Rational Subgrouping Strategies for Multi Cavity Injection Molding Systems
Rational subgrouping in multi-cavity tooling requires tracking cavities as distinct streams to avoid pooling offsets that inflate variance and distort Cpk.
Shewhart Control Limits
Meaning
Statistically calculated horizontal boundaries set at three standard deviations above and below a process average separate common cause variation from assignable disturbances on control charts. Manufacturing operators apply shewhart control limits to establish whether a repetitive process functions in statistical control without making assumptions about engineering tolerance requirements. These boundaries apply exclusively to rational subgroup data gathered under consistent physical conditions, stopping where external parameter adjustments or deliberate machine changes occur.
Threshold Calculations
Empirical boundary calculations avoid emotional over-adjustment by machine operators. Establishing shewhart control limits requires sampling output over regular rational intervals and estimating common cause dispersion through average ranges or standard deviations. The upper and lower boundaries reside exactly at the center line plus or minus three sigma, capturing ninety-nine point seven percent of normal variation.
Adjusting equipment inside these limits introduces excess variability, destabilizing otherwise steady operations.
Tolerance Contrast
Specification bands dictate commercial customer acceptance, whereas statistical process bounds reflect physical machine stability. Conflating customer tolerance lines with shewhart control limits leads to premature production approval of volatile lines. A machine operating within customer limits may still exhibit uncontrolled shifts that generate out-of-spec scrap when ambient conditions drift.
Demonstration audits mandate proving statistical stability via control boundaries prior to granting high-volume production sign-off.
Process Stability
Operational control charts provide the baseline documentation for production capability indices like Cp and Cpk. When points drift beyond shewhart control limits, operators initiate immediate troubleshooting to uncover special causes like broken tools, defective material lots or operator errors. Verifiable process control minimizes scrap rates, stabilizes continuous yields and guarantees baseline repeatability across high-volume factory lines.