Qhubio
    Process Capability · Cp · Cpk · Pp · Ppk

    Quality Statistics Knowledge Center

    4 min read Last updated

    A structured reference for process capability: definitions, interpretation, industry targets, conversions, and process-state diagnosis. Start with the calculator if you already have measurement data; use the linked pages to interpret the result and translate it into engineering action.

    Open the Process Capability Calculator Plan a Process Capability Study

    What process capability measures

    Process capability quantifies how well a stable, in-control process meets engineering specification. Cp, Cpk, Pp, and Ppk compress the study into a comparable number, but only when the measurement system, stability, and distribution assumptions are defensible.

    • Cp — potential short-term capability, assuming centering.
    • Cpk — actual short-term capability, including centering.
    • Pp — potential long-term capability using overall sigma.
    • Ppk — actual long-term capability using overall sigma.

    Why capability is not the same as stability

    A capability index is meaningful only after the process is shown to be in statistical control. Otherwise the number may be mathematically correct and still operationally useless.

    • Use the right control chart for the data stream.
    • Verify MSA / Gage R&R before you draw conclusions.
    • Check normality or use a documented non-normal method.
    • Separate mixed streams before calculating capability.

    How to use this hub

    • Run the numbers first. The calculator turns raw measurements into Cp, Cpk, Pp, Ppk, sigma level, and predicted PPM.
    • Interpret the result. Use the pages on Cp vs Cpk, Cpk vs Ppk, Cpk thresholds, and negative Cpk to understand what the number means.
    • Match the requirement. Go to the industry pages when a customer requirement, drawing note, or Control Plan matters more than the generic threshold.
    • Diagnose the process state. If the number is poor, move to the process-state pages before trying to “improve capability” blindly.

    What the strongest articles must answer

    • What does the number actually prove, and what does it not prove?
    • Which assumptions are required: stability, MSA, normality, centering, or independence?
    • What is the next action: re-center, reduce variation, separate streams, or escalate containment?
    • How does the answer change for PPAP, launch, ongoing production, or machine capability?
    • Which linked page carries the next decision step?

    Typical capability targets — reference values, not universal rules

    These values appear often in automotive and manufacturing practice, but they are not universal acceptance criteria. The governing requirement is still the CSR, drawing, Control Plan, validation protocol, or internal standard.

    • Cpk / Ppk ≥ 1.33 — common reference for stable production.
    • Cpk / Ppk ≥ 1.67 — common launch / PPAP reference.
    • Cpk / Ppk ≥ 2.00 — sometimes used for critical or high-risk characteristics.
    • Cmk ≥ 1.67 — typical machine-capability acceptance.

    Learning path

    1. 1. Compute
      Run the calculator

      Paste measurements and spec limits into the calculator to get Cp, Cpk, Pp, Ppk, sigma level, and predicted PPM.

      Open calculator
    2. 2. Stabilize
      Prove the data is valid

      Before you trust a capability number, confirm control, MSA, and distribution shape. If those are wrong, everything downstream is wrong.

      Read analysis steps
    3. 3. Interpret
      Read Cp vs Cpk and Cpk vs Ppk

      Decide whether the issue is centering, variation, drift, or a mismatch between short-term and long-term study methods.

      Interpret indices
    4. 4. Compare
      Match the governing requirement

      Check CSR, drawing note, Control Plan, or validation protocol. Common targets are references, not universal rules.

      Compare requirement
    5. 5. Diagnose
      Find the process state

      If Cpk is low, determine whether the fix is re-centering, variation reduction, or re-establishing stability.

      Diagnose state
    6. 6. Convert
      Translate between units

      Use the conversion pages when scorecards mix PPM, sigma, Cp, Cpk, Pp, and Ppk.

