Process Capability Study

How long should the study run, what subgrouping should I use, and how do I keep the process untouched during the window?

Studies are categorized as short-term (Cmk / machine capability), initial (Cp/Cpk during PPAP), or long-term (Pp/Ppk during ongoing production).

Common program conventions — confirm the governing requirement

These values are common examples, not universal requirements. Acceptance is controlled by the applicable customer, contract, drawing, customer-specific requirement, Control Plan, validation protocol, or approved internal standard. The study type, sigma estimator, sample plan, and process condition must match that governing requirement.

• Cmk ≥ 1.67 — a common machine-capability acceptance example when specified.
• Cpk ≥ 1.67 — a common PPAP or launch example when specified.
• Ppk ≥ 1.33 — a common ongoing-production example when specified.

1. Document the study charter: purpose, index required, acceptance criterion, sample plan.
2. Baseline the process at nominal setpoints; do not run at the edges of the tolerance.
3. Communicate a "no-touch" policy to operators for the study duration.
4. Run MSA before the study begins, not after.
5. Collect subgroups at planned intervals; record time, operator, tool ID, and lot.
6. Track any special-cause event (tool change, material lot, jam) in a study log.
7. Discard subgroups only with a documented special-cause explanation.
8. Compute and report Cp/Cpk and Pp/Ppk with the full study metadata.

• Collecting five parts per hour for eight hours and calling it a Ppk study.
• Excluding subgroups that "look bad" without an assignable cause.
• Running the study on a freshly serviced machine and reporting it as steady-state capability.
• Ignoring warm-up drift; the first hour dominates the histogram tail.

• A "capability study" that lasts less than one full tool life will overstate Ppk.
• Setup variation dominates most short-run studies — the first 5–10 parts are almost always non-representative.
• If the operator adjusts the machine to keep parts inside spec, the study measures the operator's skill, not the process.
• For high-cavity molds and multi-spindle machines, run one study per stream and one pooled study; the difference exposes stream-to-stream variance.

• New processes still in tuning — the study will be obsolete on the next parameter change.
• Destructive tests without a validated indirect measurement plan.
• Very low-volume characteristics (n < 30) — use interval estimates, not a single Cpk value.

• Study duration drives the Cpk-vs-Ppk gap: longer window, larger gap.
• Subgroup design drives within-subgroup σ, which drives Cp and Cpk.

1. Define the study window and subgrouping before data collection.
2. Keep the process unchanged during the baseline study.
3. Document the measurement system and operating conditions.
4. Use the free Process Capability Calculator to compute Cp, Cpk, Pp, Ppk, sigma level, and predicted PPM from your measurement data — no sign-up required.

• Attach the sampling log to the capability report; without it, results cannot be reproduced.
• Never mix a machine capability (Cmk) result with a production capability (Ppk) result in the same statement.

• ✓Study charter approved before data collection
• ✓MSA run and accepted before the study window
• ✓No process adjustments during the window (or all adjustments logged)
• ✓Subgrouping matches the intent (short-term vs long-term)
• ✓Streams (cavities, spindles, lanes) separated or explicitly pooled
• ✓Full metadata (time, operator, lot, tool) recorded for every subgroup

• Process condition: statistical stability is required.
• Distribution assumption: use a distribution model justified for the data.
• Confirm process stability and measurement-system adequacy before interpreting a capability index.
• Use a justified distribution model or non-normal method when the normal model is unsuitable.