Small smears and lumps appear on top of finished surfaces, in places the toolhead only travelled across. Not hairs between features — those are stringing — but deposits sitting on the part, sometimes dragged into a comet shape by the nozzle that left them.
An open nozzle is not a closed valve
There is nothing shutting off the tip. What holds plastic in is the balance between the melt's viscosity, its surface tension at the orifice, and whatever pressure is pushing on it from above. Warm the plastic further and viscosity falls; leave it sitting in a hot block and it expands as it comes to temperature. Both push material out.
That is why a machine can ooze while completely idle. A nozzle sitting at printing temperature for ten minutes before the print begins will produce a hanging drip with the extruder motor untouched — pure thermal expansion and gravity, no commanded flow at all.
The two moments it costs you
During travel. Residual melt pressure and expansion push material out while the head moves, and it lands wherever the head happens to be.
At the start of the next path. Everything that leaked out is material that is not in the nozzle when printing resumes. The first few millimetres of the new extrusion are then thin, so a machine that oozes also under-fills — which is why fixing this often improves surface quality in places that appear unrelated.
Make sure it is coming out of the orifice
There is a look-alike worth ruling out in ten seconds. If plastic is appearing on the outside of the nozzle rather than being deposited from the tip — a growing skirt of material creeping around the flats of the nozzle, or a burnt lump on the heater block — nothing is oozing. The hotend is leaking at its internal seal, and that is plastic escaping around the nozzle threads, which needs dismantling rather than tuning. Oozing leaves clean material on the print; a leak leaves darkened material on the machine.
Reducing the leak
- Lower the nozzle temperature toward the bottom of the material's workable range. This is the biggest lever, and a temperature tower tells you how far you can go before layer bonding suffers.
- Enable Z hop on travel moves. It does not stop the leak, but it lifts the tip clear so the drip is not smeared into the surface. Small values are enough.
- Tune retraction properly rather than maximally — enough to relieve pressure, not so much that softened filament is dragged up the heatbreak, which risks heat creep.
- Cut the heat-soak. If your start routine heats the hotend and then waits for the bed, invert it: bring the bed up first, heat the nozzle last. A hotend that reaches temperature thirty seconds before printing starts loses far less material than one that sits hot for five minutes.
- Use a longer prime line, and a wipe at the end of it, so whatever accumulated during startup is deposited somewhere harmless before the part begins.
- Enable combing so travel stays inside the part where a small deposit lands on infill that will be covered anyway.
Where a purge tower earns its place
On multi-material machines and on any print with frequent tool or colour changes, a purge or wipe tower gives the nozzle somewhere to unload before it returns to the part. It costs material and it is worth it — the same drip that ruins a visible top surface is invisible on a tower.
For single-material printing, a small sacrificial object beside the part serves the same purpose on a model that has a lot of travel over finished surfaces.
What not to try
Turning retraction up until the oozing stops. Past a certain point you are not preventing the leak, you are pulling molten plastic into the cold section of the hotend, and the failure that produces is far worse than a cosmetic smear.
Printing with the fan at maximum to "freeze" the drips. Part cooling airflow does not reach the inside of the orifice, and on materials that need a warm environment it costs layer strength for no benefit here.
The materials that leak most
PETG and PCTG, whose melts are fluid and eager. Polyamides, which are fluid, hygroscopic and hot all at once — a wet nylon does not merely ooze, it actively spits. TPU oozes because its melt is soft and the filament path is compliant, so retraction reaches the tip poorly. PLA at a sensible temperature barely does this, and PLA that is oozing badly is usually being printed a good deal hotter than it needs.