Statistical Quality Control and a Capability Analysis

Paul Johnson, University of California San Francisco

Ling Huang, Saint Mary’s College, Moraga, CA.

(QC 01)

ABSTRACT

A SAS® macro is used to perform a quality control analysis of an advanced ceramics  production line for Imperial Chemical Industries (ICI) Ceramics in Auburn, California. XBAR and RANGE charts were used to identify possible problems with process performance.

The supplier gave certain bounds for the lower and upper limits of the 10Boron (10B) isotope within the boron carbide loadings (B4C).

It was these 10B loadings that were of interest, since it was questionable whether or not impurities had caused the percentage of the 10Boron Isotope to be outside the limits specified by the supplier. Boron has two isotopes 10B and 11B.

Capability indices are calculated. A second SAS macro is provided to estimate the out-of-spec probability. The Bootstrap is the simulation approach used to estimate this ‘out-of-spec’ probability. This is the probability that the control process is not performing within the lower and upper control limit bounds. These limits are either specified by the user or set equal to 3 standard deviations below and above the process mean. The macros require base SAS, SAS/STAT, SAS/QC and SAS/GRAPH software to run.