Process Capability Index
Why is Cpk usually more important than Cp?
Cpk captures both dispersion and centering, while Cp only measures spread relative to spec width. On the shop floor, centering errors from tool changes, offset edits, or thermal drift are common, making Cpk the more operationally useful metric.
What does a Cpk of 1.00 mean?
A Cpk of 1.00 means the process spread fits within the tolerance band at the current centering, leaving little margin for normal variation. NIST notes that capability interpretation depends on how stable process output compares with specification limits.
What does a low Cpk imply on a machine?
A low Cpk indicates either the process mean is shifted off target, variation is too high, or both. In practice, this often traces to tool wear, poor workholding, thermal drift, or inconsistent material condition.
Process Capability Index (Cpk) is a statistical measure that evaluates how well a stable manufacturing process produces parts within specified tolerance limits. It accounts for both process spread and centering of the mean between the limits, making it the most practical index for CNC machining and millwork quality assurance.
In CNC machining and millwork, process capability is used to determine whether a machine, toolpath, fixture, cutter, or sanding/edgebanding operation can repeatedly hold critical dimensions like bore diameter, pocket depth, panel thickness, or edge-band overhang within customer tolerances. The shop-floor procedure involves running a stable process, collecting measured parts, calculating standard deviation and mean, and then computing Cp (spread vs. spec width) and Cpk (spread plus centering). A high Cp with low Cpk indicates precise but off-center output, while low Cp means variation exceeds tolerance. This analysis is applied to features like drilled hole location, turned diameters, milled slots, and multi-operation tolerance stacks. In millwork, capability is vital for panel thickness, reveal consistency, and glue-line position, as humidity, feed speed, cutter wear, or adhesive temperature shifts can reduce Cpk below acceptable levels.
Unstable process data: Calculating capability while the CNC spindle is warming up, a cutter is wearing, or a fixture is slipping yields misleading Cp/Cpk because indices require statistical control.
Confusing Cp with Cpk: A process can have strong spread (high Cp) but fail Cpk if the mean shifts toward one limit, leading to repeat rejects on one side of tolerance after offset changes or thermal growth.
Biased sampling: Measuring parts only after setup or from one shift misses wear, chip buildup, or environmental drift, giving false confidence in CNC and millwork validation runs.