This ph. D. project was aimed at developing and validating techniques for dimensional metrology of: miniaturized items, microsystem components, and surfaces. In particular the study was focused on techniques based on: AFM-CMM integration and Scanning Electron Microscopy (SEM).
Development of a Large range AFM instrument was addressed, starting from an existing set up with an AFM scan head mounted on the z axis of a CMM. This instrument had already been used for performing Surface Mapping. This is a measuring procedure, consisting of merging AFM data sets, recorded after successive repositioning of the AFM through the CMM axes. However, the existing manual stitching of images was affected by heavy inaccuracies. During the Ph.D. work, stitching software was developed, which performs stitching of AFM images based on partial overlapping and cross correlation. This method allows compensation of the (large) CMM positioning errors.
A geometrical (three dimensional) model, for the Large range AFM was produced and calibration issues discussed following the three dimensional approach. Furthermore, a novel measuring procedure, based on two images, for eliminating the effects of vertical drift was proposed and the principle demonstrated on software gauges. Simulations of Surface Mapping were done, based on the model developed. Direct performance verification of the Large Range AFM was eventually carried out, and lateral metrology was possible, in the millimeter range, with accuracy in the order of one percent, with this instrument. Uncertainty is dominated by residual non linearity after off line correction.
SEM based stereo-photogrammetry was also studied. A commercially available software package was purchased. The working hypothesis for the package in use was eucentric tilting. This is only an idealization and it is not achievable in real measurements. A positioning procedure that realizes pseudo-eucentric tilting was developed. A model for calculating accuracy of topography calculation was developed and calibration of instrumental independent variables performed accordingly. SEM based stereo-photogrammetry was proved to be more suitable for high aspect ratio structures or rough surfaces rather than smooth ones. However, the method was seen to be affected by high uncertainties as concerns vertical elevation. A novel design for a calibration artefact, suited for testing performance of three dimensional SEM at high magnifications was, proposed and this has resulted into a patent application.
The final part of the thesis is devoted to applications of dimensional metrology to case studies. Three applications are presented, two of them are investigations of surface metrology, while the third is about extraction and evaluation of geometrical features from measurement data files