Print Cooling: When and How Much Fan to Use
Slug: print-cooling-when-how-much-fan-use Category: Materials & Technology Original word count: ~420 Enhanced word count: ~1,750
Part cooling fans are among the most influential yet most misunderstood components on an FDM 3D printer. The difference between a perfectly crisp overhang and a droopy, sagging mess often comes down to nothing more than fan speed and timing. At 3DCentral, where we print tens of thousands of decorative collectibles on our 200-plus printer farm in Quebec, dialing in the exact cooling profile for each material and geometry type is essential to maintaining the surface quality our customers expect.
Why Cooling Matters in FDM Printing
When molten filament exits the nozzle at 200 to 250 degrees Celsius, it needs to solidify quickly enough to hold its intended shape before the next layer adds weight on top of it. Without active cooling, the deposited material remains soft and pliable for too long. On overhangs and bridges — sections where the filament spans open air without support from below — insufficient cooling causes the material to sag under its own weight before it solidifies.
The part cooling fan accelerates the transition from molten to solid by blowing air directly across the freshly deposited material. This forced convection dramatically increases heat dissipation compared to passive cooling through still air. The result is faster solidification, crisper overhangs, better bridge performance, and sharper detail on small features.
However, cooling is not universally beneficial. Too much airflow creates problems of its own, and the optimal cooling strategy varies significantly between materials. Getting cooling right requires understanding both the benefits and the trade-offs.
PLA: Maximum Cooling for Maximum Quality
PLA responds extremely well to aggressive cooling. Its relatively low glass transition temperature — around 55 to 60 degrees Celsius — means it solidifies quickly once cooling begins, and the rapid solidification preserves fine detail and produces excellent overhang performance.
For PLA printing, run the part cooling fan at 100 percent after the first two to three layers. There is rarely a reason to use less than full fan speed on PLA after the initial layers. The material’s crystallization behavior actually benefits from rapid cooling — slower cooling can allow more crystallization, which sometimes leads to warping on larger flat surfaces.
At 3DCentral, every PLA printer in our fleet runs part cooling at 100 percent from layer three onward. This setting is baked into our standard PLA profiles and not adjusted on a per-model basis. The consistency simplifies our workflow — operators do not need to evaluate each model’s cooling requirements.
PETG: Moderate Cooling with Careful Control
PETG has a fundamentally different relationship with cooling than PLA. Its higher glass transition temperature and different crystallization behavior mean it responds poorly to the aggressive cooling that PLA thrives on.
Running a PETG print with 100 percent fan speed typically produces poor layer adhesion and a rough, matte surface finish. The rapid cooling prevents adequate inter-layer bonding and can cause internal stress that leads to delamination. PETG needs time for the polymer chains in adjacent layers to intermingle, and aggressive cooling cuts that bonding window short.
The sweet spot for PETG cooling generally falls between 30 and 50 percent fan speed. Some operators prefer even lower — 20 to 30 percent — for thick-walled parts where surface finish matters less than structural integrity. For thin walls and delicate features, 40 to 50 percent provides enough cooling to maintain shape without destroying layer bonds.
At 3DCentral, our standard PETG profile uses 40 percent fan speed. For models with significant overhangs, we bump to 50 percent. Models with very thin walls or tall narrow sections may warrant different per-model adjustments, but these represent a small minority of our production prints.
Silk PLA: Reduced Cooling for the Best Sheen
Silk PLA filaments achieve their distinctive metallic luster through polymer additives that alter the material’s flow and surface properties. These additives respond best to slightly slower cooling than standard PLA — too much fan speed can dull the characteristic shimmer by disrupting the surface layer formation that creates the pearlescent effect.
We run Silk PLA at 70 to 80 percent fan speed, a modest reduction from standard PLA’s 100 percent. This allows the surface layer to form its lustrous finish while still providing enough cooling for decent overhang performance. The trade-off is slight — overhangs may not be quite as crisp as with standard PLA at full cooling, but the improved surface finish on visible areas more than compensates for minor overhang softening on collectible display pieces.
