Precision metal and engineering plastic parts from prototype through repeat production.
Tooling development, molded parts and production support for repeat plastic manufacturing.
Functional prototypes, complex geometry and low-volume parts without conventional tooling.
Explore 3D PrintingStart with geometry, material, quantity and critical requirements before selecting the route.
Request Engineering ReviewSend your CAD, drawing, material and quantity for an initial manufacturing review.
Request Engineering ReviewA high Cpk value does not automatically prove that an injection molding process is production-ready. For a critical-to-quality (CTQ) dimension, capability becomes meaningful only when the process is stable, specification limits are valid, the measurement system is reliable, and the sampling basis represents the actual molding condition.
This guide focuses on interpreting Cpk for injection-molded CTQ characteristics. It does not replace process-window development, FAI inspection or PPAP submission. For the wider qualification sequence, see the injection molding validation procedure .
Capability Fundamentals
Cp and Cpk both compare process variation with the allowable specification range, but they answer different questions. Cp evaluates potential capability based on process spread, while Cpk also considers how well the process is centered between the specification limits.
Cp asks whether the natural process variation can fit inside the specification window. It assumes the process is centered and therefore does not show whether production is drifting toward either specification limit.
Cpk evaluates the same variation while also measuring how close the process mean is to the nearest specification limit. A process may therefore show a reasonable Cp but a lower Cpk when the molded dimension is off-center or drifting.
For injection-molded CTQs, do not interpret either index alone. A capability value is useful only after the process is stable and the specification, measurement system, sampling basis and cavity identity are clearly defined.
CTQ Selection
Cpk should not be calculated for every dimension on a molded-part drawing. A formal capability study is most useful when a characteristic is critical to function, assembly, sealing, safety, regulation or customer release and when repeated production data is required to prove that the process can hold that requirement consistently.
The correct question is not “Can we calculate Cpk?” but “Does this characteristic require statistical evidence of ongoing process capability?” The drawing, CTQ plan, validation requirement and customer quality criteria should define that decision.
Variation vs Centering
Cp and Cpk are related, but they do not answer the same question. Cp looks at how wide the process variation is relative to the specification limits, while Cpk also checks whether the process is centered inside those limits. A process can have acceptable spread and still perform poorly if the mean is shifted toward one side.
When the process spread is already too wide for the tolerance, both indicators tend to be weak. The first problem is not centering; it is that the process variation itself is too large.
This pattern often shows that the process could fit within the tolerance if it were centered better. The capability loss comes from mean shift, not necessarily from excessive spread.
This is the target condition for a CTQ characteristic: variation remains contained, and the process mean stays positioned so that the dimension is not drifting toward one specification boundary.
If Cp and Cpk are materially different, start by checking process centering. If both are weak, the first question is usually whether the molding process is generating more variation than the CTQ tolerance can absorb.
Cpk Readiness Gate
Cpk should be interpreted only after the data represents a stable and clearly defined production condition. A calculated value can look precise while still being misleading if the process, specification, sampling plan, cavity basis or measurement system is not controlled.
Special-cause shifts, startup drift or uncontrolled parameter changes should be addressed before capability is interpreted.
USL and LSL must come from the approved drawing, CTQ plan or customer requirement—not from an internal convenience limit.
The sample must represent the intended production condition, rather than a small or selectively chosen set of favorable parts.
Multi-cavity molds should preserve cavity identity when cavity behavior could affect the CTQ distribution.
The inspection method must resolve the tolerance well enough that gauge variation does not dominate the observed process variation.
If one of these five conditions is not established, treat the Cpk value as provisional evidence, not as a standalone production release decision. Resolve the data-quality or process-stability issue before using capability to support validation or PPAP evidence.
Sampling Structure
A capability study needs more than a large number of measurements. The data must preserve when the parts were molded, which cavity produced them and whether the sample represents normal production. Poor sampling structure can hide drift, cavity imbalance or short-term instability behind a single attractive Cpk value.
A short run may miss warm-up drift, material-lot effects or process changes that appear during sustained production.
Samples should allow engineers to distinguish within-subgroup variation from changes over time, rather than treating every measurement as interchangeable.
Pooling all cavities can mask one weak cavity behind stronger ones. Review cavity-level behavior before combining data into a single production capability statement.
Sample size should be defined by the validation or customer requirement, not by a universal number. The key question is whether the dataset captures the variation that production will actually experience without hiding cavity or time-related effects.
Measurement System Risk
The variation seen in a capability study is not automatically the true molding-process variation. It also contains variation introduced by the gauge, fixture, measurement method, operator and part positioning. If that contribution is too large, the calculated Cpk may misrepresent the actual process.
What the dataset actually contains
Observed Variation = Process Variation + Measurement Variation
A gauge with insufficient resolution can blur small dimensional changes or make a narrow tolerance appear noisier than the molding process really is.
Flexible molded parts can shift, deform or seat differently during inspection. The measurement method must control that influence consistently.
If repeated measurements are not sufficiently consistent, investigate the measurement system before using Cpk to judge the molding process.
Do not use Cpk as a production judgment until the measurement method is capable of distinguishing meaningful CTQ variation. For critical dimensions, the inspection system should be validated to the level required by the customer or quality plan before capability evidence is accepted.
Process Window vs Capability
Cpk measures the capability of a CTQ output under defined production conditions; it does not prove that the underlying molding process has a robust operating window. That broader task belongs to scientific molding and process window validation , where process inputs and acceptable operating limits are established before capability evidence is interpreted.
Define and control the molding conditions that materially influence the part, such as fill, pack, cooling and material state.
Confirm that acceptable parts can be produced within an approved, repeatable operating region rather than at one favorable setting.
Collect representative dimensional data from the defined production condition using the approved sampling and measurement basis.
Use Cp and Cpk to judge whether CTQ variation and centering remain compatible with the required specification limits.
Can the process remain robust across an approved operating window, and which input conditions must be controlled to maintain that state?
Under that defined production condition, is the CTQ output centered and consistent enough relative to the specification limits?
A strong Cpk result should be treated as capability evidence inside a validated process system, not as a substitute for process-window development or scientific molding studies.
Release Interpretation
Cpk should be treated as one piece of production evidence, not as a standalone pass/fail score. The same numerical result can carry very different weight depending on whether the dataset is stable, representative and measurement-system credible.
A high Cpk value can support production release only when the underlying evidence is credible and the required acceptance target has been defined by the customer specification, CTQ plan, validation requirement or approved quality plan. Cpk should never create its own universal release threshold.
Evidence Handoff
Cpk is most useful when it sits inside a controlled evidence chain. FAI confirms that initial parts meet drawing requirements, capability analysis evaluates whether selected CTQs remain statistically capable, and approved evidence then supports customer release and ongoing production control.
Use the injection molding FAI report to verify initial dimensional conformity.
Apply capability analysis only to defined CTQs using credible, representative production data.
Relevant capability results can become part of the injection molding PPAP documents when required.
Transfer approved settings and reaction limits into the injection molding process control sheet .
Share the drawing, CTQ requirement and intended production condition so the engineering team can review the appropriate validation and capability evidence path.
