Why print time cannot be computed from volume
The intuitive model of a 3D printer is a nozzle laying plastic at a constant rate, and under that model print time is just volume divided by flow. If it worked, this page would be a single division and there would be no error band on it. It does not work, for one reason: a printer spends a large and unpredictable share of its life not extruding at the commanded speed.
A move at 60 mm/s only reaches 60 mm/s if it is long enough for the machine to accelerate there. At 3000 mm/s2 acceleration, reaching 60 mm/s takes 0.6 mm of travel and 20 milliseconds, which is nothing on a 200 mm straight wall and everything on a 1.5 mm segment of a curve approximated by short lines. A part made of gentle curves is a part made of thousands of short segments, and the nozzle never once touches the speed you set. Add travel moves between islands, retractions, seam handling, Z lifts, and the minimum layer time that cooling forces on small layers, and the gap between commanded speed and average speed becomes the dominant term.
That gap is what the motion efficiency field represents. It is a single fudge factor standing in for an enormous amount of physics, and it is why the honest output of this page is a range.
Volumetric flow, and the ceiling nobody prints past
The useful thing a volume model does give you is the flow rate your settings demand. An extruded line is not a cylinder and not a rectangle; it is a rectangle with rounded ends, squashed between the nozzle and the layer below. Its cross-section is (line width minus layer height) times layer height, plus pi times (layer height over two) squared. For a 0.42 mm wide line at 0.2 mm layer height that is 0.2 times 0.2, plus pi times 0.01, which comes to 0.0714 mm2. Multiply by print speed to get volumetric flow: at 60 mm/s, 4.28 mm3 per second.
That number is the one the hotend cares about. A hotend melts plastic at a finite rate set by its melt zone length, heater power and thermal contact, and once you exceed it the extruder skips or grinds, or the filament simply arrives at the nozzle not fully molten and comes out short. A conventional short-melt-zone hotend supplies roughly 8 to 12 mm3/s of PLA at typical temperatures; longer melt zones and high-flow designs go well beyond that, and PETG and nylon generally flow lower than PLA at the same setting. Whatever your figure is, this is the true speed limit on most machines. Doubling your speed setting past it does not double the throughput, it just under-extrudes.
| Layer x width | Line cross-section | Flow at 60 mm/s | Speed at 11 mm3/s |
|---|---|---|---|
| 0.12 x 0.42 | 0.0473 mm2 | 2.84 mm3/s | 233 mm/s |
| 0.20 x 0.42 | 0.0714 mm2 | 4.28 mm3/s | 154 mm/s |
| 0.28 x 0.45 | 0.1092 mm2 | 6.55 mm3/s | 101 mm/s |
| 0.32 x 0.60 | 0.1700 mm2 | 10.20 mm3/s | 65 mm/s |
| 0.40 x 0.80 | 0.2857 mm2 | 17.14 mm3/s | 39 mm/s |
Read the last column as the thing that actually changes your print times. Going from a 0.4 mm nozzle at 0.2 layers to a 0.6 mm nozzle at 0.32 layers more than doubles the plastic per millimetre of travel, so even at a lower commanded speed the job finishes far sooner.
The per-layer term, and why tall thin things are slow
Each layer carries a fixed cost that has nothing to do with how much plastic is in it: the Z move, the approach to the seam, the retraction and prime, and above all the minimum layer time. Cooling logic will not let a small layer finish in two seconds, because the plastic underneath would still be soft, so it slows the whole layer down or parks the head. On a 5 mm diameter spire, the per-layer cost is the entire print.
This is the mechanism behind a piece of practical advice that sounds like folklore: printing four small parts at once often takes barely longer than printing one. Four copies quadruple the extrusion time but share the layer count, and if the layer count was carrying the job, almost nothing changes. It also explains why the same volume of plastic can take four hours as a flat plate and fourteen as a tower.
Using the error band instead of ignoring it
The default band here is plus or minus 30 percent, which on a 14 hour estimate spans roughly 10 to 18 hours. That is not false modesty. Against a slicer estimate on real geometry, a volume-and-speed model landing inside 30 percent is doing well, and on small detailed parts it can be out by a factor of two in either direction. Narrow the band only after you have compared this page against your own slicer on your own parts several times and found a motion efficiency figure that fits your machine and your typical geometry.
The estimate is worth having for the decisions you make before slicing: whether a job fits in an overnight window, whether a nozzle change is worth the tuning it will cost, whether to quote a two-day turnaround or a one-week one. For anything where being wrong is expensive, slice it and read the number the slicer gives you. It has simulated the toolpath; this has multiplied two numbers together.
Work out the filament volume first if all you have is a model volume, then price the finished job with the filament cost calculator or the print quote calculator.
Questions people ask
Why is my slicer estimate so different from this one?
Because the slicer knows the toolpath and this page does not. The slicer has generated every move, applied your acceleration and jerk settings to each one, applied minimum layer times where cooling demands them, and added up the result. This page takes a volume, divides it by a flow rate, and then multiplies by a single efficiency factor to stand in for all of that. On large open geometry the two land close. On a small part full of short curved segments the slicer can be double what this says, because the machine never approaches the commanded speed. Trust the slicer. Use this when you have not sliced yet.
What motion efficiency should I use?
Calibrate it once against your own machine rather than taking the default. Slice three prints you consider typical, note the slicer time, then run each through this page and adjust the efficiency until the numbers agree. Most people land somewhere between 40 and 70 percent. Detailed organic models with lots of short segments sit at the low end, big flat functional parts at the high end, and machines with input shaping and high acceleration limits sit higher than older machines at the same commanded speed. One figure will not fit every model you print, which is part of why the band is wide.
Will printing faster actually finish sooner?
Only up to the point where the hotend runs out. Enter your flow limit and the calculator will show you the speed at which the settings stop being achievable. Past that the machine under-extrudes rather than speeding up, and you get thin walls, poor layer adhesion and gaps. If you want a genuinely faster print, the lever with the most travel in it is line cross-section, not speed: a taller layer and a wider line move more plastic per millimetre travelled, so the same job finishes sooner even at the same or lower commanded speed. The cost is surface finish and fine detail.
Does infill percentage change print time much?
Less than people expect above about 20 percent, and more than expected below about 10. Infill volume scales with the percentage, so going from 15 to 30 percent roughly doubles the infill plastic, but infill is usually a minority of the total once walls, top and bottom skins are counted. Infill is also printed in long straight lines at the highest speed the profile allows, so it is the most time-efficient plastic on the whole print. Walls are the slow part: they are short segments, printed slower for finish, and there are two or three of them on every perimeter. Adding a wall usually costs more time than adding ten points of infill.
Can I use this for resin printing?
No. Resin printing has an entirely different time model. An MSLA machine cures a whole layer at once, so time depends on layer count and exposure and lift settings, not on how much resin is in the layer. A full build plate and a single small part take almost exactly the same time at the same height. Everything on this page is about a nozzle laying a line and is meaningless for a light engine. The resin print cost calculator covers the resin side.