Methods for evaluation of systematic geometric deviations in machined parts and their relationships to process variables

Increasing demands for precision-machined parts put a greater emphasis on a chieving a better understanding of the relationships between manufacturing process variables and deviations from perfect geometric forms. To achieve t his understanding, it is important to have high-quality metrology data on p art features, which (in some sense) span the range of variability in form f or the process under study. The problem is complex, involving the machine t ool itself, materials, fixtures, cutters, feeds, speeds, and many other fac tors. In this report, a method is introduced that can identify key deviatio ns from perfect form and can elucidate their dependence on some of the fact ors enumerated above. This work presents two useful models for form errors of cylindrical features and develops special cases of those models to suit specific requirements. One is an analytical model based on Chebyshev polyno mials to model the axial errors and Fourier series to model angular depende ncies. The second modeling approach uses the techniques of principal compon ent analysis to extract a basis set of characteristic shapes directly from the measurement data. This report includes a full development of the mathem atical basis for the analysis and concludes with some application examples. (C) 1999 Elsevier Science Inc. All rights reserved.