A unified methodology for the application of surface metrology
This thesis addresses the growing "divergence" in the field of surface metrology through the presentation of a practical system for unification. A technical and economic review of applied surface metrology is presented, highlighting the problems associated with the many advances in instrumentation - particularly in light of the growing industrial dependence on surface metrology. This background serves as the basis for the development of a scheme whereby surface specification, instrumentation, and analysis can be concisely and completely defined and, more importantly, controlled. Several technical aspects of surface metrology are addressed in the development of the scheme. First the topic of specification and reference geometries is addressed, where it is argued that least squares methods should provide the most stable basis for assessment. Stylus/radius convolution and the associated wavelength transmissions are also considered and experimental investigations are undertaken as to describe their influences on measured data sets. The treatment of unwanted asperities is investigated and a new, robust algorithm developed and presented. The study of wavelength limitation approaches concludes that a sub-set of current methods is technologically acceptable and therefore economically attractive. A review of parameterization concentrates on a means for selecting a "unified" set of parameters and guidelines for the incorporation of future parameters. Finally, it is shown that the this proposed scheme addresses the underlying divergence in surface metrology in a manner which is practical in the context of application, technically justified in the context of standardization, and extensible in the context of further research.