ‘baking’ such 3D printing filaments prior to printing to remove absorbed moisture is well-established and with many commercial solutions available, the exact extent to which these different polymers are affected, and what these changes look like on a molecular level are generally less well-known.

Another question with such hygroscopic materials is whether the same issues of embrittlement, swelling, and long-term damage inflicted by moisture exposure that affects filaments prior to printing affects these materials post-printing, and how this affects the lifespan of FDM-printed items. In a 2022 paper by Adedotun D. Banjo and colleagues much of what we know today is summarized in addition to an examination of the molecular effects of moisture exposure on polylactic acid (PLA) and nylon 6.

The scientific literature on FDM filaments makes clear that beyond the glossy marketing there is a wonderful world of materials science to explore, one which can teach us a lot about how to get good FDM prints and how durable they will be long-term.

Why Water Wrecks Polymers

Although the effects of moisture exposure on FDM filaments tend to get tossed together into a simplified model of ‘moisture absorption’, there are actually quite different mechanisms at play for these different polymers. A good example of this from the Banjo et al paper is the difference between nylon 6 and polylactic acid (PLA). While nylon 6 is very hygroscopic, PLA is mostly hydrophobic, yet this does not save PLA from getting degraded as well from moisture exposure.

In the case of nylon 6 (C₆H₁₁NO), the highly polar functional groups such as amides (−C(=O)−(N) ), amines (−NH2) and carbonyls (C=O) make this polymer hydrophilic. As these functional groups are exposed to moisture, the resulting hydrolysis of the amide bonds gradually affects the material properties of the polymer like its tensile strength.

A few percent moisture in the polymer filament prior to passing through the hot extruder of an FDM printer will correspondingly cause issues as this moisture rapidly evaporates. And after printing a nylon object, moisture will once again hydrolyze the amide bonds, weakening the material over time. This is something that can be avoided somewhat by sealing the object against moisture intrusion, but this is rarely practical for functional parts. This degradation of polyamides can be observed by the cracking of nylon gears in toys gearboxes, servo motors, and similar high-stress applications.

In the case of PLA ((C3H4O2)n), far fewer polar functional groups are present, making PLA effectively hydrophobic, although it is soluble in various organic solvents like ethyl acetate. PLA’s weakness lies in its ester bonds, which are subject to hydrolysis and can thus be broken like amides. This type of hydrolysis in PLA is very slow, however, with studies finding that it barely degrades even submerged in water. The often cited ‘composting’ of PLA thus requires chemical hydrolysis, making options like incineration the faster and easier route for disposal. As a result, for long-term stability PLA does rate highly, regardless of its other material properties.

Naturally, in the case of all hygroscopic polymers the rate of degradation depends on both the moisture content of the air, and the temperature. In the earlier referenced study by D. Banjo et al., the FDM printed samples were fully submerged into water to accelerate the process, with three types of polymers tested at 21 °C and 70 °C.

Freshly Baked Polymer

Drawing the moisture out of the polymer again can be done in a variety of ways, with applying heat over an extended period of time being the most common. The application of thermal energy motivates the water molecules to make their way out of the polymer again, but it is important to understand here that hydrolysis is a permanent, non-reversible process. This means that the focus is primarily on removing any absorbed water that can be problematic during extrusion, and to prevent further degradation of the polymer over time.

A paper presented by Xuejun Fan at the IEEE EuroSimE conference in 2008 titled “Mechanics of Moisture for Polymers: Fundamental Concepts and Model Study” covers the fundamental concepts related to moisture intrusion which ultimately enable the degradation. In particular it is of note that the effects of submersion (water sorption) versus exposure to the air (moisture sorption) lead to very different transport mechanisms, and that there’s a distinction between bound and unbound water inside the polymer. This unbound water is contained within microscopic pores that exist within the material, and would thus be a good target for forced eviction using thermal means.

Exactly how much heat has to be applied and for which duration differs wildly, based mostly on the type of material, with commercial filament dryers generally having presets programmed into them. Filament drying charts are available from a wide variety of sources, such as from Bambu Lab. They recommend drying PLA filament at 50 °C – 60 °C for eight hours, while Prusa recommends drying PLA for six hours at 45 °C (and PA11CF reinforced nylon at 90 °C). This highlights just how hard it is to make definite statements here other than not heating up a spool of filament to the point where it softens and sticks together. The question of ‘how long’ would be ideally answered with ‘until all the moisture is gone’, but since this is hard to quantify without specialized equipment, longer can be said to be better.

Perhaps the biggest take-away here is that preventing moisture from getting even near the polymer is by far the best option, meaning that keeping spools of filament in vacuum bags with desiccant gel between printing sessions is highly recommended.

Endurance

If water molecules cause physical damage to the polymer structure, how severe is the impact? Obviously having unbound moisture in the filament is a terrible thing when trying to melt it for printing, but how long can an FDM printed part be expected to last once it’s finished and put into use in some kind of moist environment?

For PLA and nylon we can see the effects illustrated in the D. Banjo et al. study, with parameters like moisture absorption, crystallinity changes, and mechanical performance examined.

Perhaps most fascinating about these results is the performance of PLA, which at first appears to outperform nylon, as the latter initially shows a sharp decrease in mechanical properties early on. However, nylon stabilizes while PLA’s properties at either water temperature completely fall off a cliff after about a week of being submerged. This brittleness of PLA is already its main weakness when it’s not subjected to hydrolysis, and clearly accelerated aging in this fashion shows just how quickly it becomes a liability.

The big asterisk here is of course that this study used an absolute worst-case scenario for FDM-printed polymers, with water sorption in relatively warm to very warm water. Even so, it’s illustrative of just how much different polymers can differ, and why picking the optimal polymer for an FDM print completely depends on the environment. Clearly PLA is totally fine for many situations where its disadvantages are not an issue, while for more demanding situations nylon, ABS/ASA, or PC may be the better choice.

Keeping filament dry, vacuum-packed and far away from moisture will significantly improve printing with it as well as its longevity. Printed parts can have their surface treated to seal them against moisture, which can make them last much longer without mechanical degradation. Ultimately FDM printing is just a part of the messy world of materials science and manufacturing, both of which are simultaneously marvels of modern science while also giving engineers terrible nightmares.