Methodology for the robust design of air bleed systems in aeronautics

Doctorant : Gabriel DE CARVALHO FERREIRA SILVA

Laboratoire INSA : AMPERE

École doctorale : ED160 : EEA (Electronique, Electrotechnique, Automatique)

At the design stage, product engineers try to find the perfect balance between the physical parameters of the components and the performance required to properly design a system. It is therefore very important to understand how the choice of physical parameters at the design stage will impact the dynamic behavior of the final product. The problem requires skills in modeling complex multi-physics systems and control theory.
This work investigates the air bleed system of aircraft in collaboration with Liebherr Aerospace Toulouse. Although the structure of each valve and the arrangement of valves in the air bleed system are well known, the parameters that critically impact performance are difficult to identify. Due to the complexity of the model required to reproduce the behavior of valves, its non-linearity and the high degree of coupling between parameters, existing tools quickly show their limits. That is why the aim of the work presented in this thesis is to propose a new methodology that closely address the dynamic aspects of these valves from the design stage and provide a relevant evaluation of their dynamic performance in relation with the design requirements. This new methodology is presented and described from its theoretical concepts to its application to practical problems; it is based on a specific modeling step followed by a numerically solved stability condition, which is formulated as a constrained optimization based on linear matrix inequalities (LMI). The main originality of this work is based on the use of optimization tools to find, not an optimal parameter value, but the admissible ranges for a set of parameters which guarantee the required dynamic behavior.

 

23/10/2024 à 21h (CEST) / 24/10/2024 à 6h (AEDT)

Doctorant : Alexandru TECHERES

Laboratoire INSA : MATEIS

Ecole doctorale : ED34 : Matériaux de Lyon

The strengthening of Al alloys via Sc and Zr has been well documented in the literature. The formation of L12 Al3Sc and Al3Zr leads to improvements of mechanical properties through formation of precipitates during artificial ageing. However, the scarcity and cost of Sc have always been a challenge towards adoption on an industrial scale. Nonetheless, new mineral deposits discovered in Australia, together with novel processing methods create the expectation that the price of Sc will become more accessible.
In the context of a circular economy, recycling is the norm. However, this introduces impurities in the recycled material which originate from the imperfect sorting of scraps. The precipitation kinetics of Al3Sc in high purity, controlled composition Al alloys is well understood. However, the precipitation process in the presence of impurities such as Fe and Si has not been studied.
Therefore, the problematic of the current thesis focuses of the effect of impurities on the precipitation kinetics in Al-Sc-Zr during artificial ageing at various temperatures. Previous reports in the literature indicate an accelerating effect of Si and there are limited reports on the effect of Fe.
Using a combinatorial method, in this work we investigate the effect of Si on the precipitation kinetics via samples with a macroscopic composition gradient (also called diffusion couples). Additionally, two sets of samples are investigated with different Fe concentration to comparatively study the effect of Fe. The precipitation kinetics is observed indirectly via small angle x-ray scattering (SAXS) and hardness measurements across the samples with macroscopic composition gradient. Furthermore, advanced characterisation techniques such as Scanning Electron Microscopy, Transmission Electron Microscopy and Atom Probe Tomography are employed to analyse conditions determined as critical during the ageing.
A strong acceleration effect was observed, with increased Si, agreeing with previous reports. However, a saturation effect was identified, which seems to depend on the ageing temperature. This is discussed as a function of Si-vacancy interaction in the precipitation of Al3Sc. Modelling in a Kampmann-Wagner Numerical framework was performed assuming various numbers of heterogenous nucleation sites. To match experimental results, it was required to increase the diffusion coefficient of Sc in the presence of Si, further highlighting the acceleration effect.
It was identified via TEM and APT that the presence of Si leads to smaller mean radius and larger volume fraction of precipitates, which promote larger increments of the mechanical properties. Furthermore, it was observed that Si replaces Al in the structure of the precipitates and that the concentration of Si in the precipitates increases with the concentration in the solid solution.
Finally, it was observed that Fe can replace Al in the precipitates too. Moreover, in the presence of Fe, Si is found in higher concentrations inside the precipitate than when Fe is virtually absent. However, only a minimal difference is observed in terms of hardening between the alloys with different Fe content and only early in the ageing process. Analytical strengthening calculations seem to predict well the hardening in a Fe-free alloy but poorly in the Fe-rich one. This indirectly suggests that Fe has an effect on the precipitation kinetics, but more investigations are needed to fully understand the effect of Fe on precipitation in these alloys.

 

Eco-design and processing of polymer composites towards the shielding of EMI by absorption: Processing, structure, and properties relationship

Doctorante : Emna MASGHOUNI

Laboratoire INSA : IMP

École doctorale : ED34 : Matériaux de Lyon

 

Multi-criteria decision making for sustainable engineering systems

Doctorant : Majid BASEER

Laboratoire INSA : CETHIL UMR5008

École doctorale : ED162 MEGA

The trend towards sustainability in the building sector is gaining momentum, driven by environmental regulations and societal will. Although new building projects adhere to established standards, they replace only 1-3% of existing buildings annually. Renovation projects often present challenges such as the presence of uncertain data, varied stakeholder’s interests, and complex sustainability factors. To address these limitations, this thesis aims to integrate uncertainty into the decision-making process. The thesis develops a multi-criteria decision-making (MCDM) framework based on sustainability factors encompassing economic, environmental, social, and technical aspects for energy renovation of existing buildings. In contrast to existing deterministic MCDM methods, this framework incorporates uncertain data, yielding more realistic outcomes. A review of established MCDM methods was conducted, followed by a SWOT analysis to select a suitable method for the research problem. The application of MCDM methods in building energy renovations was examined to understand sustainability factors comprehensively. To address uncertainty, probability distribution and Monte Carlo simulation are integrated with MCDM methods. These tools represent uncertainty, simulate decision-making, and handle ambiguity. The novel MCDM framework, probabilistic ELECTRE Tri, has been developed based on the ELECTRE Tri method, probability distribution, revised Simos method, and Monte Carlo simulation. This framework was specifically developed for the classification of scenarios for building energy renovation. It was validated through a case study of a social housing project in Lyon, France. The results from the probabilistic ELECTRE Tri method were compared with those from ELECTRE Tri, underscoring the significance of incorporating uncertainty in decision-making. The developed framework enhances transparency, adaptability, flexibility, and user understanding, benefiting stakeholders. It is generic and can improve the objectivity and consistency of decision- making. Additionally, it advances sustainable building renovation and aligns with global environmental and energy efficiency goals. Keywords: multi-criteria decision making, ELECTRE Tri, uncertainty, probability distribution, energy renovation, sustainability.

Pilotage d'une chaîne de conversion active et Analyse de Cycle de Vie pour le petit éolien

Doctorant : Adrien PREVOST

Laboratoire INSA : AMPERE

École doctorale : ED160 : EEA (Electronique, Electrotechnique, Automatique)

Cette thèse porte sur l’étude d’une chaîne de conversion active pour le petit éolien. Dans sa configuration initiale, l’éolienne étudiée est équipée d’une génératrice synchrone à aimants permanents raccordée à une batterie via un redresseur à diodes. Dans ce travail, nous proposons une modification du système visant à augmenter son efficacité : remplacer le redresseur à diodes par un redresseur actif, tout en questionnant les impacts environnementaux et sociétaux de cette modification. Dans la première phase du travail, nous avons réalisé un banc d’essais permettant d’émuler le comportement de l’éolienne, afin d’évaluer sa performance en conditions stationnaires et dynamiques. Nous avons ensuite modélisé le système génératrice-convertisseur dans le but de proposer une loi de commande. Pour ce faire, nous avons établi puis validé expérimentalement un modèle réluctant et un modèle par éléments finis de la génératrice. Nous avons ensuite proposé une loi de commande du système avec un redresseur actif qui a montré des gains de performance significatifs en comparaison avec la configuration classique de redresseur à diodes. Nous avons notamment montré que la stratégie pouvait se passer de capteur mécanique de position grâce à un observateur tout en étant robuste aux incertitudes du modèle. Du point de vue environnemental, une Analyse de Cycle de Vie (ACV) du système a montré que les parties les plus impactantes étaient la structure du mât et les batteries. Nous avons proposé un cadre de pensée unifié pour considérer l’optimisation technico-environnementale du système. Enfin, nous avons étudié la communauté du petit éolien auto-construit en France pour comprendre quel effet pourrait avoir l'introduction d'un redresseur actif sur l'appropriation de l'éolienne.

 

 

 

 

Méthodologie de caractérisation socio-organisationnelle des adresses IPs appliquée à la sécurité

Doctorante : Mme Camille MORIOT

Laboratoire INSA : CITI

École doctorale : ED512 : InfoMaths (Informatique et Mathématiques de Lyon)

Internet est un système clé dans la société contemporaine. Il s'agit d'un système complexe réparti entre de nombreuses organisations ayant une variété de rôles et d'intérêts. Depuis leur création, les cyberattaques sont devenues des actifs précieux, car elles donnent aux rivaux des avantages, par exemple dans les domaines politique ou économique. Il est nécessaire d'analyser ces attaques, d'identifier leurs singularités et les mécanismes sur lesquels elles s'appuient afin de les contrer. Cela permettra d'établir des signatures plus précises et plus pertinentes et aidera la conception des contre-mesures. Un des aspects d'analyse des attaques sont les infrastructures utilisées par les attaquants pour générer les attaques. De nombreux outils aujourd'hui permettent de caractériser l'aspect technique des machines qui composent ces infrastructures. Mais comme les attaques ont lieu dans un environnement social, politique, économique et organisationnel, nous revendiquons qu'il est nécessaire d'évaluer ces machines d'un point de vue organisationnel. 

