INVESTIGATION OF CHARACTERISTIC MEASURES FOR THE ANALYSIS AND SYNTHESIS OF PRECISION-MACHINED SURFACES

Error prediction and control are key factors in precision machining. T hese factors rely on the development of formal approaches for analyzin g and characterizing error sources in manufacturing. One such approach is the development of mathematical measures of precision, where preci sion, in this context, is defined as surface variations of manufacture d part profiles. in this paper, we discuss a novel investigation of fo ur mathematical measures. These four methods, namely the autocorrelati on function, the Fourier spectrum, a fractal-wavelet representation, a nd the Karhunen-Loeve expansion, are applied to surfaces produced from grinding processes. The first two methods provide a basis for the inv estigation, as they are commonly used in the literature for qualitativ e signal characterization of manufacturing surfaces. However, the frac tal-wavelet method and Karhunen-Loeve expansion have never been applie d to the analysis and synthesis of surface variations. While other fra ctal methods have been used to characterize surface-finish variations, a wavelet formalism is a new approach, especially at the scales of bo th surface finish and tolerances. A combination of the first three tec hniques is shown to give a proper minimum set of characteristic precis ion measures for representing grinding surfaces. This combination is a clear contribution to the field of analysis of surface characteristic s. it is also shown that the Karhunen-Loeve technique is a novel alter native to represent surface errors. The existence of characteristic me asures of surface precision should aid designers in choosing process a nd design parameters and in comparing the precision between competing machining processes.