A Simple Measurement Capability Technique


Enos, Lesley   (1990, ASQC)   Alpha Industries, Inc., Woburn, MA

Annual Quality Congress, San Francisco, CA    Vol. 44    No. 0
QICID: 9522    May 1990    pp. 634-639
List $10.00
Member $5.00

This article is not available online. Contact us to receive a scan of the archive, in PDF format.

Article Abstract

The major goal in the work of quality improvement professionals is to reduce variation and produce consistency in products and services. Understanding the causes and taking steps to reduce variation is the foundation to improving customer satisfaction. First attempts to understand product variation genera6lly begin with control charts and/or Pareto analysis of existing data. However, one key element is commonly overlooked - the measurement system. The assumption that the measurement system is "in control" can be misleading and dangerous.

Typically, when a measurement system is analyzed for capability, a quick study is performed neatly combining product, process and operator components. Ideally, to determine the accuracy of the measurement system, an industry standard is purchased and compared to the measurement system. After the gauge study is performed, the data is entered into a computer program and the resulting measurement variation is displayed. This method looks appealing since it can be quick and easy while combining many components at one time. However, erroneous results are common, blaming measurement variation for what is actual product variation.

At Alpha Industries, as in many semiconductor facilities, microscopic tolerances in the angstroms is common. Measurement capability plays a major role when working in such tight tolerances. The goal of the measurement system contributing no more than 10% of the total variation can be virtually impossible with small tolerances. There are times where state-of-the-art measurement equipment used to measure tight tolerances is extremely expensive or not yet invented. And to complicate matters, industry standards to determine measurement accuracy and precision also have not been invented. A measurement system guide has been developed at Alpha Industries which addresses these problems and is used towards reducing measurement variation.

This paper will discuss an existing measurement capability system which determines the cause of variation and answers the following questions with examples of actual measurement studies:

  • - What happens when the measurement system exceeds 10% of the total product tolerance? or,
  • - What happens if it is physically impossible for the measurement system to be within 10% of the total product variation? Actual examples of control chart techniques and analysis concentrating on the reducing measurement variation are provided.
  • - What is a simple analysis of variance technique which determines the root causes of measurement variation? Examples of operator training decreasing measurement variation.
  • - How is the smallest allowable tolerance for a measurement system determined? Examples of smallest tolerance capability determined and using this knowledge in product design.



Browse QIC Articles Chronologically:     Previous Article     Next Article

New Search

Featured advertisers

ASQ is a global community of people passionate about quality, who use the tools, their ideas and expertise to make our world work better. ASQ: The Global Voice of Quality.