3D printing tolerances: make parts that fit

1. Separate clearance from dimensional error

Clearance is the intended space between mating surfaces. A tolerance is the permitted variation around a specified dimension. Hobby discussions often use “tolerance” for both, so write down exactly which number you mean.

A nominal 20 mm peg and nominal 20 mm hole have zero designed diametral clearance. Expecting them to slide together assumes both finished surfaces match their drawings perfectly. Prusa’s modeling guide explains why fit changes with geometry, orientation, material, and calibration; it does not establish a universal clearance.

First decide the behavior you want: easy removal, a sliding lid, or a snug locating feature. “It fits” is too vague to compare two test pieces.

2. State whether a gap is per side or total

A note saying only “0.2 mm clearance” could describe two different designs. Label the dimension on the drawing or filename. For a rectangular insert, calculate width and depth independently; a fit along one axis does not guarantee a fit along the other.

These numbers illustrate the arithmetic. Choose test values around the needs of your own project rather than treating this example as a recommended fit.

3. Measure the surfaces that actually meet

Identify the contact area before reaching for a caliper. A lid may locate on an inner lip while its decorative outer wall never touches the box. Measuring the outside of the lid alone would miss the relevant dimension.

Measure the mating object more than once and at several positions. Note tapers, corner radii, seams, and protrusions. If a recess is narrower at the bottom, its opening is an incomplete measurement. Keep the part and instrument in the same orientation when comparing repeats.

Record the nominal model size separately from the cooled print size. The slicer measures geometry; it cannot certify the manufactured fit. Prusa’s measurement tool documentation describes measuring model features and also notes limitations with overlapping parts.

4. Make a fit coupon that represents the joint

Instead of printing an entire organizer, create a short section containing its lip and mating recess. Preserve the intended wall thickness, corner shape, print orientation, and contact length as far as practical. A random peg test is less informative for a long sliding lid.

1. Choose a small set of clearly labeled clearance variants.
2. Print them with the material, nozzle, and profile planned for the final object.
3. Let them cool and remove only the finishing material you would also remove on the final part.
4. Compare insertion, movement, removal, and visible rubbing.
5. Record the chosen variant and print a representative section again before scaling up the job.

A short coupon cannot reproduce every behavior of a long part. Treat it as a way to reject unsuitable fits cheaply, then verify the full contact length.

5. Fix the source of the mismatch

If the part catches only at its bed-facing edge, inspect for elephant’s foot. If one corner lifts, investigate warping. Neither symptom calls for automatically shrinking the entire model.

If the model itself is the wrong size in the slicer, check units and scale first. If the model is correct but every test varies unpredictably, establish a repeatable print before fine-tuning the design.

A global scale change also changes hole spacing, wall thickness, and other features. Prefer editing the specific mating dimension when you control the design. Save the original and name each revision so a later fit improvement can be traced to a real change.

6. Carry the successful test into the final print

Keep the same orientation and profile for the validated feature. If you rotate the full model to save supports, repeat the fit check in that orientation. Our orientation guide explains how the surfaces and geometry presented to the printer change.

Use the model-fit calculator for a separate question: whether the overall dimensions fit the build volume. It does not predict mating clearance or include supports and brims.

Save the test label, model revision, material, measured dimensions, and fit judgment in the print log. A useful result reads “variant B slides by hand after cooling,” alongside dimensions, rather than “printer tolerance is 0.2 mm.” That record is specific enough to reuse and honest about what was tested.