Use the free tool below: your data never leaves your browser — all work happens on your device, and nothing is uploaded to any server.
Why printed parts come out smaller than drawn
Almost every 3D printing material shrinks as it cools from printing temperature to room temperature. The plastic leaves the nozzle hot and expanded, then contracts as it cools, and the finished part ends up slightly smaller than the model. For a display piece, a one or two percent difference is invisible. For a part that must fit a bearing, slide into a slot, or match a mating piece, it is the difference between working and useless. Shrinkage compensation fixes this by printing the model slightly larger, so that after shrinkage it lands exactly on the target size.
The math is simple once the shrinkage percentage is known. If a material shrinks two percent, a printed dimension equals the model dimension times 0.98. To end up at 100 mm, the model must be printed at 100 divided by 0.98, which is about 102.04 mm, a scale factor of 102.04 percent. The calculator does this division for any target dimension and shrinkage percentage. The hard part is not the math: it is knowing the true shrinkage percentage for your material, printer, and settings.
How to measure your own shrinkage
Print a test piece of known size, such as a 100 mm bar or a 50 mm cube, in the material and settings you plan to use. Let it cool completely to room temperature, ideally overnight, because some materials keep shrinking slightly for hours. Measure it with calipers at several points and average the readings. If a 100 mm bar measures 98.5 mm, the shrinkage is 1.5 percent. That measured number beats any published figure, because shrinkage depends on printing temperature, cooling, infill, and even the shape of the part, not just the material name.
Measure in more than one direction. Many printers shrink differently along X and Y than along Z, because layers cool differently in the vertical direction. If your bar measures 98.5 mm along X but 99.2 mm along Z, you need different compensation per axis, which most slicers support through separate X, Y, and Z scale factors. The calculator gives a single-axis figure, so run it once per axis with that axis’s measured shrinkage when the difference matters.
A worked example
Suppose you need a gear with a finished outer diameter of 60 mm, printed in a material that shrinks 1.5 percent on your printer. The scale factor is 100 divided by 98.5, which is 101.52 percent, and the printed diameter must be 60.91 mm. You scale the model to 101.52 percent in the slicer, print it, and measure 60.0 mm after cooling. The gear now meshes correctly with its partner, where the uncompensated version would have been nearly a millimetre too small and would have rattled or slipped.
Consider a larger part: a 200 mm enclosure panel in ABS, which shrinks around 2 percent on many printers. Without compensation the panel comes out at 196 mm, four millimetres short, and it will not cover its opening. With compensation at 102.04 percent, the printed panel is 204.08 mm and shrinks back to 200 mm. Four millimetres of error is not a fine detail issue: it is a failed part, and it is entirely avoidable with one measurement and one division.
What shrinkage compensation cannot fix
Compensation fixes uniform shrinkage, where the whole part shrinks evenly. It does not fix warping, where corners lift off the bed and the part distorts unevenly. Warping is a bed adhesion and temperature problem, solved with a heated bed, an enclosure, brims, and appropriate temperatures, not with scaling. Scaling a warped part just gives you a larger warped part. Similarly, compensation does not fix holes that print small for a different reason, such as over-extrusion or the way slicers trace circles. Hole size deserves its own test and its own correction.
Also note that shrinkage compensation interacts with tolerances. If you scale a part up by two percent, every hole in it also grows by two percent before shrinking back, which is exactly what you want. But if you then add a separate hole-size correction on top, you may double-count. Apply compensation first, print a test, measure the actual holes, and only then add any further correction the measurements justify.
Frequently asked questions
Does PLA shrink much? PLA shrinks relatively little, often under one percent, which is why it is forgiving for dimensional work. ABS, nylon, and some engineering materials shrink more and need compensation.
Should I compensate in the model or the slicer? The slicer is better, because the original model stays at true size and the compensation stays visible as a slicer setting you can change per material.
Does infill affect shrinkage? Yes. Higher infill and thicker walls generally shrink slightly more, because there is more hot plastic contracting. Measure with the settings you will actually use.
These figures are estimates and general guidance only, not a guaranteed print result. Your material manufacturer’s datasheet and your printer’s own calibration govern every setting. Use the calculator above with shrinkage measured from your own test piece, and verify the finished part with calipers.
Related calculators
- Spanish version
- Resin Exposure Time Calculator – Adjusted Exposure for a New Layer Height
- Model Scaling Calculator – Exact Scale Percentage to a Target Dimension