When two 3D printed parts do not fit, the problem is rarely solved by one universal tolerance number. Fit is the result of a system: machine condition, measurement method, extrusion behavior, material, orientation, geometry, first-layer effects, slicer compensation, and the kind of fit the design actually requires.
Start with metrology, not compensation
Before changing XY compensation, hole expansion, extrusion multiplier, or CAD dimensions, define what is being measured and how.
Record the CAD nominal dimension, measurement points, outside/inside/depth method, tool resolution, part temperature, and repeated readings. A single caliper reading is not a process-capability study. Measure the same feature the same way and look for repeatability before treating the number as truth.
Verify the printer's mechanical baseline
A loose belt, worn motion component, binding axis, unstable frame, or backlash symptom can look like a slicer-tolerance problem. Compensating in CAD for a mechanically unstable printer bakes the machine fault into the design.
Inspect the motion system, belts, rails, wheels, fasteners, homing behavior, and simple baseline geometry first. The sequence is important: restore repeatability before chasing absolute dimensional accuracy.
Validate the measurement tool
The caliper can create its own error. Clean the measuring faces, zero the tool, compare it with a known reference, avoid excessive jaw pressure on printed polymer, and repeat the reading.
You do not need a laboratory instrument for every functional print, but you do need a tool whose limits and repeatability you understand. Switching between instruments mid-calibration without recording it destroys useful comparison data.
There is no universal tolerance value
Prusa's current design guidance explicitly notes that there is no single universal tolerance because fit depends on factors including model size, orientation, geometry, calibration, settings, and material properties.
That matches practical experience. A sliding fit, hinge, snap fit, press fit, captive feature, fastener hole, and decorative alignment feature should not all use the same clearance.
Define the functional relationship first: should the parts move freely, locate precisely, grip through friction, flex into place, or never touch? Then choose a starting clearance appropriate to that job and validate it on the actual process.
Material and orientation change the answer
Shrinkage and warping vary by material. Hole geometry can print differently from outside dimensions. Vertical holes, horizontal holes, thin walls, sharp corners, and long unsupported dimensions can behave differently even on the same printer.
Record the material, nozzle, layer height, orientation, temperature, cooling strategy, slicer version, and compensation settings with the test result. A tolerance validated for one configuration should not be assumed to transfer unchanged to another.
First-layer geometry can break an otherwise correct fit
A squished first layer can create an outward lip commonly called elephant foot. On a tight-fitting part, that small edge can prevent assembly even when the upper geometry is dimensionally acceptable.
PrusaSlicer includes elephant-foot compensation specifically to reduce this effect. The correct response is to diagnose whether the interference is localized to the first layer before applying a global compensation that changes the whole part.
Use fit coupons instead of repeatedly reprinting the full part
A calibration coupon can test several hole sizes, shaft clearances, wall dimensions, or mating features in one small print. This makes iteration faster and produces a reusable record.
Create a fit ladder around the intended nominal dimension, print it with the target material and orientation, measure the result, identify the best functional fit, then transfer that evidence into the production design.
A repeatable calibration sequence
- Define the fit: sliding, moving, locating, friction, press, snap, or cosmetic.
- Verify mechanics: make sure the printer is repeatable before compensating.
- Validate the tool: confirm the measurement method and instrument.
- Print a reference: use controlled geometry or a fit ladder.
- Measure repeatedly: separate random variation from consistent offset.
- Classify the error: outside dimension, hole, first layer, material shrinkage, or geometry-specific behavior.
- Adjust one control: CAD clearance, slicer compensation, process setting, or hardware condition.
- Reprint and verify: confirm the destination fit rather than assuming the change worked.
- Record the process: preserve the validated printer/material/settings combination.
Current technical guidance worth checking
- Prusa Knowledge Base — Modeling with 3D printing in mind
- Prusa Knowledge Base — Elephant foot compensation
Use the complete operating system
3D Print Dimensional Accuracy & Tolerance Calibration System™ turns the measurement-first approach into a 20-chapter professional reference with calibration checklists, decision worksheets, fit scenarios, implementation standards, process logs, and validation controls for functional printed parts.
Explore 3D Print Dimensional Accuracy & Tolerance Calibration System™
Related resources
- 3D Print Dimensional Accuracy & Tolerance Calibration System™
- Bambu Lab 3D Printing™: Build a More Reliable Workflow From First Layer to Finished Part
- Formlabs Professional 3D Printing™: Resin Success Is an End-to-End Workflow
- Inside The Repost Reach Loop: A Practical Guide
- Explore the Mindset Journal topic guides