This family has one axis, and it explains the whole shelf: how much the polymer has been prevented from crystallising.
A regular polymer chain packs into ordered crystalline regions as it cools. Crystals are denser than the disordered melt they form from, so a crystallising polymer shrinks considerably, unevenly, and for hours after the print stops. Suppress crystallisation and the material becomes clear, dimensionally calm and forgiving. Every member here is the same terephthalate polyester with a different amount of suppression built in.
| Material | Crystallisation | Choose it for |
|---|---|---|
| PET | not suppressed | recycling projects and PET's specific chemistry, accepting a narrow process window |
| PETG | suppressed by a minority co-monomer | almost every functional part below about 70 °C |
| PCTG | suppressed further, the ring diol dominant | clear parts that must stay clear after being flexed or struck |
PETG is the middle entry and it is where roughly everyone should start. The other two are answers to specific questions.
When this family is the right shelf
The case for a copolyester is that it does the job of PLA and the job of a styrenic at once, without the drawbacks of either. It survives outdoors where PLA goes chalky, tolerates a warm car better than PLA, takes impacts without shattering, and prints without an enclosure, without a strong smell and with a fraction of the warping.
That combination is why it is the default answer for functional work: outdoor brackets, protective covers, water-contact fittings, printer parts, anything mechanical below about 70 °C.
Where the family runs out is heat and detail. If the part goes above roughly 80 °C, the styrenics are the next shelf. If the part is a display model or a miniature, PLA prints it sharper.
What every member shares
They hydrolyse. This is the family-defining vulnerability, and it is the reason every member of the family carries a drying requirement rather than a drying suggestion. The damage is chemical rather than cosmetic: the polymer comes out of the nozzle with a lower molecular weight than it went in with, and no amount of tuning recovers it afterwards. The whole family sits at 65 to 70 °C for four to eight hours.
They stick too well. All three bond to smooth polyetherimide sheets hard enough to tear the plate. A textured surface or a glue-stick release layer is the standard defence, and the inversion catches people out: glue stick is preventing adhesion here rather than helping it.
They droop. The melt is viscous and slow to set, so unsupported spans sag where PLA holds a line. Design around it rather than tuning around it.
They are denser than PLA, running 1.22 to 1.40 g/cm³, so a spool holds less filament and a given model weighs more. Any estimate produced from a filament profile that was set up for PLA and never revisited will come out low across this whole shelf.
The one setting people get wrong
Nozzle 230 to 275 °C depending on member, bed 70 to 90 °C, enclosure helpful rather than required. The one setting that separates a good copolyester print from a poor one is part cooling: run it around half, not full. These polymers bond between layers by chain diffusion across the interface, and a hard fan freezes each bead before that has happened.
The other family habit is stringing, and the reflex to attack it through retraction settings is usually wrong. Dry the spool first and re-run the same print before changing a single value.
The decision, in one line each
Functional part, outdoors or indoors, below 70 °C? PETG, and stop reading. Clear part that gets bent or knocked and must not go white at the fold? PCTG. Turning bottles into filament, or you need PET's gas-barrier and chemical behaviour specifically? PET, and expect it to be a genuine skill.
For a stiffer, flatter, less stringy version of the middle option, the fibre-filled grade lives with the composites rather than here.