      Convert metrics

    Limitations and assumptions of the normal model

    • PPM predictions from Cp/Cpk assume approximate normality; skewed or heavy-tailed data will mislead the tail estimate.
    • Cpk describes the nearest specification limit only. Total two-sided defects need centering or both tail distances.
    • Short-term indices (Cp/Cpk) and long-term indices (Pp/Ppk) answer different questions.
    • A high Cpk on an unstable process, a process with weak MSA, or a mixed-stream process is not trustworthy.

    How to strengthen a weak article

    • Add a worked example with a realistic USL, LSL, mean, sigma, and interpretation.
    • Add a diagnostic block: centering vs variation vs instability vs mixed streams vs MSA.
    • Add a section that maps the index to action, not just meaning.
    • Link forward to the next decision page and back to the conceptual page that supports it.

    Key takeaways

    • Confirm stability and MSA before trusting any capability index.
    • Report the study method (short-term vs long-term) alongside the index — a number without context is not auditable.
    • Use the calculator to compute, this hub to interpret, and the industry pages to compare against the actual requirement for the part.
    • If Cpk is low, decide whether variation or centering is the driver before choosing corrective action.

    Frequently asked questions

    Knowledge directory

    Every page in the cluster, grouped by intent. Start with interpretation when you have a calculated value; start with industry when you have a customer requirement; start with process-state when the result looks wrong.

    Capability Interpretation

    Start here when you already have a number. These pages must do more than define Cp or Cpk — they should explain what the index does, what it does not prove, what assumptions are required, and what decision follows next.

    Cp vs Cpk: Difference, Formulas, and Interpretation
    Cp measures potential capability, Cpk measures actual capability including centering. Learn the difference, formulas, and how to interpret both for production processes.
    Process Capability Index Explained
    The process capability index family (Cp, Cpk, Pp, Ppk) measures how well a process meets specification. Definitions, formulas, and target values explained.
    Process Capability Analysis: Step-by-Step
    Process capability analysis evaluates whether a stable process meets specification. Learn the data, normality check, sigma estimation, and reporting steps.
    Process Capability Study
    A process capability study collects 25–100+ measurements from a stable process to compute Cp, Cpk, Pp, and Ppk for PPAP and SPC reporting.
    Negative Cpk: Causes and Corrective Actions
    A negative Cpk means the process mean lies beyond the nearest specification limit. Learn what it implies, verify the study, and choose the appropriate containment response.
    Cpk Below 1: Risk, PPM, and Recovery
    A Cpk under 1.00 means the process produces measurable defects. Learn the PPM rate, sigma level, and the actions required to bring Cpk above target.
    Cpk = 1.00: The Bare-Minimum Capability
    Interpret Cpk 1.00, including its 3σ nearest-limit distance, statistical assumptions, expected PPM, and why it provides little operating margin.
    Cpk = 1.33: The 4σ Target
    Interpret Cpk 1.33, its approximate 4σ nearest-limit distance, expected defect rate assumptions, and common—but not universal—production targets.
    Cpk = 1.67: The PPAP Launch Target
    Interpret Cpk 1.67, its approximate 5σ nearest-limit distance, expected PPM assumptions, and its use in some customer launch requirements.
    Cpk = 2.00: Six Sigma Capability
    Interpret Cpk 2.00, its 6σ nearest-limit distance, modeled PPM with and without a 1.5σ convention, and when demanding targets may be specified.
    Cp = 1.00: Spread Equals Tolerance
    Cp = 1.00 means the process spread (6σ) exactly equals the tolerance width. There is no safety margin even when perfectly centered.
    Cp = 1.33: Minimum Potential Target
    Cp 1.33 means the tolerance is approximately 1.33 times the natural 6σ spread. Interpret it with centering, stability, and the applicable requirement.
    Cp = 1.67: PPAP-Level Potential
    Cp 1.67 describes strong potential capability, but PPAP acceptance and the required index depend on the customer-specific requirement and study method.
    Cp = 2.00: Six Sigma Potential
    Cp = 2.00 corresponds to Six Sigma potential capability: the process spread is half the specification width.

