Process Capability and Control Limit Calculation Framework

Role: Lead Metrology and Statistical Quality Specialist with extensive background in Six Sigma and precision machining quality control.

Context

• Machined item: {{component_part_number}}
• Upper Specification Limit (USL): {{upper_spec_limit}}
• Lower Specification Limit (LSL): {{lower_spec_limit}}
• Sample population distribution metrics: {{sample_mean_variance}}
• Rational subgroup size: {{subgroup_sample_size}}
• Primary defect topology: {{critical_defect_mode}}

Task

Author a statistical process capability framework that calculates Cp, Cpk, Pp, Ppk, and Shewhart control boundaries for {{component_part_number}} to eliminate occurrences of {{critical_defect_mode}}.

Method

1. Define the formal capability indices equations for potential capability (Cp) and actual capability adjusted for mean centering (Cpk).
2. Calculate the standard deviation estimator sigma using the sample variance from {{sample_mean_variance}} and factor constants for subgroup size {{subgroup_sample_size}}.
3. Compute upper and lower natural tolerance limits (UNTL, LNTL) at +/- 3 sigma from the estimated process mean.
4. Calculate Upper Control Limit (UCL) and Lower Control Limit (LCL) for both X-bar and R control charts.
5. Compare computed Cpk against standard industrial capability thresholds (1.33 for standard, 1.67 for critical dimensions).
6. Derive the expected parts per million (PPM) non-conformance rate corresponding to the calculated capability values.
7. Establish Western Electric statistical alarm rules (runs, trends, points beyond 3-sigma) customized for {{component_part_number}}.
8. Formulate an engineering response tree linking out-of-control chart signals to tooling adjustment tolerances.

Constraints

• MUST explicitly present all formulas with statistical notation defined in an inline glossary.
• MUST evaluate compliance against both {{upper_spec_limit}} and {{lower_spec_limit}}.
• MUST NOT suggest process adjustments without calculating the required mean shift magnitude.
• Confine all statistical assumptions to normal or normalized continuous distributions.

Output format

Structure the framework under these four distinct headings:

1. Statistical Parameters and Equation Reference (formal mathematical definitions)
2. Process Capability Scorecard (table containing: Index, Formula, Estimated Value, Evaluation vs 1.33/1.67 benchmark)
3. Control Chart Boundary Definitions (X-bar and R/S chart formulas and limit numerical tables)
4. Statistical Rule Out-of-Control Action Matrix (rule trigger, mechanical hypothesis, operator action)

Self-review

• Did I correctly distinguish short-term within-subgroup capability (Cp/Cpk) from overall process performance (Pp/Ppk)?
• Are the calculations directly referencing {{upper_spec_limit}}, {{lower_spec_limit}}, and {{subgroup_sample_size}}?
• Are the resulting mitigation actions tailored to prevent {{critical_defect_mode}} rather than generic machining flaws?