Skip to content
3D PRINTEROn Demand

FDM Dimensional Accuracy by Part Size

Dimensional accuracy is how close a finished printed dimension lands to the number in the CAD file, and published FDM specs state it differently depending on how big that dimension is.

Does a printed part's accuracy change with its size?

Yes: Stratasys publishes FDM accuracy as a fixed floor or a percent of the dimension, whichever is greater, so a small feature is governed by the fixed floor and a larger one is governed by the percentage instead. This percent-plus-floor structure is the size relationship, not a separate rule layered on top of it.

Where is the crossover between the floor and the percentage?

Arithmetic on Stratasys's own published floor and percentage puts the crossover at about 100 mm for the F123 class and about 85 mm for the Fortus 450mc, the dimension where the fixed floor and the percentage figure equal each other. Neither crossover number is stated directly on either spec sheet; both come from dividing that machine's published floor by its published percentage.

MachineFloor (mm)Percent of dimensionCrossover, floor divided by percent (mm)Provenance
F123 class (F370)0.2000.2 percent (0.002 mm per mm)about 100Stratasys F370 spec sheet, arithmetic on the published figures
Fortus 450mc0.1270.15 percent (0.0015 mm per mm)about 85Stratasys Fortus 450mc spec sheet, arithmetic on the published figures
Crossover point where the fixed floor equals the percentage figure, arithmetic on each machine's own published spec

Does a real study confirm small features drift further from nominal?

Yes: Wei and coauthors, printing PLA gears, measured a dimensional deviation of minus 0.045 to plus 0.060 mm on the main gear and a wider minus 0.150 to plus 0.078 mm on the smaller pinion, in the same study and on the same part.

FeatureDeviation range (mm)Provenance
Main gear-0.045 to +0.060Polymers 2025, PMC11820634
Pinion (smaller feature)-0.150 to +0.078Polymers 2025, PMC11820634
Wei and coauthors, Polymers 2025: measured dimensional deviation by gear size

Does a filament datasheet tell you how much a part will shrink?

No: Prusament's ASA technical data sheet, read in full, carries no shrinkage percentage row, and Stratasys treats its own shrink rates as proprietary information rather than publishing them.

That absence is itself the finding, not a gap in this page's research. If a shrinkage number for your exact material and printer existed in a datasheet, it would be cited here with the same rigor as every other figure on this page.

What this page does not tell you

  • This page does not give you one accuracy number for your specific part. Geometry, wall thickness, orientation and cooling all change what a printer actually holds on a given feature.
  • The crossover sizes are arithmetic performed on two published Stratasys specs, not a number either spec states directly, and they describe those two machines only, not FDM printing in general.
  • This page does not give you a shrinkage percentage for your material and printer. No filament datasheet checked here publishes one, and this page does not fill that gap with an unattributed number.

When to print a test coupon instead

Print a small test coupon at the actual size of the feature you care about before committing a full run, especially near or below a machine's fixed accuracy floor. Send the file through a review request and tell us which dimension worries you. A person reviews your request and follows up by email.

Sources

  • F370 Spec Sheet, Stratasys

    States printed dimensional accuracy of 0.200 mm or 0.002 mm per mm, whichever is greater. Footnote states accuracy is geometry dependent, derived from statistical data at 95 percent dimensional yield, and Z accuracy carries an additional tolerance of -0.000/+slice height.

  • Fortus 450mc Spec Sheet, Stratasys

    States printed dimensional accuracy of 0.127 mm or 0.0015 mm per mm, whichever is greater.

  • Polymers 2025 (PMC11820634), MDPI Polymers, PMC open access

    PLA gear study. Measured a printed bore 0.345 mm undersize of nominal in one test setup. Separately measured dimensional deviation of minus 0.045 to plus 0.060 mm on the main gear and minus 0.150 to plus 0.078 mm on the smaller pinion, evidence that a small feature drifts further from nominal than a large one on the same part.

  • ASA Prusament Technical Data Sheet (2022), Prusament

    Full filament technical data sheet, read in full. Carries no shrinkage percentage row.

Related

Published 5 September 2026.