Abstract
This paper discusses a statistical technique that can be used to investigate a family of lubricants with multivariate data in search of groups or clusters. A computer method for partitioning n lubricants into g groups is described. From this data grouping and perhaps further tests, a lubrication specialist could select lubricants with desirable properties for specific applications. The technique is illustrated on actual data collected on eight different properties of 21 wire-rope lubricants. All trade names of lubricants and names of properties are coded for the discussion.
Presented at the 26th ASLE Annual Meeting in Boston, Massachusetts, May 3–6, 1971
Notes
Presented at the 26th ASLE Annual Meeting in Boston, Massachusetts, May 3–6, 1971