3D Printer Dimensional Accuracy: Why Parts Print Out of Tolerance and How to Calibrate

Where the gap between the CAD dimension and the printed part comes from

An FDM print is never an exact copy of the model. Molten filament gets laid down by a round nozzle, then it cools and contracts. According to the Simplify3D documentation, the most common factors that hurt dimensional accuracy are under-extrusion or over-extrusion, thermal contraction, filament quality, and nozzle position on the first layer. Each one acts on the part differently, and that difference is exactly what lets you tell them apart.

Too much flow widens every wall by a roughly constant amount. Thermal shrinkage grows with the size of the part. A squished first layer only affects the bottom of the object. Filament with an inconsistent diameter produces errors that change from one spool to the next, sometimes from one meter to the next.

So before you change anything, grab a pair of calipers and measure the same test part in several places: near the bottom, near the top, on a small dimension and on a large one. The shape of the error tells you where to start, and the full order is laid out in the guide on 3D printer calibration.

Measure away from the first layer or you will chase a ghost

The Simplify3D documentation stresses a point that gets overlooked constantly: first layer nozzle height can throw off the next 10 to 20 layers. If you print a 0.2 mm layer but the nozzle sits only 0.1 mm above the bed, the extra plastic spreads out and creates a first layer that is slightly too wide. The layers above inherit that excess, and the bottom of the part comes out oversized while the rest is fine.

That is the mechanism behind elephant foot, and it trips up plenty of people who measure their test cube at the base. To avoid that bias, Simplify3D recommends printing a model with 50 to 100 layers and measuring only the top 20 or so, far from the bed's influence.

If your top measurements are good but the bottom is not, the problem is not your printer's accuracy. Go back to the nozzle height procedure in the first layer guide, then measure again. If the error is the same at the top and the bottom, the first layer is cleared and you can move on to the next step, which is flow rate.

Dial in the extrusion multiplier with a vase-mode cube

The extrusion multiplier, also called flow rate, controls how much filament comes out of the nozzle. According to Simplify3D, a multiplier that is too low gives parts smaller than intended, with gaps between perimeters and holes in top surfaces. Too high, and you get parts larger than intended with top layers that bulge upward. The Prusa Knowledge Base adds that the ideal value differs for every material and color, and can even change from spool to spool, especially with cheap no-name filament.

The precise method Prusa describes is to print a cube in vase mode, measure each wall's thickness at three or more points, and average the readings. The new multiplier equals the target extrusion width, 0.45 mm in Prusa's example, divided by the average measured wall thickness. Reprint, remeasure, and repeat until the numbers settle.

Prusa notes that adjusted values usually land between 0.9 and 1.1, which shows up as 90 to 110 in the printer's Tune menu, and suggests nudging by 1 to 2 percent at a time with the visual method. The pages on under-extrusion and over-extrusion cover the symptoms in more depth.

Constant error: horizontal size compensation in the slicer

Once flow is calibrated, a stable offset sometimes remains. The part is always too wide by the same amount whether it is 20 or 100 mm across. That offset usually comes from the geometry of the extruded bead, which spills slightly past the theoretical line, outward on outer contours and inward on holes.

Simplify3D offers a setting for this case called Horizontal size compensation, found on the Other tab of the Process Settings. A value of -0.1 mm shrinks the model by 0.1 mm in X and Y. The documentation points out that it works best when the error is consistent regardless of model size. Other popular slicers have an equivalent parameter, usually tucked into the advanced print settings.

This fix belongs at the end of the list, never at the start. If you paper over an over-extrusion problem with a geometric offset, your holes stay too small and your top surfaces stay bulged. Also check that the correction does not depend on the material. Keeping a separate filament profile per material type keeps you from correcting the same flaw twice, or applying a PETG value to PLA.

Error that grows with part size: thermal shrinkage

If the error increases with size, thermal contraction is almost always the culprit. Simplify3D gives the example of a 20 mm part that is 0.1 mm too small while a 100 mm part is 0.5 mm too small. The relationship is proportional, so it has to be the plastic shrinking as it cools. High temperature materials such as ABS are affected most, which lines up with the precautions on the page about ABS and ASA.

The fix is a scale factor. In the example above, 0.1 divided by 20 gives 0.5 percent shrinkage, so you import the model at 100.5 percent. Simplify3D lets you automate that resize with an Import Action applied to every new model. Most slicers at least let you scale manually per axis, so note the value in your profiles.

That factor is specific to each material and shifts with the ambient temperature of the room you print in. An enclosure makes results more consistent from one print to the next, but it does not eliminate shrinkage. You will still have a percentage to apply, just a steadier one, which is exactly what you want for repeatable parts.

Checklist order and the tools you need to measure right

To sum up the method, stick to this order: first layer first, then extrusion multiplier, then horizontal compensation for a constant error, and finally scaling for a proportional error. Each step assumes the previous one is done. Otherwise you end up fixing a symptom with the wrong tool, and the defect comes back the moment you swap spools.

On the hardware side, a digital caliper covers the vast majority of needs. Prusa also mentions micrometers for the precise method. Always measure at the same spot, several times, and average the readings, because a bead of plastic is never perfectly smooth. If you are still figuring out what gear to buy, the startup budget tool lists the accessories worth planning for.

Finally, accept that an FDM printer has limited accuracy by nature. For tight assemblies, design clearance into the file instead of hunting for absolute zero, and test the adjustment on a small part before committing to a full bed. A mechanical part that fits on the first try is the real measure of success.

Sources and limits

• Simplify3D — Dimensional Accuracy
• Prusa Knowledge Base — Extrusion multiplier calibration

We organize manufacturer and community documentation; we have not measured these products ourselves. The manual for your exact model takes precedence over general guidance.