How to use this tool
- Keep measurements in original time order; do not sort them by magnitude.
- Paste readings or upload a local CSV/XLSX file and select the measurement column.
- Enter the engineering specification limits, not control limits.
- Review the I-chart, capability estimates and stability/sample warnings together.
Calculation method and formula
For individual observations, estimated within σ = mean moving range / 1.128. Cp = (USL − LSL)/(6σwithin), Cpk = min[(USL − mean),(mean − LSL)]/(3σwithin). Pp and Ppk use the overall sample σ.
In a ten-observation example with LSL 9.5 and USL 10.5, the illustrative dataset produces Cp ≈ 1.611 and Pp ≈ 1.414. Ten values are too few for confident process capability conclusions.
What the calculation does not cover
- A capable process must also be statistically stable; Cp/Cpk alone are not a production acceptance decision.
- I-MR assumes a specific within-sigma estimate and time ordering; subgroup or non-normal data require different methods.
- At least 50 observations are recommended here; verify normality and measurement system adequacy independently.
Reference: NIST Engineering Statistics Handbook
Common questions
What is the difference between Cpk and Ppk?
Cpk uses a within-process variation estimate; Ppk uses overall sample variation, which can include shifts and longer-term changes.
Can I upload a spreadsheet privately?
CSV/XLSX input is read in the browser. XLSX requires downloading an Excel parsing library, but the file data are not sent to CHAAAM.
Engineering use: This calculator is a transparent estimate for its defined model; confirm units, applicability, standards and engineering decisions independently.