Uncertainty Estimation for Performance Evaluation of a Confocal Microscope as Metrology Equipment

Both in industry and research, the quality control of micrometric manufactured parts is based on the measurement of parameters whose traceability is sometimes difficult to guarantee. In some of these parts, the confocal microscopy shows great aptitudes to characterize a measurand qualitatively and quantitatively. The confocal microscopy allows the acquisition of 2D and 3D images that are easily manipulated. Nowadays, this equipment is manufactured by many different brands, each of them claiming a resolution probably not in accord to their real performance. The Laser Center (Technical University of Madrid) has a confocal microscope to verify the dimensions of the micro mechanizing in their own research projects. The present study pretends to confirm that the magnitudes obtained are true and reliable. To achieve this, a methodology for confocal microscope calibration is proposed, as well as an experimental phase for dimensionally valuing the equipment by 4 different standard positions, with its seven magnifications and the six objective lenses that the equipment currently has, in the x–y and z axis. From the results the uncertainty will be estimated along with an effect analysis of the different magnifications in each of the objective lenses.

Caracterización geométrica de huellas de dureza Brinell mediante equipos ópticos. Modelo con microscopía confocal

En esta tesis se desarrolla una metodología alternativa para la determinación de la dureza Brinell a partir de imágenes obtenidas mediante microscopía confocal, que se ha mostrado robusta para mejorar los resultados de medición del diámetro en condiciones de reproducibilidad. Las validaciones realizadas evidencian su posibilidad real de implementación, especialmente para la certificación de patrones de dureza. Los estudios experimentales realizados ponen de manifiesto que la medición del diámetro de una huella de dureza Brinell, siguiendo la metodología tradicional, depende de la posición del patrón, de las características del equipo empleado y del propio operador. Dicha medida resulta crítica y las dificultades para identificar el borde de la huella incorporan a menudo una fuente adicional de incertidumbre difícil de soslayar. En esta investigación se han desarrollado dos modelos matemáticos que permiten identificar de forma unívoca el diámetro de la huella en el punto donde se produce el límite de contacto entre el indentador y el material de la probeta durante la realización del ensayo. Ambos modelos han sido implementados en Matlab® y se ha verificado su validez mediante datos sintéticos. Asimismo, se ha realizado una validación experimental sobre patrones de dureza certificados, empleando un microscopio confocal marca Leica, modelo DCM 3D disponible en el Laboratorio de Investigación de Materiales de Interés Tecnológico (LIMIT) de la Escuela Técnica Superior de Ingeniería y Diseño Industrial de la Universidad Politécnica de Madrid (ETSIDI – UPM). Dicha validación ha puesto de manifiesto la utilidad de esta nueva metodología por cuanto permite caracterizar las huellas, estimar las incertidumbres de medida y garantizar la trazabilidad metrológica de los resultados. ABSTRACT This PhD thesis presents an alternative methodology to determine the Brinell hardness from the images obtained by confocal microscopy that has proved to be robust to improve the results of indentation diameter measurements in reproducibility conditions. The validations carried out show the real possibility of its implementation, especially for calibration of hardness reference blocks. Experimental studies performed worldwide show that the measurement of the indentation diameter in a Brinell hardness test depends, when the traditional methodology is applied, on the position of the test block, the equipment characteristics and the operator. This measurement is critical and the difficulties to identify the edge of the indentation often bring an additional source of uncertainty with them that is hard to avoid. In this research two specific mathematical models have been developed to identify unambiguously the indentation diameter at the point where the edge of the boundary between the indenter and the test block is found during the test. Both models have been implemented on Matlab® and their validity has been verified by synthetic data An additional experimental validation with calibrated hardness reference blocks has been carried out using a Leica-brand confocal microscope, model DCM 3D, available in the Laboratory for Research on Materials of Technological Interest (LIMIT in its Spanish acronym) of the Escuela Técnica Superior de Ingeniería y Diseño Industrial de la Universidad Politécnica de Madrid (ETSIDI-UPM). This validation has shown the utility of this new methodology since it allows to characterize the indentation, to estimate the measurement uncertainties and to ensure the metrological traceability of the results.