Add "reproducible research" page

This commit is contained in:
Thomas Dehaeze 2025-04-14 21:57:33 +02:00
parent a8d0fa1e30
commit 80d41cf79d

View File

@ -185,40 +185,40 @@ Prof. Gérard Scorletti\newline
Dr. Olivier Mathon\newline
European Synchrotron Radiation Facility (Grenoble, France)
* TODO Abstract
* Abstract
:PROPERTIES:
:UNNUMBERED: notoc
:END:
The recent advent of the 4th generation light sources in synchrotron radiation facilities has resulted in X-ray beams that are 100 times more brilliant with the capability to be focused down to smaller, sub-micron sizes.
While these advancements open unprecedented scientific opportunities, they simultaneously present significant challenges, especially regarding end-stations that must achieve enhanced sample positioning accuracy and scan speeds.
The advent of $4^{\text{th}}$ generation synchrotron light sources has yielded X-ray beams with a 100-fold increase in brilliance and sub-micron focusing capabilities, offering unprecedented scientific opportunities while necessitating end-stations with enhanced sample positioning accuracy.
At the European Synchrotron (ESRF), the ID31 beamline features an end-station for positioning samples along complex trajectories.
However, its micrometer-range accuracy, limited by thermal drifts and mechanical vibrations, prevents maintaining the point of interest on the focused beam during experiments.
At the European Synchrotron Radiation Facility (ESRF), the ID31 beamline is equipped with an end-station designed for positioning samples along complex trajectories.
A key experimental application is diffraction-tomography, wherein the beam is focused on the sample, which is continuously rotated.
However, the accuracy of this end-station is currently limited to the micrometer range due to thermal drifts, mechanical vibrations, and insufficient precision of various mechanical guiding elements.
Consequently, it fails to maintain the point of interest of the sample on the focused beam throughout experiments.
This thesis aims to develop a system for actively stabilizing the sample's position in the nanometer range while the end-station executes complex trajectories.
The system comprises an external metrology measuring the sample's position, an high dynamic stabilization stage fixed between the end-station and the sample, and dedicated control architecture.
The development of such a system presents several challenges that are addressed in this thesis.
The first challenge relates to the design methodology.
The performance of this complex mechatronic system may be affected by various phenomena.
Several dynamical models of increasing complexity and accuracy were employed to predict performance and anticipate design flaws early in the project.
This methodical design approach facilitated convergence to a solution that fully satisfies the requirements.
The second challenge stems from the control requirements, specifically the need to stabilize samples with masses ranging from 1 to 50 kg, which necessitated the development of specialized control architectures.
Finally, experimental validation was performed through the construction and testing of the Nano Active Stabilization System on the ID31 beamline.
The thesis demonstrates the feasibility of enhancing the positioning performance of an existing end-station through the implementation of an active stabilization system, thereby contributing to the advancement of experimental capabilities in synchrotron radiation facilities.
To address this limitation, this thesis aims to develop a system for actively stabilizing the sample's position down to the nanometer range while the end-station moves the sample through the beam.
The developed system integrates an external metrology for sample position measurement, an active stabilization stage mounted between the end-station and the sample, and a dedicated control architecture.
The design of this system presented key challenges, first of which involved the design process.
To effectively predict how this complex mechatronic system would perform, a series of dynamical models with increasing accuracy were employed.
These models allowed simulation of the system's behavior at different design stages, identifying potential weaknesses early on before physical construction, ultimately leading to a design that fully satisfies the requirements.
The second challenge stems from control requirements, specifically the need to stabilize samples with masses from $1$ to $50\,\text{kg}$, which necessitated the development of specialized robust control architectures.
Finally, the developed Nano Active Stabilization System underwent thorough experimental validation on the ID31 beamline, validating both its performance and the underlying concept.
\vspace{-1em}
\begingroup
\let\clearpage\relax
\chapter*{Résumé}
\endgroup
* TODO Résumé
:PROPERTIES:
:UNNUMBERED: notoc
:END:
L'avènement des sources de lumière synchrotron de $4^{\text{ème}}$ génération a produit des faisceaux de rayons X avec une brillance multipliée par 100 et des capacités de focalisation sub-microniques, offrant des opportunités scientifiques sans précédent tout en nécessitant des stations expérimentales avec une précision de positionnement d'échantillons améliorée.
À l'Installation Européenne de Rayonnement Synchrotron (ESRF), la ligne de lumière ID31 dispose d'une station expérimentale conçue pour positionner des échantillons le long de trajectoires complexes.
Cependant, sa précision de l'ordre du micromètre, limitée par des effets tels que les dérives thermiques et les vibrations mécaniques, empêche de maintenir le point d'intérêt sur le faisceau focalisé durant les expériences.
Pour remédier à cette limitation, cette thèse vise à développer un système permettant de stabiliser activement la position de l'échantillon pendant que la station expérimentale déplace l'échantillon à travers le faisceau.
Le système développé intègre une métrologie externe pour la mesure de la position de l'échantillon, une platine de stabilisation active montée entre la station expérimentale et l'échantillon, et une architecture de contrôle dédiée.
La conception de ce système présente des défis majeurs, dont le premier concerne le processus de conception lui-même.
Pour prédire efficacement les performances, une série de modèles dynamiques ont été utilisés.
Ces modèles ont permis de simuler le comportement du système aux différentes étapes de conception, identifiant ainsi les limitations potentielles, pour aboutir à une conception répondant aux spécifications.
Le deuxième défi provient des exigences de contrôle, notamment la nécessité de stabiliser des échantillons dont la masse peut varier de $1$ à $50\,\text{kg}$, ce qui a nécessité le développement d'architectures de contrôle robustes.
Enfin, le Système de Stabilisation Active développé a fait l'objet d'une validation expérimentale sur la ligne de lumière ID31, validant à la fois ses performances et le concept sous-jacent.
* Acknowledgments
:PROPERTIES:
@ -262,7 +262,22 @@ And to Juliette, for being incredibly supportive through the inevitable tough ti
:UNNUMBERED: notoc
:END:
The foundation of this PhD thesis is built upon the principles of reproducible research.
Reproducible research is the practice of ensuring that the results of a study can be independently verified by others using the original data, code, and documentation.
This approach was adopted to increase transparency and trust in the presented research findings.
Furthermore, it is anticipated that the methods and data shared will facilitate knowledge transfer and reuse within the scientific community, thereby reducing research redundancy and increasing overall efficiency.
It is hoped that some aspects of this work may be reused by the synchrotron community.
The fundamental objective has been to ensure that anyone should be capable of reproducing precisely the same results and figures as presented in this manuscript.
To achieve this goal of reproducibility, comprehensive sharing of all elements has been implemented.
This includes the mathematical models developed, raw experimental data collected, and scripts utilized to generate the figures.
For those wishing to engage with the reproducible aspects of this work, all data and code are freely accessible in *add zenodo link*.
The organization of the code mirrors that of the manuscript, with corresponding chapters and sections.
All materials have been made available under the MIT License, permitting free reuse.
This approach represents a modest contribution toward a more open, reliable, and collaborative scientific ecosystem.
* Grants :ignore: