3D printing tolerances: test the interface, not a magic number

Tolerance and clearance are related, but they are not the same thing

Tolerance describes the acceptable variation around a target dimension. Clearance is the intentional space designed between mating surfaces. Your CAD or generator can define the intended clearance, while the printer, material and slicer determine how closely the finished part reproduces it.

This distinction matters because adding “0.2 mm tolerance” does not say where the allowance is applied. A useful design states which surface moves, whether the value is per side or across the complete diameter, and what kind of fit the part needs.

Name the clearance before choosing the number

If a 20.00 mm shaft needs 0.20 mm of radial clearance and the shaft stays nominal, the matching circular bore is 20.40 mm. The 0.20 mm allowance exists on both sides of the shaft, so the diametral difference is 0.40 mm.

That formula defines the intended geometry; it does not promise the printer will produce the exact measured bore. Hole shrinkage, first-layer expansion and extrusion behaviour still need to be tested.

Why one 3D-printing tolerance does not fit every job

Fit can change when any part of the print system changes. The main variables include:

• printer motion, extrusion calibration and maintenance;
• material family, brand, pigment, moisture and cooling behaviour;
• nozzle diameter, line width and layer height;
• part orientation and whether a mating surface is vertical, horizontal or bridged;
• first-layer expansion, seam placement and support contact;
• slicer version, profile and compensation settings; and
• the intended fit: free-running, sliding, locating, press, threaded or deliberately loose.

Even two interfaces on the same part may deserve different experiments. A shaft bore, a printed thread and a Gridfinity base do not share the same geometry or failure mode, so a single printer-wide “tolerance” is usually too blunt.

A repeatable coupon-first workflow

1. Measure the real mating part. Use the actual shaft, hole, container, panel or receiver—not only a nominal catalogue size.
2. Choose the fit you need. State whether the part should slide, locate, rotate, thread together or remain removable.
3. Name the allowance. Record radial, diametral or per-side planar clearance so it cannot be applied twice.
4. Print a small trial set. Compare at least a tighter, selected and looser variant around a reasonable starting value.
5. Keep the production setup consistent. Use the intended material, nozzle, layer height, orientation and slicer profile.
6. Test the real behaviour. Check assembly, release, rotation or engagement as appropriate—without inferring strength or safety.
7. Record the result. Save the selected clearance with the print setup and use it only where the interface is comparable.

A coupon is useful because it is fast and cheap to reprint. It isolates the fit question from the time, filament and extra geometry of the final part.

For trial design, physical markers and recording fields, continue with the complete fit-coupon calibration guide.

Two examples from editable EazyMake templates

Gridfinity base in a receiving plate

The 3 × 2 Gridfinity baseplate starter begins with a 0.25 mm per-side planar clearance and provides a separate fit pack. The starting value makes the experiment concrete; the printed trial establishes which option works on the recorded setup.

The broader Gridfinity generator also exposes drawer-fit layouts, bins and supported feature-specific trials.

Backlash and bore fit in a spur gear pair

The 2:1 spur gear pair starts with 0.20 mm backlash, but the shaft bores are separate mating interfaces. The gears may need one trial for tooth engagement and another for the intended bore fit.

See the spur-gear generator for the shared geometry and backlash workflow behind that starter.

Neither result proves torque capacity, wear life or application safety. Those require a separate test plan for the final mechanism.

Record enough information to reproduce the fit

A useful fit result should identify:

• the interface and whether the clearance is radial, diametral or planar;
• the measured mating dimension;
• printer model and relevant calibration state;
• material, including grade or brand when behaviour matters;
• nozzle diameter and layer height;
• slicer and profile identity;
• part orientation;
• the tested clearance variants; and
• the observed physical result.

This turns “0.2 worked for me” into evidence that can be reused thoughtfully. If the print setup or interface changes materially, treat the old result as a starting point and run the small test again.