Capability summary for many characteristics: guide and PPAP case study

Try it while you read

Open the free Capability Summary tool and click Load sample: it reproduces the case study below.

Theory: one table instead of twenty studies

A machined part often has 10 to 50 drawing characteristics. For a production part approval (PPAP) or a launch review the customer wants to see the capability of every special characteristic, and the quality engineer wants to see at a glance where the problems are. The capability summary applies the same calculation as the full capability study to every column: σ within from rational subgroups, σ overall from all values, the four indices, the expected PPM and a normality test.

The result is ranked by Ppk so the weakest characteristic is on top. A bar chart compares Cpk and Ppk of all characteristics against the required level, which makes drift (Ppk well below Cpk) visible immediately.

Key point

The summary is a triage tool. A characteristic that fails here still needs its own study with control charts, because the reason (unstable, off-centre or too wide) decides the action.

Step by step

Measure the same parts for all characteristics, typically 100 parts in subgroups of 5 from a run at rate.
Name the columns directly in the header of the table, fill in the LSL and USL rows (Target is optional) and paste the measurements underneath.
Set the subgroup size and the requirement: 1.67 for an initial PPAP study (Ppk), 1.33 for ongoing production (Cpk).
Read the table and the chart. Status is Capable, Not capable or Not stable.
Plan actions for every characteristic below the requirement and document them with the PPAP submission.

Case study: five characteristics of a shaft

For a PPAP run, 100 shafts are measured on five characteristics (the Load sample data set). Requirement: Cpk and Ppk ≥ 1.33.

Characteristic	Cp	Cpk	Ppk	PPM	Status
Ø25 h7	1.18	0.97	0.96	1,928	Not capable
Length 120	1.34	1.25	1.32	42	Not capable
Ø12 bore	1.08	0.98	1.03	1,078	Not capable
Runout (upper limit only)	–	2.02	2.10	0	Capable
Groove depth	1.31	0.92	0.93	2,591	Not stable

Reading the table

Groove depth is the weakest and also unstable: first find the special cause on the control chart.
Ø25 h7 has Cp 1.18 but Cpk 0.97: it is off centre and too wide. Centring alone would not be enough.
Length 120 is just below the limit (Ppk 1.32, 42 PPM): a small centring correction is likely sufficient.
Runout is comfortably capable.

Result

Only 1 of 5 characteristics is capable, so the PPAP cannot be submitted as is. The team prioritises groove depth and Ø25, contains with 100 % inspection, and reports the plan in the launch review.

Common mistakes

Mixing parts from different cavities or fixtures in one column.
Using the summary alone for approval without looking at the control charts.
Wrong limits for one-sided characteristics (entering 0 as LSL instead of leaving it empty).

Frequently asked questions

Can I mix one-sided and two-sided characteristics?

Yes. Leave LSL empty for characteristics like runout or roughness that only have an upper limit. Cp and Pp are then not shown, Cpk and Ppk are the one-sided indices.

Why is a characteristic marked "Not stable" even though Ppk is shown?

Its control chart has special causes. The indices are calculated for reference but should not be used for approval until the instability is understood.