Click to rename — the title is used in the PDF report too
Drag the sliders below — watch the curve widen, narrow, and drift off target.
Process (±3σ)TargetProcess meanOut of spec
Cp
1.50
Cpk
1.00
How tight the process spread is versus the spec width. Ignores where the process is centered.
Same spread, but measured against the nearest spec limit — so centering matters.
Off-center direction
Cpk can never exceed Cp — value was clamped to Cp.
Scenarios
Centering shift k
–
0% = on target
How far the process mean has drifted from the target, as a share of the half-tolerance: k = 1 − Cpk/Cp. 0% means perfectly centered; 100% means the mean sits right on a spec limit. This single number is the entire difference between Cp and Cpk.
Mean offset
–
in σ from target
The same drift expressed in process standard deviations (σ). Positive = toward USL, negative = toward LSL — the Cp/Cpk values are identical either way. For reference, the classic Six Sigma model assumes a long-term mean shift of ±1.5σ — which is exactly why "Six Sigma quality" (Cp = 2.0) corresponds to Cpk = 1.5 and 3.4 defects per million.
Expected defects
–
parts per million
Estimated parts per million falling outside LSL or USL, assuming a stable, normally distributed process. Benchmarks: Cpk 1.00 ≈ 2,700 PPM · Cpk 1.33 ≈ 63 PPM · Cpk 1.67 ≈ 0.6 PPM. Real long-term defect rates are usually worse, because the mean drifts over time.
Expected yield
–
within specification
The share of production expected to land inside specification — the complement of the defect rate. Note how deceptive percentages are at high capability: 99.7% yield sounds excellent, yet it still means 2,700 defective parts in every million.