Printed gears and moving parts: what actually works
Clearances, materials and orientation for printed mechanisms, and why printing an assembly already assembled works better than it sounds.
A printed mechanism works or does not work on a single figure: the clearance between the parts that move. Everything else — gear module, tooth count, material — matters afterwards. If two parts touch with zero clearance in CAD, in reality they come out fused.
Clearances that work
| Joint | Total clearance | Note |
|---|---|---|
| Shaft rotating in a hole | 0.4 to 0.5 mm | Holes come out slightly under nominal |
| Gear teeth | 0.3 to 0.4 mm between flanks | Better loose and noisy than tight and seized |
| Print-in-place hinge | 0.3 mm per face | Less than that and it prints as one piece |
| Slide or guide | 0.4 mm | Split between the two parts |
Gears: what genuinely matters
- Big teeth beat fine teeth. A small module gives many tiny teeth the nozzle cannot reproduce crisply and that wear out fast. With a 0.4 mm nozzle, a tooth under about a millimetre and a half wide is already trouble.
- Print gears flat, axis vertical. The teeth are then formed by each layer’s perimeter, which is the most precise thing an FDM printer does.
- Use PETG rather than PLA if they will turn a lot. PLA wears, and more importantly it deforms under the heat friction itself generates.
- Use high infill or go solid. A tooth at 15% infill is a hollow tooth: it collapses as soon as it transmits any torque.
- Do not expect clockwork precision. A printed gear train drives a blind, a demonstration mechanism or a toy. To transmit real torque, the printed part is the prototype, not the product.
Printing the assembly already assembled
It is one of the things 3D printing does that nothing else allows: a hinge, a chain or a mechanism that comes off the machine already articulated, with nothing to assemble. It works surprisingly well if you respect the clearance and design so the facing surfaces need no support between them — support trapped inside a joint is impossible to remove and ruins the part. The trade-off is that if something goes wrong there is no component to replace: the whole thing gets reprinted.
Can you print gears that actually work?
Yes, and it is a regular job, especially as a spare for mechanisms nobody makes any more. What is worth knowing is what they are for: a gear in a blind or an appliance, yes; one transmitting real torque continuously will wear, and you should treat it as a temporary fix or a prototype.
Why did my mechanism come out fused into one block?
Because the designed clearance was smaller than the printer’s tolerance. A gap of one or two tenths closes as the material is laid down, and the two parts weld together. The fix is raising the clearance to 0.3 or 0.4 mm, or printing in resin, whose tolerance is far finer.
Is resin any good for mechanisms?
For detail and fine clearances it is far better: it holds gaps of a tenth where FDM needs three or four. The problem is brittleness — a standard resin tooth shears off suddenly — which is why, where there is load, tough resin or FDM in PETG are usually better calls.