Title :
Analysis of Component Failure Problee Using Partially Ordered Sets
Author_Institution :
Los Alamos Scientific Laboratory, University of California Los Alamos, New Mexico
Abstract :
It is often desirable, due to expense in time and money, to find the cause of component failures directly from available operational data rather than to resort to controlled tests. Such a method for analyzing component failures cause by manufacturing processing problems is illustrated by an application to a group of early failures among Scyllac capacitors. The application of the method proves that the failures were predominantly caused by inadequate drying of the capacitor packs. The method involves the use of partially ordered sets and can be considered as an abstraction of multivariate regression. The advantages and disadvantages of the method with respect to multivariate regression are discussed.
Keywords :
Atomic measurements; Capacitors; Data analysis; Failure analysis; Inspection; Laboratories; Manufacturing processes; Multivariate regression; Performance evaluation; Testing;
Journal_Title :
Nuclear Science, IEEE Transactions on
DOI :
10.1109/TNS.1971.4326349