This process capability calculator tells you whether a process can hold its tolerance. Paste measurements in production order and enter the lower and upper specification limits, and it gives Cp, Cpk, Pp and Ppk with a rating, the expected parts per million out of spec, and a histogram with the limits and a fitted normal curve you can download as SVG. One-sided limits work too. Paste a table of parts and characteristics to calculate them all. It runs in your browser.
| Cpk | Sigma level | Expected ppm out of spec |
|---|---|---|
| 0.67 | 2σ | 45,500 |
| 1.00 | 3σ | 2,700 |
| 1.33 | 4σ | 63 |
| 1.67 | 5σ | 0.57 |
| 2.00 | 6σ | 0.002 |
Cp compares the tolerance width with the process spread; Cpk also accounts for how far the mean is from the nearer limit. A centred process has Cp = Cpk.
1.33 is a common minimum, 1.67 is often required for new or safety-critical characteristics, and below 1.0 the process makes out-of-spec parts.
Cpk uses short-term within-process variation (moving range); Ppk uses the overall standard deviation, so it includes drift between subgroups.
At least 30, and 100 or more for a reliable estimate. Keep them in time order.
No. Everything runs in your browser.
The mean and overall standard deviation come from the data; within-process sigma is the average moving range ÷ d2 (1.128). Cp = (USL − LSL) ÷ 6σwithin and Cpk = min(USL − mean, mean − LSL) ÷ 3σwithin. Pp and Ppk use the overall standard deviation. Expected parts per million outside the limits come from the normal distribution with the overall sigma. One-sided specifications give Cpk and Ppk only. Batch mode reads a row per characteristic with the values in one cell.