1 Manufacturing Engineering, Department of Mechanical Engineering, Technical University of Denmark2 Department of Mechanical Engineering, Technical University of Denmark3 Materials and Surface Engineering, Department of Mechanical Engineering, Technical University of Denmark4 Department of Management Engineering, Technical University of Denmark5 Danish Fundamental Metrology
The present work deals with a study concerning 3D-SEM metrology as a tool for coordinate measurements at the nanometer scale. The relevance of 3D-SEM, based on stereophotogrammetry technique, has been highlighted with respect to the other measuring instruments nowadays available and the main issues to be addressed concerning uncertainty evaluation have been discussed. Most recent developments in the field of micro and nano-metrology, in terms of measuring machines and techniques, are described pointing out advantages and limitations. The importance of multi-sensor and multi-orientation strategy for geometrical reconstructions is discussed through an experimental example, together with point cloud stitching methodology and the currently used algorithms for feature extraction. Theoretical basis of stereo-pair technique, based on two SEM images obtained by tilting the SEM stage of a desired amount, leading to 3D reconstructions, are given and the main phases involved in stereophotogrammetry technique are described underlying the most relevant error sources in the case of 2D and 3D-SEM metrology. An uncertainty evaluation has been thus carried out in accordance with ISO GUM, following a holistic approach, to quantify the influence of the different error sources on the stereo-pair reconstruction procedure. As a case study, a wire gauge with a known reference diameter has been employed. Although stereo-pairs are more commonly obtained trough a SEM stage tilting, a new methodology has been developed based on object rotations inside the SEM chamber, since the item under consideration had a cylindrical shape. This technique permits multi-orientation measurements enabling the reconstruction of the complete object geometry. The main error sources considered, when performing 3D-SEM reconstructions, are point cloud processing and feature extraction, instrument setting parameters and image quality. Moreover, a comparison of the results obtained trough a theoretical and an experimental uncertainty evaluation of stereo-pair technique has been performed. All these effects have been quantified through a series of experimental investigations often based on the Design of Experiments (DOE) approach. A final uncertainty budget table has been produced for the case of multi-orientation reconstructions obtained by applying 3D-SEM technique to three cylindrical items: two reference wire gauges and a hypodermic needle. In the last part of the work, the development and application of two novel multiplestep heights artefacts, intended for 3D-SEM calibration, is addressed. Experimental results of the different step-height values, measured from 3D-SEM reconstructions, are compared with the calibrated ones obtained from reference measurements performed by means of stylus profilometer and with measurements carried out using an Infinite Focus instrument.
Main Research Area:
De Chiffre, Leonardo, Horsewell, Andy, Hansen, Hans Nørgaard, Dirscherl, Kai