Use measurements, not a material-shrinkage table
- Print at least two outside test lengths per axis you need, preferably at different sizes. For a hole, print at least two hole diameters along the relevant X/Y axis. Keep conditions comparable.
- Measure the finished part with suitable tools, away from elephant foot and visibly warped regions. Enter the designed and measured values in the same units.
- Fit the models, inspect the observed residuals, then solve the CAD dimension or fit. Prefer targets inside the measured size span; verify a second test print before committing to expensive parts.
Model and diagnostics
For two or more distinct designed sizes, this calculator fits printed = slope × CAD + offset by ordinary least squares for each axis/feature group. Inverse design is (desired printed − offset) ÷ slope. A single row uses printed / CAD as a proportional ratio and assumes offset zero. The maximum residual is the largest absolute difference between a measured row and its fitted line; it is not a prediction bound, confidence interval, or printer specification.
Results beyond the minimum and maximum calibrated CAD size are marked as extrapolations. A fit can still be wrong inside its range if prints warp or process settings change. No fixed shrinkage percentage is assigned by material.
Outside, holes, and first layers differ
A hole can print smaller even when outside dimensions are close to target. Global XY compensation or CAD scaling can move both feature classes; do not transfer a hole model to an outside dimension, or vice versa. First-layer squish (“elephant foot”) is a local effect; it is not a general X/Y/Z scale. Prusa’s first-layer compensation guide explains this distinction. UltiMaker’s hole-expansion note describes why a hole-specific offset can matter.
Prusa’s modeling and tolerances guidance emphasizes that fit depends on model size, orientation, calibration, print settings, and material. The clearance planner uses a signed diameter difference only; it cannot predict friction, surface finish, strength, or assembly force.
Frequently asked questions
Why do I need two different test sizes?
One sample cannot tell proportional shrinkage from a fixed-size perimeter or measurement offset. Two sizes identify a line, and three or more help reveal residual disagreement. Use multiple size ranges and repeat a print when fit is critical.
Is the displayed CAD percentage a slicer shrinkage setting?
No. It is the size needed for one requested dimension compared with your current CAD size. Slicers can use different scale and horizontal-expansion conventions, and the fitted offset may not be corrected by a uniform scale. Check the slicer’s own documentation and a test print.
Can I calibrate a hole using outside measurements?
No. The tool keeps them separate. A measured actual shaft can be entered directly, but solving a hole still requires a hole measurement on the selected axis.
Does a negative clearance guarantee a press fit?
No. It defines a target hole diameter smaller than the shaft’s modeled/measured diameter. Actual interference, strength, and insertion force depend on process, roughness, shape, orientation, and material; print and test a fit coupon.