Fits, Clearances, and Tolerances for Customized 3D Prints

A printed hinge makes fit errors obvious. Too little clearance fuses or binds the pin and knuckles. Too much clearance creates wobble and poor alignment. A pin that fits one material and orientation may fail after the production setup changes.

This article uses the parametric-hinge project as its running example. One exposed gap variable controls both rotational pin clearance and axial spacing between adjacent knuckles, while derived expressions keep pin size, knuckle width, and knuckle pitch synchronized.

Beyond Sliders: Designing Safe Parameter Spaces for Printable Products explains how to constrain related inputs. Here, the focus is physical evidence: calibrate a hinge under the production conditions you will use, select a passing clearance band, and record it as part of the production profile.

“Make it fit” is not a complete hinge requirement. Define how the joint should assemble and move.

The project’s print-in-place relationship is a clearance fit. The joint must rotate without allowing excessive radial or axial play. Choose that behavior before choosing a numeric gap.

Four related ideas are easy to blur together:

• Nominal size is the intended reference dimension in the design.
• Tolerance is the permitted or expected range after production.
• Allowance is the intentional dimensional difference that creates the desired fit.
• Clearance is the resulting positive space between produced mating surfaces. Negative clearance represents interference.

For a pin and bore, the modeled diametral clearance is

In the hinge project, gap is the modeled diametral clearance and the axial spacing between neighboring knuckles. State whether a documented value is diametral or per side: a 0.40 mm diametral clearance provides 0.20 mm of radial space on each side when the parts are centered.

Measure the printed pin and bore separately because they often have different dimensional bias. Let

and

The achieved hinge clearance is

To target a known printed clearance, model

This is why one global scale factor is rarely enough. An outside pin can print oversize while a bore prints undersize. Record both errors and calibrate the gap value under the production conditions that will be used for the hinge.

A hinge clearance belongs to a production profile, not to a printer model in the abstract. Record the printer, nozzle or imaging setup, material and conditioning, layer height, line width or exposure profile, hinge orientation, support strategy, cooling, slicer compensation, and post-processing.

Orientation changes bore shape, seam placement, first-layer behavior, cleanup, and strength. A horizontal bore may sag, a seam can create a local bind, and first-layer expansion can close axial gaps. Lock one production orientation and calibrate the same support and finishing state used for the finished hinge. Treat another orientation as a different profile until it passes the same motion and strength checks.

Published guidance is useful for choosing a test range, but it is not a substitute for validation. Prusa’s modeling guidance notes that no universal tolerance applies across model size, orientation, geometry, calibration, settings, and material. Formlabs’ overview of engineering fits likewise ties successful tolerancing to the selected process and material. Use published numbers as starting points, then keep the measurements from your own hinge workflow.

Calibrate the hinge with the same machine, material, layer height, orientation, and finishing steps used for production.

Use the project’s gap range to generate a ladder of candidate clearances that crosses the expected passing value. Keep leafT, knuckleN, pin mode, and the production setup fixed during the first comparison. After printing and normal cleanup, record:

• whether the leaves separate and rotate without damage;
• whether the hinge moves through the required throw;
• radial wobble at the pin;
• axial movement between knuckles; and
• any fused region, sag, seam, or first-layer obstruction.

The selected gap should become part of a named, versioned Rig Cad production preset rather than a number the operator retypes from memory.

From hinge requirement to validated Rig Cad preset

Diagram: Hinge requirements lead to a locked production profile, a printed gap ladder, measured results, a selected passing gap band, a saved preset, and final verification.

flowchart TD
  S1["Define rotation and play requirements"] --> S2
  S2["Lock the production profile"] --> S3
  S3["Generate a hinge gap ladder"] --> S4
  S4["Print and finish the samples"] --> S5
  S5["Measure motion and play"] --> S6
  S6["Select the passing gap band"] --> S7
  S7["Save a versioned Rig Cad preset"] --> S8
  S8["Verify the finished hinge"]

A useful coupon should reproduce the hinge’s active surfaces and print orientation.

Label each candidate or keep an unambiguous map from sample to gap. Include pin and bore reference dimensions that can be measured. The goal is a small test that reproduces the hinge’s likely failures.

