The filament breaks at the bend between the spool and the extruder, or inside the feed tube, on a spool that printed perfectly a few months ago. Sometimes it snaps as you pull it off the reel; sometimes the print stops halfway through with a short stub still in the extruder.
Two causes that look identical
Water. Several polymers absorb moisture from the air, and in some of them water attacks the polymer backbone over time, cutting long chains into shorter ones. Short chains cannot distribute stress along their length, so the material loses its ability to bend.
Age. Ultraviolet light and ordinary oxidation do the same thing by a different route. A spool on an open shelf near a window is being slowly degraded by daylight, and every polymer here has a shelf life measured in years rather than decades.
Both leave you holding filament that snaps. Neither can be identified by looking at it. What separates them is that one is reversible and the other is not.
The test
Dry the spool at the temperature this site lists for that material — 45 °C for PLA, 65 °C for PETG and ABS, 80 °C for the polyamides — for the full listed duration, then bend a length of it around your finger.
- It bends without breaking: it was moisture. Print it, and fix the storage.
- It still snaps at the same radius: the polymer has degraded. No amount of further drying will restore it, because what was lost was chain length rather than dryness.
That second answer is worth accepting quickly. A degraded spool will keep producing weak parts and clogs, and the hours spent nursing it cost more than the spool did.
The third cause, which is mechanical
Filament can also snap because of how it was wound rather than what it is made of.
The last part of a spool sits at the smallest radius, so it has been held at the tightest curve for the whole of its life. On a cheap spool with a small core, that curvature is permanent by the time you get to it, and a brittle-ish material breaks trying to straighten out.
Similarly, a spool that was wound under high tension puts every wrap in permanent bending stress. If a spool consistently snaps in the last quarter and was fine before that, the material is probably fine and the reel is the problem.
Storage that actually works
- Sealed container with fresh desiccant. Not a plastic bag with the top folded over. Indicating silica gel changes colour so you can tell when it has stopped working — replace or regenerate it when it does.
- Out of daylight. A cupboard, a box, anywhere that is not a shelf by a window. This costs nothing and it is the whole of the ultraviolet problem.
- Secure the loose end to the reel whenever a spool goes back in the box. Nothing to do with brittleness — it is how you avoid meeting a tangle at hour six of the next print.
- Buy what you will use. A spool bought for one project and opened once a year is being stored in the worst way that exists, which is opened.
What snapping filament does to a print
It rarely fails cleanly. A break inside the feed tube leaves the print running on the remainder in the hotend, so it under-extrudes for a while before it stops — which shows up as a band of thin walls at whatever height it happened, then nothing. A break at the extruder can leave a short stub jammed in the drive gear that has to be removed before anything else works.
If a print failed this way, check the whole path before restarting: a stub left in the tube will stop the next load dead.
The materials that go brittle
PLA is the most common, because it is the most widely owned and it ages noticeably over a year or two. PVA and BVOH are the extreme cases: a material formulated to dissolve in water pursues it out of the air just as enthusiastically, and leaving one uncovered over a weekend can be enough. The polyamides absorb more water than anything else here and go both brittle and unprintable. PETG resists this better than its reputation suggests: wet PETG prints badly long before it snaps.
Anything fibre-filled deserves its own note. The base polymer may be fine while the composite is brittle, because the fibre creates stress concentrations at every particle. Those spools are less bendable when new and should not be judged against a plain filament's flexibility.