Cette thèse propose une méthodologie originale de catégorisation des adresses IP, à l'aide de 6 étiquettes décrivant deux axes : un axe technologique et un axe organisationnel. Nous proposons également un outil d'investigation, IPSeen, qui implémente cette méthodologie, en affectant les étiquettes aux adresses IP. Il s'appuie sur différentes sources de données : Wikidata, RDAP, Onyphe, GeoIPLite. Deux versions d'IPSeen sont proposées et évaluées dans ce manuscrit. Ces deux versions se différencient par leur rapidité et leur niveau de précision. 

Enfin, nous appliquons notre méthodologie à un ensemble de données réelles de suivi d'infrastructure de type command and control. L'analyse produite propose une description des infrastructures des organisations qui maintiennent les machines participant aux infrastructures d'attaques. Nous montrons que notre approche apporte un éclairage essentiel sur la compréhension des attaques, en complément des nombreuses caractérisations techniques par ailleurs disponibles.

 

 

« Advanced signal processing methods for source identification using references »

Doctorant : Muhammad Nabil Mustafa ALBEZZAWY

Laboratoire INSA : LVA

École doctorale : ED162 : MEGA de Lyon (Mécanique, Energétique, Génie civil, Acoustique)

Les techniques de référence à rang réduit sont couramment employées pour résoudre les problèmes d’extraction de source et de resynchronisation de champs physiques, lorsque le nombre de références dépasse celui des sources incohérentes. Dans ce cas, la matrice inter-spectrale devient mal conditionnée, rendant la solution des moindres carrés invalide. Bien que la décomposition en valeurs singulières tronquée (DVST) soit utilisée pour résoudre ce problème, elle n'est valable que pour un bruit scalaire sur les références. De plus, il est difficile de définir un seuil de troncature lorsque les valeurs singulières diminuent progressivement. Cette thèse propose une solution nommée technique de référence maximale-cohérente (RMC), basée sur la recherche d’un ensemble de références virtuelles maximalement corrélées avec les mesures de champ. Cette technique est optimale, surtout en présence d’un bruit corrélé sur la référence. Cependant, elle nécessite également une troncature des valeurs propres, exigeant la connaissance ou l’estimation préalable du nombre de sources incohérentes, un problème inverse mal posé et peu étudié. La thèse présente trois méthodes d’énumération des sources applicables à toutes les techniques de référence : un test du rapport de vraisemblance contre le modèle saturé, une technique de bootstrap paramétrique et une approche de validation croisée. Une étude comparative basée sur des données numériques et expérimentales montre deux résultats importants. D'abord, le nombre de fenêtres spectrales utilisées affecte grandement la performance des trois méthodes, qui se comportent différemment selon ce nombre. Ensuite, le bootstrap paramétrique s’avère être la meilleure méthode en termes de précision et de robustesse par rapport au nombre de fenêtres utilisées. Enfin, la technique RMC accompagnée de bootstrap a été utilisée pour l’extraction de source et la resynchronisation de données réelles provenant d’expériences en laboratoire et d’un moteur électrique, fournissant de meilleurs résultats que la solution des moindres carrés et la DVST dans les mêmes conditions.

 

Characterization of the fiber-matrix interface fracture properties in long fiber composites

Doctorant : Hugo GIRARD

Laboratoire INSA : MATEIS

École doctorale : ED34 : Matériaux de Lyon

Fiber-matrix interface in long fiber composite is a key aspect of global composite mechanical properties since it drives damage initiation and load transfer. Fiber-matrix interface debonding is usually the first type of damage that occurs when the composite is subjected to transverse loading. After initiation, the interface debonding propagates and often kinks into the matrix, leading to further critical defects for the structure. As a result, it is crucial to accurately characterize the fiber-matrix interface in order to prevent or control damage in composites. Going beyond existing experimental methods currently focused on interface shear fracture properties, single-fiber microcomposite tensile sample loaded transversely are developed to simultaneously characterize opening and shear fracture properties. An accurate experimental characterization of the fiber-matrix debonding process allowed the identification of the interface fracture properties using adequate 2D and 3D numerical approaches and related fracture models such as the Coupled Criterion (CC) and Cohesive Zone Models (CZM). Both the CC and the CZM are able to reproduce the experimentally observed debonding process in 2D, the 3D model being able to describe the free surface singularity. In 3D, the fracture property identification yields tensile strengths and critical energy release rates respectively slightly higher and in the same order of magnitude than those identified in 2D. The 3D model does not enable identifying the shear fracture properties, unlike in 2D. In 2D the optimal initiation crack shapes correspond to i) the stress isocontours for small brittleness numbers, ii) the energy-based shapes for large brittleness numbers and iii) neither of them for intermediate brittleness numbers. The 2D stress isocontours-based debonding shapes provide an accurate estimate of the initiation loading. In 3D, the optimum initiation crack always corresponds to energy-based debonding shapes and the 3D stress isocontours-based debonding shapes may thus overestimate the initiation loading by up to 30%.

