Does 3D printed resin shrink? How much, and how to compensate

Yes. Resin shrinks as it cures. Makers list 3 to 6% shrinkage by volume, and printed parts measured in X and Y usually come out 0.3 to 0.6% small, so a 100 mm part lands near 99.5 mm.

How much does 3D printing resin shrink?

Common hobby resins shrink 3 to 6% by volume, according to their makers: 3 to 4% for Anycubic 16K Standard, 4.3 to 5.6% for Anycubic Standard Resin V2, and about 6% for most Siraya Tech resins. Measured on printed parts, the linear loss in X and Y is smaller, typically 0.3 to 0.6%.

Resin	Published shrinkage	Type
Anycubic 16K Standard Resin	3 to 4%	volume
Anycubic Standard Resin V2	4.3 to 5.6%	volume
Anycubic ABS-Like Resin 2.0	3.9 to 4.9%	volume
Siraya Tech resins (most)	about 6% (1.5 to 2% per axis)	volume, fully cured
Siraya Tech Sculpt Ultra	about 2.5%	volume, fully cured

Elegoo’s ABS-Like Resin page claims “low volume shrinkage and linear shrinkage” but gives no figure. No published figure turned up from Sunlu or Phrozen. Anycubic’s and Siraya Tech’s pages don’t say how they measured.

Printed-part data comes from Jan Mrázek, who measured resins with a 100 mm test bar (printer not stated): Siraya Tech Fast Gray 0.48%, Fast White 0.46%, Fast Navy Gray 0.34%, Fast Mecha 0.21%, QTS Strong 0.32% and QTS Model 0.90%. He found tough resins shrank more, and composite resins much less.

Volume shrinkage vs linear shrinkage

Volume shrinkage is how much the cured solid loses in volume. Linear shrinkage is how much shorter a dimension gets, the number your calipers see. If a resin shrinks evenly in every direction:

linear = 1 - (1 - volume)^(1/3)

6% volume  -> 1 - 0.94^(1/3)  = 2.0% linear
4% volume  -> 1 - 0.96^(1/3)  = 1.35% linear
1.8% volume -> 1 - 0.982^(1/3) = 0.6% linear

For small values that is close to a third. Mrázek pairs 0.3 to 0.6% linear with 0.9 to 1.8% by volume, and Siraya Tech pairs its 6% with 1.5 to 2% per axis.

Printed parts measure well under the converted data sheet number: 0.3 to 0.6% measured, against 1 to 2% from the published volumes. The data sheet figure isn’t a printed-part measurement, and a printed part doesn’t shrink evenly. Mrázek notes that Z shrinkage never reaches the part, because each new layer fills the space the cured one gave up. Scaling by the converted data sheet figure would over-size the part, so measure instead.

Why does resin shrink, and when?

Resin shrinks because curing turns loose molecules into one tightly bonded solid. Liqcreate explains that weak Van der Waals contacts between monomer units are replaced by short covalent bonds. The cured part takes up less space than the liquid it came from.

It happens in two stages:

• In the printer, as each layer cures. This shows up in X and Y; Z is refilled layer by layer.
• During and after post-cure. Formlabs says post-curing “also causes minor shrinkage” and that its print settings already account for both stages.

Mrázek found parts kept shrinking for a couple of days, and curing at 50 °C cut that to about an hour. In a later reply he said Siraya Tech Fast kept shrinking for about 3 months after final cure, fading over time.

No maker publishes how its total splits between printing and post-cure. To find it for your resin, measure the part after washing, then again after curing and a few days’ rest.

Does ABS-like resin shrink?

Yes. ABS-like resin is a photopolymer, not ABS. Formlabs describes it as “a photopolymer resin with a UV acrylate base and added toughening agents meant to mimic ABS properties”, while ABS itself is a thermoplastic. Sunlu’s guide calls ABS-like a category of UV-curing photopolymers engineered to replicate ABS. Shrinkage figures for filament ABS don’t apply.

Anycubic lists its ABS-Like Resin 2.0 at 3.9 to 4.9% by volume, between its 16K Standard (3 to 4%) and Standard V2 (4.3 to 5.6%). So “ABS-like” isn’t a low-shrinkage label on its own. Mrázek’s finding that tough resins tended to shrink more is a reason to test before trusting a dimension.

How to compensate for resin shrinkage

Compensation takes two corrections: a percentage scale for shrinkage and a fixed offset for exposure bleed. Bleed is light spill at the layer edges that makes outer walls fat and holes tight. Shrinkage error grows with size, while bleed error stays constant.

1. Print a test part at 100% scale with compensation off. Use outer and inner features at several lengths. Mrázek’s bar has pairs 20, 50 and 100 mm apart. Formlabs’ method also uses several X and Y dimensions.
2. Wash, post-cure on your normal schedule, and let it settle before measuring. Use the same cure every time, since a different cure gives a different size.
3. Measure outside and inside dimensions. If outer features are long and holes short by the same amount, that is bleed, so use the offset. If the error grows with length, that is shrinkage, so use the scale.
4. Turn shrinkage into a scale: scale = 1 / (1 - s). For 0.48%, 1 / 0.9952 = 100.48%. The shrinkage compensation calculator does this from your designed and measured sizes.
5. Scale only in the printer’s X and Y, not Z. Mrázek warns that Chitubox scales a tilted model in its original orientation, so tilt it elsewhere first if it prints at an angle.

Where the settings are:

• Chitubox added tolerance compensation and bottom tolerance compensation (a fixed offset) in version 1.9.4, and a shrinkage scaling percentage in 1.9.5. Liqcreate’s test with its Flame Retardant HDT resin on a Saturn 3 settled on 100.65% scaling plus a 0.015 mm offset, for an average deviation of 0.017 mm on 1 to 24 mm blocks. In an earlier test with its ESD resin on an Elegoo Mars 3, uncompensated parts came out 117 to 192 µm short outside, with holes 82 to 185 µm narrow.
• PreForm (Formlabs) has X, Y and Z correction factors and an Outer Boundary Offset in the Print Settings Editor. Formlabs’ method fits a slope (the scale) and intercept (the offset) from several measured dimensions.

Run the test for each resin, even each color, because Mrázek found they differ. He also saw print temperature move shrinkage by roughly 0.1 percentage points.