Resources

What's a good Cpk?

Cpk tells you how well a process fits inside its tolerance, accounting for both spread and centering. It compares the distance from the process average to the nearest spec limit with three standard deviations of the process.

The usual targets

• Below 1.0: the process is making parts outside tolerance. Inspect 100% while you fix it.
• 1.0 to 1.33: marginal. It works on a good day.
• 1.33: the common minimum for an established process, about 63 ppm out of spec if centered and normal.
• 1.67 or more: the target many aerospace and automotive customers set for key characteristics and new processes.

Cp vs Cpk, Cpk vs Ppk

Cp ignores centering: it's the best Cpk you could get if you centered the process. If Cp is high and Cpk is low, adjust the aim before buying a new machine. Cpk uses short-term (within-subgroup) variation; Ppk uses the overall standard deviation of all the data. A big gap between them means the process drifts between subgroups: tool wear, temperature, material lots, or shift changes.

Three ways Cpk lies

1. The process isn't stable. Check a control chart first; capability of an unstable process predicts nothing.
2. Too few points. Thirty readings give a very uncertain Cpk; 100 or more is better.
3. The gage is too coarse. If your measurement system eats most of the tolerance, Cpk is measuring the gage. Run a Gage R&R.

Try the free Cp/Cpk calculator.

General guidance, not legal advice. Standards are referenced by name only. Written by Matt Norman, Norman Quality Tools (Veteran-Owned & Operated).