    Industry Requirements

    Use these pages to map a calculated result to the real acceptance criterion that governs the part: CSR, drawing note, Control Plan, PPAP protocol, validation protocol, or internal standard. Common numbers such as 1.33 and 1.67 are reference points, not universal rules.

    Automotive Cpk Requirements
    Understand common automotive capability reference values and how customer-specific requirements, PPAP, the drawing, and the Control Plan determine the actual target.
    CNC Machining Cpk Requirements
    Understand how drawings, customer requirements, machine studies, tool wear, thermal drift, sampling, and production conditions determine CNC capability acceptance.
    Injection Molding Cpk Requirements
    Plan injection-molding capability studies around customer criteria, cavity-specific results, shrinkage, process settings, and material-lot variation.
    Medical Device Cpk Requirements
    Understand how statistically justified capability criteria can support medical-device process validation without inventing universal ISO or FDA Cpk thresholds.
    Aerospace Cpk Requirements
    Understand aerospace capability planning for key characteristics while separating AS9100 requirements from customer- and program-specific numerical targets.
    Stamping Cpk Requirements
    Plan stamping capability studies around customer criteria, die wear, coil-lot variation, press conditions, springback, and measurement strategy.
    Welding Cpk Requirements
    Plan welding capability studies for measurable strength or geometry characteristics while preserving qualification, validation, NDT, and customer requirements.

    Process State

    These pages diagnose what the index is telling you about the process itself: off-center mean, too much variation, drift, unstable streams, mixed cavities, or a measurement problem that invalidates the result.

    Process Mean Outside Specification
    When the process mean lies outside a specification limit, Cpk is negative. Verify the study, estimate risk under stated assumptions, and follow the approved reaction plan.
    Process Variation Too High
    When Cp is low the process spread does not fit the tolerance. Diagnose the common-cause variation sources before chasing centering.
    Process Not Centered
    When Cpk is far below Cp, the mean has drifted off target. Causes include tool wear, offsets, fixturing, and setup. Step-by-step recovery.
    Process Capable and Centered
    A capable, centered process (Cpk ≥ 1.33 with Cp ≈ Cpk) is the target state. Move from corrective action to predictive monitoring.
    Process Capability Improvement
    Practical playbook to improve Cp and Cpk: separate centering from variation, attack dominant variation source, validate with a fresh study.
    High Defect Rate Process
    If your process is producing thousands of PPM, Cpk is well below 1. Use the calculator to quantify the gap and prioritize containment.
    Process Producing Defects: A Capability Diagnosis
    When defects are being produced, capability analysis tells you whether centering, variation, or both are to blame — and how to fix it.

    Capability Conversion

    Use these pages when scorecards, customers, or internal systems talk in different units. They translate between Cpk, Cp, Pp, Ppk, sigma level, and predicted PPM under explicit assumptions.

    PPM to Cpk Conversion
    Convert nearest-tail PPM to an equivalent Cpk under a stable normal model. Understand why total two-sided PPM cannot be converted without additional assumptions.
    Cpk to PPM Conversion
    Convert Cpk to modeled nearest-tail PPM under a stable normal distribution, with clearly labeled assumptions for centered two-sided estimates.
    Sigma Level to Cpk Conversion
    Convert sigma level (Z-score) to Cpk: Cpk = sigma-level / 3 (short term) or (sigma-level − 1.5) / 3 (with the Six Sigma shift).
    Cpk to Sigma Level Conversion
    Convert Cpk to sigma level (Z-score). Cpk × 3 gives the short-term sigma level; add 1.5 for the long-term Six Sigma convention.
    Cp vs Pp: Short-Term vs Long-Term
    Cp uses within-subgroup (short-term) sigma. Pp uses overall (long-term) sigma. The difference quantifies process drift.
    Cpk vs Ppk: Differences and Reporting
    Cpk uses short-term within-subgroup sigma; Ppk uses long-term overall sigma. Customers usually want Ppk for ongoing production.
    Process Capability Formulas
    All four process capability formulas in one place, with worked examples and the difference between within-subgroup and overall sigma.