The project derives the pin, bore, and knuckle spacing from a small set of variables.

pinD     = autoPin ? leafT : pinCustom
pinHoleR = pinD / 2 + gap / 4
pinR     = pinD / 2 - gap / 4

The resulting diametral difference is exactly gap:

The same gap also spaces the knuckles:

knuckleW     = (leafH + gap) / knuckleN - gap
knucklePitch = knuckleW + gap

autoPin keeps the nominal pin diameter equal to leafT. Turn it off to expose pinCustom. hasPin selects the fused-pin assembly or a hole for a separate pin. These choices change the pin construction, while gap remains the calibrated fit input shared by the mating features.

One shared gap is appropriate only when the passing radial-clearance and axial-spacing bands overlap. If the hinge needs different values to control rotation and end play, split the relationship into separate radial and axial gap variables and calibrate each one explicitly.

The hinge’s product dimensions and manufacturing allowance have different meanings.

Changing hinge size should not silently change the calibrated production evidence. If a new leafT, knuckleN, pin mode, or orientation creates a substantially different joint, revalidate the relevant range before treating the old gap preset as production-ready.

One moving hinge proves possibility, not repeatability. Print several copies under normal production conditions and record the printed pin diameter, bore diameter, radial play, axial play, and any location-specific binding. Three to five copies are a practical starting check for a small workflow, not a statistical guarantee.

The worst-case radial clearance across observed samples is

If $C_{min}$ is negative, the observed bands overlap and some hinges can bind even when the average sample moves correctly. Also evaluate axial spacing: repeated knuckles can accumulate small dimensional errors along the hinge axis.

Formlabs describes this repeatability mindset in its Formtest reliability program: dimensional, positive, negative, structural, and functional features are tested repeatedly rather than inferred from one attractive print.

Define hinge acceptance as actions and measurements, not impressions.

Use the same cleanup state, motion cycle, reference points, and test method for calibration and production. A simple fixture or go/no-go gauge is more repeatable than asking whether the hinge “feels right.”

Dimensional correction can happen at several layers: machine calibration, slicer-wide compensation, Rig Cad expressions, or manual finishing. Stacking undocumented corrections makes the next hinge difficult to predict.

Use this hierarchy:

1. correct machine or material problems that affect every part;
2. use slicer compensation for a repeatable process-wide bias;
3. store the hinge’s validated gap in a named production preset; and
4. reserve drilling, reaming, sanding, or pin fitting for operations deliberately included in the work instructions.

Do not enlarge a hinge bore in CAD and also apply an undocumented hole-compensation setting for the same bias. Apply each correction once, record where it lives, and recheck the hinge preset when a global setting changes.

Route the hinge to recalibration or manual review when:

• the printer, material, nozzle, layer height, orientation, slicer compensation, or finishing differs from the validated profile;
• gap falls outside the tested passing band;
• leafT, knuckleN, pin mode, or engagement length leaves the validated family;
• a pin, bore, or knuckle dimension is outside its measured acceptance band;
• a repeated sample or go/no-go gauge fails;
• the hinge does not complete the required motion cycle;
• a finishing operation was not included in the product promise; or
• failure could create a safety or load-bearing risk.

A mesh can be valid while the hinge binds, wobbles, or breaks. Stopping production when the evidence no longer applies is part of the system.

The hinge separates product geometry from production evidence. Product variables define the leaves, knuckles, and pin mode; derived expressions construct their relationships; a calibrated gap records the allowance that produced acceptable motion under one named production profile.

Changing the product dimensions regenerates the model. Changing the printer, material, orientation, or fit requirement changes the evidence. When a configuration leaves the validated family, recalibrate instead of assuming the old clearance still applies.

Continue with Versioned Production: Presets, Reorders, and Change Control for Parametric Goods to preserve the project revision, hinge settings, production profile, artifact, and inspection result behind a successful joint.

The previous article, Beyond Sliders: Designing Safe Parameter Spaces for Printable Products, defines which combinations belong in the product family. Continue with Versioned Production: Presets, Reorders, and Change Control for Parametric Goods to preserve the project, production profile, artifact, and inspection evidence behind a successful fit. The Parametric Goods overview contains the full reading path.