Do 3D printed parts shrink? While cooling, and over time

Yes. 3D printed parts shrink as they cool from nozzle temperature to room temperature, by less than 1% in X and Y for PLA, PETG and ABS in a printed-part study. Once cooled, a part keeps its size unless heat, moisture or a steady load acts on it.

How much do PLA, PETG and ABS shrink after printing?

Weiser et al. at Binghamton University (2021) printed parts in PLA (polylactic acid), PETG (glycol-modified polyester) and ABS (acrylonitrile butadiene styrene) on a QIDI X-One2 and measured them with calipers. Bambu Lab’s wiki names the cause as thermal contraction while the plastic cools, most noticeable in X and Y.

Material	Shrinkage in X and Y	Z	Heat limit from the maker’s data sheet
PLA	0.35 to 0.63%	0.80%	Glass transition 60 °C
PETG	0.54 to 0.76%	0.73%	Glass transition 69 °C
ABS	0.59 to 0.69%	grew 0.92%	Heat deflection 87 °C at 0.45 MPa

The heat limits come from Bambu Lab’s PLA Basic, PETG Basic and ABS data sheets. On a 100 mm part, 0.6% shrinkage is 0.6 mm. Your printer, settings and spool will give different numbers. Print a test block, measure it and enter the result in the shrinkage compensation calculator to get the scale or slicer value.

How long after printing does a part keep shrinking?

Cooling shrinkage ends when the whole part reaches room temperature. None of the data sheets or slicer guides cited here gives a time, since it varies with wall thickness, bed temperature and enclosure. Bambu Lab’s test procedure says only to measure after the part cools to room temperature.

Measuring too early reads large. ABS expands 80 to 110 millionths of its length per °C, according to INEOS Styrolution’s data for its Terluran GP-35 ABS resin. A 100 mm ABS part measured 10 °C above room temperature reads 0.08 to 0.11 mm long. That is up to a fifth of the shrinkage you are trying to measure. Leave the part and the calipers in the same room until the part feels no warmer than the bench, then measure.

Do 3D printed parts shrink over time?

Not at room temperature, in any way that affects a fit. A cooled PLA, PETG or ABS part holds its size until something acts on it: heat near its glass transition (60 °C for Bambu Lab’s PLA Basic), annealing or a constant load that slowly stretches it. Nylon is the exception. It absorbs water from the air and grows.

Heat near the glass transition

Above the glass transition the plastic softens, and stresses frozen in during printing can relax and move the part. Bambu Lab’s data sheets put PLA Basic’s glass transition at 60 °C and its heat deflection temperature at 57 °C under a 0.45 MPa load. PETG Basic’s figures are 69 °C and 71 °C. A PLA part left in a hot car or next to a heater can warp or shrink long before it melts; how PLA behaves in heat has the temperatures.

Annealing does this on purpose. In Prusa’s annealing tests, PLA, PETG, ABS and ASA samples shrank along their longest dimension and grew in Z.

Constant load: creep

A part under a steady load stretches slowly. This is creep, not shrinkage, and heat speeds it up. Dogan (2022) loaded printed specimens of six materials for three hours at 25, 40 and 60 °C. Creep rose with temperature and stress in every material. PLA had the worst creep resistance, and the author recommends using PLA parts at room temperature with no load or only very low static loads. Polycarbonate crept least. A PLA hook or shelf bracket that sags over months is creeping.

Do 3D prints shrink or expand after printing?

Both, for different reasons. Every print shrinks once while it cools. After that, parts expand and contract reversibly with temperature, by about 0.1 mm per 100 mm for a 10 °C change in ABS. Nylon parts also grow as they absorb water from the air, and annealing shrinks a part along its length and grows it in Z.

Temperature swings are reversible

Thermal expansion goes back when the temperature does. BASF’s data sheet gives 102 millionths per °C along the flow direction for Ultramid B3S, an unfilled nylon 6, close to the ABS figure above. It matters when a plastic part has to fit a metal one across a wide temperature range, or when you measure a part that is still warm.

Nylon swells with moisture

BASF’s Ultramid brochure says nylon 6 and 6/6 (PA6, PA66) grow 0.2 to 0.3% in length for each weight percent of water absorbed. Its B3S data sheet gives 3% water at 23 °C and 50% relative humidity. Multiplied, that is roughly 0.6 to 0.9%, about as much as the part shrank on the printer.

Polymaker measured this on printed 40 mm cubes. They shrank slightly when annealed, and then most grew past their original size after moisture conditioning; Fiberon PA12-CF10, a carbon-filled nylon 12, was the exception. Polymaker gives no standard percentage, because the change depends on geometry, size and infill. Measure a nylon part in its service state: fresh and dry, or conditioned to room air. The guide to dimensionally accurate nylon parts covers drying, conditioning and the combined correction.

Does PLA shrink over time?

No. PLA shrinks while it cools and then stays put at room temperature. It changes size later only if it is heated toward 55 to 60 °C or carries a sustained load. Its glass transition is the lowest of the three filaments in the table, and it had the worst creep resistance of the six materials in Dogan’s test.

Does PETG shrink over time?

No. PETG shrinks 0.54 to 0.76% in X and Y while cooling and holds that size afterwards. With a glass transition of 69 °C in Bambu Lab’s PETG Basic data sheet, it tolerates warmer places than PLA before it starts to soften and move.

For a full routine that separates shrinkage from flow and axis errors, see how to calibrate dimensional accuracy per material.