First Layer: Always Zero Fan
Regardless of material, the first layer should always print with the part cooling fan completely off. The first layer must bond to the build surface — the heated bed and build plate — and cooling interferes with that adhesion process. The heated bed maintains the bottom surface of the print above the glass transition temperature so the material conforms to and grips the build surface. Blowing cool air across that first layer fights the heated bed and can cause corner lifting, edge curling, or complete first-layer detachment.
Most slicer software defaults to zero fan on the first layer and ramps to the target speed over the first two to four layers. This graduated approach allows the initial layers to establish solid bed adhesion before cooling begins affecting layer formation. At 3DCentral, our profiles specify zero fan for the first three layers with a linear ramp to target speed over layers four and five.
Directional Considerations and Fan Duct Design
Not all cooling is equal. The direction and distribution of airflow across the print matters as much as the volume of air moved. Most consumer 3D printers ship with a single-sided radial fan and duct that blows air from one direction. This asymmetric cooling can produce different overhang quality on different sides of the same print — the side facing the fan cools faster and prints cleaner, while the opposite side receives less airflow and may show slight sagging.
Dual-fan or 360-degree cooling duct upgrades address this asymmetry by surrounding the nozzle with even airflow from all directions. For printers used in production, we consider symmetric cooling a necessity rather than a luxury. Our production printers use dual-fan setups with optimized duct geometry that we have tested and refined over thousands of hours of print time.
The duct geometry itself matters. A well-designed duct focuses airflow on the zone immediately below the nozzle where fresh filament is being deposited, not on the nozzle itself. Cooling the nozzle can cause heat creep — thermal energy migrating upward through the filament in the heat break — leading to jams. The target is the printed part surface, not the hot end.
Bridging: Maximum Cooling for Open Spans
Bridges — horizontal spans between two supported points with nothing underneath — represent the most cooling-demanding feature in FDM printing. When filament is extruded across open air, gravity pulls it downward while surface tension and rapid solidification attempt to hold it straight.
For bridges, virtually every material benefits from maximum cooling. Even PETG, which normally runs at moderate fan speeds, performs better on bridges with fan speed boosted to 80 to 100 percent. Most modern slicers allow bridge-specific fan speed overrides, and we take advantage of this feature extensively.
Bridge cooling also benefits from slower print speeds. The combination of slower extrusion and faster cooling gives each segment of bridge filament more time to solidify before the nozzle moves on, reducing sag. Our bridge speed settings are typically 30 to 50 percent of normal perimeter speed, combined with maximum fan speed.
Small Features and Minimum Layer Time
Tiny features — thin pillars, pointed tips, small details — present a cooling challenge because the print head deposits material and immediately returns to deposit more before the previous layer has fully cooled. This creates cumulative heat buildup that softens the entire small feature, causing it to deform.
Most slicers include a minimum layer time setting that forces the printer to slow down on small layers so the material has time to cool. Typical values range from 5 to 15 seconds. If the layer would complete in less than the minimum time at normal speed, the printer reduces speed to stretch the layer duration.
At 3DCentral, we set minimum layer time to 10 seconds for PLA and 12 seconds for PETG. For models with many small pointed features — dragon horns, gnome hat tips, duck bills — these settings prevent the mushy deformation that would otherwise ruin the details that make our collectibles appealing.
Cooling Profiles in Practice
Building reliable cooling profiles requires systematic testing, not guesswork. Print a standard overhang test — most slicer libraries include one — at several fan speed settings and compare the results. The speed that produces clean 50-degree overhangs without visible layer separation is your working target for that material.
Document your settings and apply them consistently. At 3DCentral, every material in our inventory has a validated cooling profile stored in our slicer’s configuration. When a new material enters production, it goes through a standardized qualification process that includes cooling optimization before any customer orders are printed with it.
