Bothe, Davis R. (1991, ASQC) International Quality Institute, Inc., Northville, MI
This abstract is an edited version of the author's original. The text of this paper is the same as QICID 9833. Do not order both.
Many methods measure the process capability of a single characteristic of a particular product. These measurements estimate the percent of parts having this characteristic within print specifications. This paper presents a new technique for measuring "product capability" that measures the ability of a product to meet all of the customer's requirements.
A product capability study first determines the probability of each characteristic for a given product being within print specifications, then calculates the combined probability that all characteristics are within specifications. This probability can be converted into a "product Cpk" value so different products can be compared and improvements in product capability monitored.
This method works for attribute data as well as for as many characteristics as a product may have. Examples are given involving the determination of a product capability index for two products, one with three different characteristics, the other with two. The three characteristics of the first product all involve variable data, two with bilateral specifications, the third with a unilateral specification. The second product has one characteristic measured with variable data and the other with attribute data. Assumptions relating to the independence of the individual characteristics are discussed.
The product Cpk concept may be extended to determine the capability of an entire assembly (or machining) line to produce products meeting all requirements.
Machinery,Product development,Variation,Capability study,Attributes control charts,Statistics