Normal Process Capability (CpK): Calculate and Interpret Process Capability
What is Process Capability (CpK)?
CpK measures how well a process fits inside its specification limits: Cp = (USL - LSL) / (6 * sigma) reflects potential capability from spread alone, while CpK = min(CPU, CPL) combines location and spread, with CPU = (USL - mu) / (3 * sigma) and CPL = (mu - LSL) / (3 * sigma). The gap between Cp and CpK shows how far the process center sits from the specification midpoint. A CpK of 1.33 or higher is generally considered capable.
When to Use It
Use normal CpK for variable-data capability assessment whenever the data follow a normal or near-normal distribution: new product introduction, ongoing production monitoring, supplier audits and PPAP submissions. Before calculating, confirm the process is in statistical control using SPC charts, because capability indices are meaningless for an out-of-control process. Use a minimum of 25-30 observations for a stable estimate, and switch to a non-normal capability tool when data are skewed.
How to Use It (Step by Step)
Enter your specification limits and paste the measurement data; the tool computes Cp, CpK, Pp, PpK, CPL and CPU with a capability grade and histogram. First compare Cp with CpK to judge centering, then read the capability grade against the customer's requirement. Check the estimated-sigma basis (within-subgroup or overall) and confirm control-chart stability before reporting the result to a customer.
Key Formulas / Example
Cp = (USL - LSL) / (6 * sigma), CPU = (USL - mu) / (3 * sigma), CPL = (mu - LSL) / (3 * sigma), and CpK = min(CPU, CPL). With USL = 10.0, LSL = 4.0, mu = 7.5 and sigma = 0.5, CPU = 2.5 / 1.5 = 1.67 and CPL = 3.5 / 1.5 = 2.33, so CpK = 1.67, indicating a capable and reasonably centered process.