Filament absorbs water from the air, and water in the hot end turns to steam. The fix has two parts: dry the spool you have, at the temperature its maker publishes, then store it sealed with desiccant so it does not get wet again. Prusa Research's knowledge base puts the underlying problem plainly: the materials used for FFF (fused filament fabrication) printing are mostly hygroscopic. The water is absorbed into the plastic, not sitting on the surface, and it comes back out at 200 °C or more, through the melt.
Bambu Lab's drying guide describes the result: the moisture vaporizes inside the nozzle, the molten filament expands and extrudes erratically, and you get stringing, oozing, holes, rough surfaces and reduced strength. Prusa names poor surface quality as the first signal, followed by low layer adhesion, blobs, and bubbling or smoke at the nozzle. If a print has gone downhill and nothing in the slicer changed, this is the first thing I check.
Which materials are most at risk
Prusa's article singles out polyamide (nylon), polypropylene, the water-soluble supports PVA and BVOH, and TPU as the materials most affected by moisture. Bambu Lab's table puts drying before use in the "required" column for nylon, PVA, PC and TPU, and only "recommended" for PLA and PETG. Bambu gives one number worth remembering for nylon: after about 3 months in open air, a convection oven at 80 to 90 °C struggles to dry a PA spool at all. Nylon and PVA are the spools that live in the sealed box, not on the shelf.
How to tell a spool is wet before you waste a print
- Listen at the nozzle during the first layer. Popping, crackling or visible steam is water leaving the melt.
- Look at the surface. Glossy gone rough or matte, with fine hairs between features, is the sign both makers name first.
- Weigh the spool against the net weight on its label, on a kitchen scale that reads to a gram. Heavier than the label means water.
- Check the desiccant. Bambu Lab's silica-gel packs change colour as they saturate; a fully changed indicator means nothing in that box has been dry for a while.
Drying temperatures and times the makers publish
The right numbers depend on the material and on the spool it is wound on, so use the figures from whoever made the filament you are holding. The figures below are the makers' own, not mine.
| Material | Prusa Research (Prusament, filament dryer) | Bambu Lab (forced-air oven) |
|---|---|---|
| PLA | 45 °C, 6 h | 50 °C, 8 h |
| PETG | 55 °C, 6 h | 60 to 65 °C, 8 h |
| TPU | 60 °C, 4 to 6 h | 70 °C, 8 h |
| ASA | 80 °C, 4 h | 75 to 85 °C, 8 h |
| PC | 85 °C, 5 h (PC Blend) | 75 to 85 °C, 8 h |
| Nylon (PA) | 90 °C, 6 h (PA11 carbon fibre) | 75 to 85 °C, 8 to 12 h |
| PVA | Not listed | 75 to 85 °C, 8 to 12 h |
Three cautions from the same pages. Bambu Lab says not to use a microwave or a kitchen oven: the heat is uneven and the hot spots near the elements can deform the spool or the filament. It says to flip the spool midway through drying, and warns that some third-party spools are not heat-resistant enough for the higher settings. If your spool's maker publishes nothing, start at the low end of the range for that material and check for softening before going higher.
Keeping it dry
- Store sealed, with desiccant, and measure it. Both makers say a closed box or bag with desiccant in it. Bambu Lab adds the target: below 20% RH a spool stays dry for 2 to 7 days by its figures. A hygrometer inside the box is how you know you are there.
- Regenerate the silica gel instead of replacing it. Bambu Lab's instructions: 80 to 90 °C in an oven on a heat-safe tray, 1 to 3 hours depending on saturation, never above 100 °C. The beads are good for 5 to 15 cycles.
- Feed nylon from a dry box while it prints. Polymaker's material notes say to run the dryer the entire time you print nylon in a humid climate. The spool loaded in the printer is the one exposed longest.
- Reseal the factory bag every time. The bag and its desiccant packet are the cheapest storage you own. The clock restarts the moment the spool is out.
When not to bother drying
A fresh PLA spool straight out of a sealed bag, printing a bracket that will live under a bench, does not need a drying cycle first; Bambu Lab lists PLA as "recommended" rather than "required". Skip repeated drying on polycarbonate, too. Bambu Lab warns that PC goes brittle after repeated heating and cooling cycles, so dry it once, keep it sealed through the job, and re-dry only if it has been exposed. And never push a spool past the temperature its plastic or cardboard hub can take to chase a number from another maker's table.
The expensive version of this problem
It is not the print that obviously failed. It is the part that looks fine, passes a quick visual check, and breaks under load later because moisture compromised the layer adhesion the whole time it was printing. Reduced strength is on both makers' symptom lists. A part that will carry real load, or has to hit a dimension, deserves a spool you dried and stored on purpose, and a decision first on whether it should be printed at all rather than machined.
If a print's surface or strength has quietly got worse without a slicer change, check the filament before the printer. Dry the spool to its maker's figure, get it into a sealed box below 20% RH, and keep nylon and PVA there between prints. If the first layer is still wrong after that, the bed leveling guide covers what a probe can and cannot see, and the tolerances guide covers what "fits" should mean. The 3D-printed tools I make here come off spools stored this way.
Checked against Prusa Research, Bambu Lab and Polymaker help pages on 2026-09-21: Drying filament (Prusa Knowledge Base), Filament drying recommendations (Bambu Lab Wiki), Printing tips by material type (Polymaker Wiki).
