Every resin printer owner eventually measures a finished part and finds it slightly smaller than the model that went in. The instinct is to call it a defect. It is not. It is arithmetic, and once you understand the terms you can compensate for it deliberately instead of chasing it with time-consuming trial and error.
The Three Shrinking Stages
A photopolymer part loses dimension at three distinct points in the process, and conflating them is the most common reason people misdiagnose drift.
| Stage | What happens | Typical order of magnitude |
|---|---|---|
| During exposure | Polymerisation pulls molecules together as bonds form | Small, but cumulative across layers |
| During washing | Uncured resin leaves; thin walls can soften and sag | Highly geometry-dependent |
| During post-curing | Remaining conversion completes; the largest dimensional change | Often the dominant term |
The practical takeaway: a part measured straight off the plate is not finished shrinking. Measuring before post-cure produces an optimistic number that will not survive the full process.
How to Measure Your Own Shrinkage Rate
Generic shrinkage figures are a starting point, not a specification. The actual rate depends on your resin batch, your exposure settings, your post-cure time, and the geometry of the part. The only reliable number is the one you measure.
- Print a calibration coupon with a long, known dimension. A 100mm bar is better than a 20mm cube, because it amplifies the error into a readable value. Add a second axis if your priorities are dimensional across a face.
- Process it exactly as you process production parts. Same wash time, same post-cure duration and temperature. The coupon is worthless if its process differs.
- Measure after at least 24 hours of rest. Post-cure continues to resolve for a period afterwards. Measure when the part has stabilised, not when it felt dry.
- Compute a percentage, not a millimetre. If a 100mm bar finishes at 98.4mm, your rate is 1.6%. That percentage is what transfers to other models.
- Repeat across three prints. One coupon can hide process noise. Three coupons reveal whether your number is stable enough to design around.
Compensating in CAD Rather Than in Slicer Software
Scaling the whole model is the tempting shortcut and it is almost always wrong. Uniform scale expands the part in every direction including the ones that were already correct — and it does not respect the fact that shrinkage is directional and geometry-dependent.
- Compensate the dimension that drifted. If the drift is along one axis, scale that axis. If a wall thickened, adjust the wall, not the model.
- Leave features that must fit others alone. Threads, press-fit bores and mating surfaces usually need their own allowance rather than global scaling.
- Design in a shrink allowance from the start. Once your rate is measured and stable, build it into the model so the printed part lands on nominal without a second operation.
- Re-check after any process change. A new resin batch, or a change to post-cure time, invalidates your measured rate. Re-run the coupon.
Settings That Tighten Tolerance
Shrinkage cannot be eliminated, but its variability can be reduced — and variability is what actually breaks tolerance targets.
| Setting | Effect on consistency | Practical adjustment |
|---|---|---|
| Exposure time | Under-exposure leaves unconverted resin that shrinks late | Dial to the resin's tested band, then hold it constant |
| Post-cure duration | Varying duration means varying final dimension | Fix the time and temperature, and log both |
| Wash time | Over-washing softens detail and thins walls | Use a timed wash, not a repeat-until-clean habit |
| Orientation | Supports and drainage affect local deformation | Keep production orientation consistent across batches |
| Resin batch | Formulation shifts between lots | Re-measure when you open a new bottle |
The pattern running through all of this is repetition. Shrinkage is not a mystery to be solved once; it is a parameter to be characterised. A shop that measures its rate, fixes its process, and designs the allowance into the model gets dimensionally reliable parts without thinking about it every time.
Where to Start
Pick a resin that exposes predictably before you optimise anything else — a generalist grade with documented settings gives you a stable baseline to measure against. The ANTINSKY Standard Lite Resin 1kg is a sensible calibration material, and ANTINSKY Standard Tough Resin covers parts that need to survive handling without chipping. Print the 100mm coupon, process it the way you process real work, and write your number down. Everything downstream gets easier from there.