3D Printed Gears: Materials, Fit and a Test Log

Define the job before choosing a filament

Write down the tooth counts, module, pressure angle, speed, expected torque, operating temperature and required running time. A hand-turned demonstration and a motor running all day are different jobs. Specify the shaft and housing too: gears cannot correct moving or misaligned axes.

The distinction between material and application is visible in igus’s printed-gear guidance, which uses application inputs and material-specific wear data. Its results for its own polymers are not a load rating for an arbitrary PLA or PETG print. Use the module guide to establish the dimensions first.

Material choice: compare requirements, not a winner

Prusa’s filament guide separates heat resistance, toughness, print difficulty and other properties. Choose a specific grade using its supplier data and a profile suitable for your printer. Polymer names alone do not establish gear wear, friction or lifespan.

PLA can be useful for checking shape and motion with a familiar printing process. PETG is another candidate for prototypes, but its toughness does not guarantee low wear. Nylon grades may be useful for moving parts; check drying, shrinkage and conditioning requirements. Filled, wear-optimized and unfilled polymers require their own data.

Slice the teeth, root and bore before printing

Use a stable printer/material profile, then inspect the toolpaths. Prusa’s design guidance explains why orientation and feature dimensions affect the print. For a spur gear, placing one flat face on the bed is a useful candidate orientation. Check other orientations where the hub or loading makes that necessary.

Look for continuous paths through each tooth root and sufficient solid material around the bore. Wall count and infill work together; neither one has a universal strength threshold. A fine layer height can improve the representation of sloped features, but it does not by itself establish accuracy or strength.

Keep supports out of working flanks where removal would damage them. Helical, herringbone and bevel parts need their own overhang and contact checks. A shape that can be generated is not automatically support-free on every machine.

Separate shaft fit, tooth clearance and center distance

Shaft fit controls the connection between a bore and its shaft. Tooth clearance controls the mesh. Nominal center distance positions the two gear axes. Changing one to hide a fault in another can make the mechanism unreliable.

The circular-hole control in the gear generator uses a radius. For a chosen model bore of 8.2 mm diameter, enter 4.1 mm radius. The export uses that radius directly; it does not silently add 0.2 mm. This arithmetic example is not a recommended allowance. Determine the desired diameter with a small coupon and your actual shaft.

Set the nominal shaft spacing with the center distance calculator. Measure the printed gears and inspect the mesh before adjusting it. Profile shift changes the tooth geometry and can affect the required working center distance; do not treat it as a generic backlash control.

A repeatable first-pair test

Generate a 20-tooth pinion and 40-tooth wheel with module 2, 20° pressure angle and zero profile shift using the spur gear generator. Their nominal center distance is 60 mm. Choose bores and face width for your fixture. This is a test procedure you can carry out, not a report of a physical test performed for this article.

Record the following before changing any setting. Print both parts, inspect and measure them, mount them without power and turn the pair through complete revolutions. Identify any repeatable tight spot. A freely turning pair passes that hand check only. Evaluate any powered test separately using a fixture, containment and stopping criteria appropriate to the application.

The dashed circles in the drawing are pitch circles. Their radii are 20 and 40 mm, so the theoretical axis spacing is 20 + 40 = 60 mm. The visible tooth tips extend farther. This standard, unshifted spur-pair calculation follows KHK’s gear dimension reference; it does not add a printer-specific allowance.

Use the log to compare one change at a time

The CSV contains a header and one empty row. Open it in Excel, Google Sheets or another spreadsheet, then use one row per trial. If the columns do not split automatically, import it as UTF-8 with a comma delimiter. Keep the trial ID with the STL files and printed parts. The two-page worksheet has the same fields for handwritten notes. Neither file contains example results.

First record the geometry, printing profile and fixture. After cooling, measure the bores, shafts and actual axis spacing with the same method. For a comparison, copy the baseline to a new trial ID, identify one changed variable and record its old and new values. Leave measurements you did not take blank. Prusa’s design and tolerance guidance explains why a fit allowance depends on geometry, orientation, process and material.

Describe what was held still and where you checked movement. Shaft movement or a flexible fixture can contribute to the play you observe, so do not label all movement as tooth backlash. Repeat the check at the same location and record any tight spot by its position in the rotation.

Application-test fields are optional. Use them only for a test you actually perform: record speed, time or cycles, temperature and the measurement location. If force is measured through a lever, record the perpendicular lever radius too. Force in newtons and torque in newton-meters are different quantities. Note the reason for stopping and visible damage. An empty torque field is more useful than an assumed strength rating.

Diagnose the failure instead of adding a fixed allowance

Binding throughout a revolution suggests checking tooth compatibility, shaft spacing and over-extrusion. One tight spot suggests runout, a seam or a damaged tooth. A poor shaft fit belongs in the bore calibration. Root cracking calls for a review of load, root geometry, material and print structure; appearance alone does not identify one guaranteed cause.

A widened first layer can also interfere with the mesh. Prusa’s elephant-foot explanation describes that effect and its compensation. Correct a measured problem rather than applying a fixed offset to every surface.

If lubrication is appropriate, check compatibility with the exact polymer and operating environment. Do not assume a lubricant halves wear or makes an under-designed gear safe. Use the planetary guide before combining rings and planets: those assemblies introduce additional interference checks.

Method, sources and corrections

This guide combines the linked manufacturer references with the generator’s own parameter definitions. The 20/40 example uses calculated dimensions; its tooth outlines come from the tool’s involute geometry. The worksheet is a proposed recording method with no measured results. No physical gear test, material ranking or service-life measurement was performed for this guide.

The editorial responsibility is GenerateSTL. To report a correction, use the public project issue tracker and identify the section, parameter or calculation involved. The public geometry engine lets you inspect how exported parts are constructed.