 

Molecular dynamics simulation of semicrystalline polymers: from molecular topology to mechanical properties

Doctorant : Junxiong WANG

Laboratoire INSA : MateiS

École doctorale : ED34 : Matériaux de Lyon

Semi-crystalline polymers have attracted widespread attention due to their wide range of industrial applications, attractive mechanical properties, and good chemical resistance. Semi-crystalline polymers exhibit excellent mechanical properties due to their unique molecular structure (crystalline and amorphous phases overlapping each other). Topological molecules, like ties, loops, … and entanglements in amorphous phase, serve as stress transmitters and can be crucial to mechanical properties. However, these microstructures cannot be studied quantitatively experimentally, and the nonequilibrium process of crystallization and how the microstructure affects mechanical properties cannot be studied at the nanoscale. The dependence of the mechanical properties of semi-crystalline polymers on topology and entanglement has been explored using a coarse-grained model through molecular dynamics simulations. From cooling a melt, and after isothermal treatment, semi-crystalline polymers with lamellar structures were obtained with different entanglement densities and topologies. The strongest mechanical properties are shown when the tensile direction is highly consistent with the crystal chain orientation. And the system with a higher entanglement density has a smaller yield stress but a significant stress-hardening regime, indicating that high entanglement density effectively increases the stress-hardening effect. Additionally, the effect of different topologies on mechanical properties has been explored. Uniaxial tensile test results show that cilia have little effect on mechanical properties. The yield stress increases with the number of loops, showing that not only the loops but also the number of topologies has a strong influence on the mechanical properties. The tie molecules appear to have a slightly greater impact on the mechanical properties than the loops, manifesting in a slight strain softening effect. These results will enhance the understanding of the relationship between microstructure and mechanical properties of semicrystalline polymers.

New Approaches for the Construction of Ternary Solute/Solvent/Non-solvent Phase Diagrams and Applications in the Field of Nanoprecipitation

Doctorante : Yiping CHEN

Laboratoire INSA : IMP

École doctorale :  ED34 : Matériaux de Lyon

Nanoprecipitation (or solvent-displacement, Ouzo effect) process is a promising technique for straightforwardly producing colloids of controlled dimensions without recourse to surfactants or any high shearing force systems. Successful applications of the nanoprecipitation process crucially rely on the ability to construct the phase diagrams for the solute/solvent/nonsolvent ternary system of interest by identifying regions (SFME, Ouzo domain…) where the hydrophobic solute aggregates at the nanometer scale. Therefore, the main aim of this thesis is to develop robust methods to construct phase diagrams, to attempt at enlarging the Ouzo domain through cautiously-chosen additives, and finally to use thus constructed phase diagrams for precipitation of novel molecules, targeting potential applications in the biological field.
Combined fluorescence microscopy (FM) and dynamic light scattering (DLS) techniques were first used to construct phase diagrams containing pyrene as fluorescent indicator and surfactants as stabilizers in oil/solvent/water systems, respectively. It has been found that under the micrometer resolution FM, the pyrene-loaded nanodispersions appear black in the Ouzo domain (owing to their nanoscale dimensions). In DLS tests, adding a non-ionic surfactant, the Ouzo domain showed a monodisperse peak, with dI > dN and PDI < 0.15. Remarkably, the Ouzo domain identified by DLS was slightly larger than that obtained by FM owing to the introduction of surfactants. In a second approach, using the DLS technique, we studied the effect of adding specific surfactants (Brij 56, Enordet J3131, Cremophor EL) on the nanoprecipitation process. With miglyol oil, the Ouzo limit was shifted up to two decades, significantly enhancing the Ouzo domain. Finally, solid molecules such as antibiotics and fluorophores were used as solutes for nanoprecipitation. One specific antibiotic showed similar efficacy against various bacteria in molecular state or under nanoparticle form in vitro testing. Nanodispersions of high Tg fluorophores keep a good colloidal stability over a long period, and maintain their fluorescence activity upon dilution, making them good candidates as biomarkers.