How to Make 3D Printed Parts Waterproof

• 1
  Key takeaways
• 2
  Where 3D printed parts actually leak
• 3
  Design the part so water has nowhere to go
• 4
  Pick a polymer that survives water
• 5
  Coatings, annealing and vapor smoothing compared
• 6
  Test the part, then decide whether to keep printing it
• 7
  Steps to make 3D printed parts waterproof
• 8
  Which sealing route fits your part

FDM parts are built from roads of molten polymer laid side by side. Each road cools and shrinks a little before the next one lands. The bond between them is never a full weld. Under 0.2 bar of water pressure, a gap of 20–40 μm between two roads is enough to weep. That gap is invisible on the outside of the part.

The worst path is between the outer perimeter and the infill. Slicers leave a small air gap there by design, so the perimeter does not get pushed around by infill overlap. Water follows that gap along the layer line and reaches the inside of the wall. If your part has four perimeters and a 0.15 mm gap, you have built a channel.

Layer lines are the second path. A 0.2 mm layer with a 0.4 mm nozzle leaves a visible ridge. The valley between two layers is a capillary. It will not pass a drip test at atmospheric pressure, but it wicks. Put the part in 60 °C water and the wicking speeds up because viscosity drops.

Threads and inserts are the third path. A printed M6 thread has a spiral gap along the flank. If that thread is the only seal on a port, it will leak at 1 bar. Sealing washers or an O-ring groove are not optional on pressurized ports.

• 1
  Perimeter-to-infill gapSet the slicer gap to 0 mm or use a solid top and bottom skin of at least 1.2 mm.
• 2
  Layer line capillarySmaller layers and hotter extrusion narrow the valley, but do not close it.
• 3
  Printed threadsSeal with an O-ring or a bonded insert, never with the thread alone.

Start with wall thickness. For a part that sees splash or short immersion, 1.6 mm of solid wall is a practical minimum. For anything above 0.5 bar or continuous contact, use 2.4 mm. That is 6 perimeters on a 0.4 mm nozzle. Below 1.2 mm the wall is mostly perimeter and the layer bond carries the load.

Avoid sharp internal corners on the wet side. A 0.5 mm radius is enough. Sharp corners concentrate stress and give coating a place to pull away. Fillet the outside too, since a rounded edge takes a gasket better than a knife edge.

Design a dedicated sealing face. A 3 mm wide flat land, machined or printed flat, gives an O-ring or gasket something to compress against. Do not rely on the general surface of the part. If the face is printed, orient it against the build plate so it comes off the printer flat.

Leave a groove for the seal. A rectangular groove 1.5 mm deep and 2.5 mm wide takes a 2 mm cord O-ring with about 20% compression. Print the groove 0.1 mm oversize in width. Printed grooves come out slightly narrow, and an undersized groove pinches the cord.

• 1
  Wall 1.6 mm minimum2.4 mm for pressure or constant wet contact.
• 2
  0.5 mm internal radiiSharp corners crack coatings and trap water.
• 3
  3 mm flat sealing landOrient it on the build plate for flatness.

PLA is the wrong choice for anything wet. It absorbs about 0.5% moisture by weight, softens near 60 °C, and hydrolyzes in warm water. A PLA part that holds cold tap water for a week may look fine. Leave it in a hot car or a dishwasher and it warps and crazes.

PETG is the default for printed water parts. It absorbs under 0.2%, prints at 235–250 °C, and bonds layers well. It is not immune to stress cracking in strong solvents, but for water and mild detergents it holds up. ASA and ABS are close behind and take UV better if the part lives outdoors.

PP and PA12 are the step up. PP is nearly non-polar, so water barely interacts with it. It is hard to print on an open machine because of warping. PA12, especially SLS PA12, is dense and isotropic. An SLS PA12 part at 95%+ density often passes a low-pressure leak test without any coating.

For hot water above 60 °C or steam, none of these polymers is a good long-term answer. PEEK and PEI exist, but they need a 400 °C hot end and a heated chamber. At that point, a machined aluminum or 316L body with an O-ring is cheaper and more reliable.

• 1
  Avoid PLA for wet partsHydrolyzes and softens at modest temperatures.
• 2
  PETG as the defaultPrints at 235–250 °C, low moisture uptake.
• 3
  SLS PA12 for densityIsotropic and often leak-tight as printed.

Epoxy and polyurethane coatings work by filling the layer valleys. Two thin coats beat one thick coat, because a thick coat traps solvent and sags at corners. Spray or dip, then cure per the datasheet. A 40–60 μm dry film is enough for splash. For immersion, build 100–150 μm in two passes.

Annealing helps by letting polymer chains relax and diffuse across the layer boundary. For PETG, hold the part at 65–70 °C for 2 hours, then cool at under 1 °C per minute. Faster cooling warps the part. PLA anneals around 60–70 °C but shrinks 1–3%, so allow for it in the drawing.

Vapor smoothing with acetone works on ABS and ASA. It melts the outer 50–100 μm into a glossy skin and closes the layer valleys. It does not seal the inside of a wall, so a smoothed part with a thin wall still wicks. Use it on the outside only, and keep the solvent away from inserts.

SLS parts can be sealed with a low-viscosity cyanoacrylate or an epoxy dip under vacuum. The vacuum pulls resin into the surface pores. Without vacuum, the resin sits on top and the pores stay open. This is the one process where a vacuum chamber is worth building.

• 1
  Two thin coats beat one thick40–60 μm dry film for splash, 100–150 μm for immersion.
• 2
  Annealing closes layer bondsPETG at 65–70 °C for 2 hours, slow cool.
• 3
  Acetone smoothing seals outside onlyThe inner wall stays porous.

Test with air, not water. Fill a sink, cap the part, and pressurize to 0.2 bar. Air bubbles show you the leak location immediately. Water testing hides the path because water wets the surface and you cannot see where it enters. Use a regulator, not a shop compressor at full line pressure.

Test at temperature, not just at room temperature. A part that holds 0.2 bar at 20 °C may fail at 50 °C because the polymer softens and the layer bond weakens. Run a second test with the part in warm water if the duty cycle includes heat.

Know the ceiling. Printed polymer parts are reasonable for splash, rain, wash-down and low-pressure static water. They are not a good answer for continuous pressure above 1 bar, hot water above 60 °C, steam, or fuel and solvent contact. At that point the failure mode is not a drip, it is a cracked wall.

The practical cutoff is around 200 parts. Below that, printed and coated is often the fastest route to a working assembly. Above it, or when the part must hold pressure for years, we machine the body from 6061 or 316L, cut the O-ring groove to ±0.005 mm, and finish the sealing face to Ra 0.8–1.6 μm. The seal then depends on a machined surface, not on a coating.

• 1
  Air test at 0.2 barBubbles show the leak path; water hides it.
• 2
  Retest warmPolymer bonds weaken as temperature rises.
• 3
  Switch to machined metal past the limitAbove 60 °C, above 1 bar, or long service life.