Assessing 3D printing process capability is essential when a component’s dimensions must reliably meet design limits. A printer may produce a part that looks acceptable, yet measurements can vary across its height or with its internal geometry. Capability analysis helps engineering and operations teams quantify that variation and decide whether a process is suitable for a specific part and tolerance.
An experimental evaluation of fused filament fabrication (FFF), also known as Fused Deposition Modeling (FDM), illustrates why capability should be assessed on representative parts rather than assumed from a machine’s nominal specifications. The results point to practical considerations for additive manufacturing quality: measure actual output, account for part geometry, and treat capability as specific to the process and application.
What process capability says about a 3D printer
Every manufacturing process has variation. In 3D printing, the finished part’s dimensions can be influenced by factors including material, printing process, operating speed, printer type and the volume or geometry of the part. Dimensional tolerances define the acceptable range for a feature; capability analysis compares the process’s observed variation with those specification limits.
Two common indices are Cp and Cpk. Cp compares the width of the specification range with the process’s variation. It indicates whether the process spread could fit within the limits, but does not show whether the process is centered. Cpk also accounts for the process mean’s position relative to the nearest specification limit. A process can therefore have a reasonable spread yet a lower Cpk if its output is shifted toward one limit.
These indices are only meaningful when the data and process conditions support their use. The evaluation discussed here notes the importance of a stable process, a suitable distribution model and representative sampling. Capability numbers should be interpreted alongside the measurement method and the conditions under which parts were printed.
How the FDM capability evaluation was conducted
The study examined rectangular PLA samples produced on an Ultimaker 2 Extended+ printer using FFF. Samples shared a 20 mm by 20 mm base, while their heights varied across four groups: 20, 40, 80 and 150 mm. Different internal infill patterns were considered, including triangular, concentric, rectilinear, gyroid and honeycomb designs. A raft was used to provide a base for the printed parts.
Rather than measuring each part at only one point, the procedure captured dimensions at the bottom, middle and top levels. X, Y and Z measurements were taken three times at these locations using a digital Vernier caliper connected to a computer, with readings recorded for analysis. The team used statistical summaries and plots, including histograms and box-and-whisker plots, to examine the spread and distribution of measurements and calculate Cpk.
This measurement approach has an important implication for quality planning: a single reading or a measurement at one height may not represent the dimensional behavior of the entire part. Sampling at multiple vertical positions can reveal variation that matters when a component has features or interfaces distributed along its height.
What the Cpk results indicate
For the samples and conditions evaluated, the reported Cpk values were below 1 across the measured levels. In the study’s interpretation, this meant the process was not capable of consistently meeting the defined requirements for those samples. The authors also observed differences among infill patterns and changes in capability with part height, indicating that a single capability value should not automatically be applied to every geometry.
For one reported sample row, Cpk declined from -0.11 at the bottom to -0.22 at the middle and -0.39 at the top. These figures describe that evaluated case, not a universal performance profile for FDM printers. They do, however, show why checking dimensions along the Z direction can matter. The study also reported that bottom-level Cpk increased with sample height across rows, while the pattern at the middle and top levels was not clear.
A negative Cpk indicates that, for the data and specification limits used in the calculation, the process mean was beyond a specification limit. The finding should be read in its experimental context: the reported capability depends on the tested printer, material, samples, measurement locations and limits. It is not a blanket conclusion about all additive manufacturing equipment or applications.
Applying capability analysis to design and production
For engineers and system integrators, the practical lesson is to connect design requirements with demonstrated process performance. Before assigning a 3D-printed component to a production or functional application, teams can use a structured assessment to:
- Define critical dimensions and limits. Identify which features affect fit, assembly or function, and establish their acceptance ranges.
- Measure representative parts. Include relevant locations and dimensions, particularly when geometry changes along the build direction.
- Evaluate the actual configuration. Consider material, infill and part geometry rather than relying on a capability result from a different setup.
- Review variation as well as averages. Use suitable statistical summaries and confirm that the data support the assumptions behind the selected capability index.
- Reassess after changes. A different geometry or internal design may produce different results, so prior measurements may not apply.
Process capability analysis also helps distinguish a machine’s stated resolution from its demonstrated ability to produce parts within a particular tolerance. The evaluation included a printer specification of 12.5-micron XYZ resolution, but its measured-part results varied by location and configuration. Resolution alone, therefore, should not be treated as evidence that a finished component will meet a specified dimensional tolerance.
Conclusion: Make capability part of the design decision
The FDM evaluation shows that dimensional capability can vary across a printed part and with its height and internal pattern. For industrial teams, the takeaway is to validate the complete application—not just the printer—by measuring representative parts against defined limits. Using capability data early in design and process selection can help identify dimensional risks, support more informed tolerance decisions and guide quality improvement before those risks reach production.
Access Official Technical PDF