• Refine Query
  • Source
  • Publication year
  • to
  • Language
  • 2
  • Tagged with
  • 4
  • 2
  • 2
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • About
  • The Global ETD Search service is a free service for researchers to find electronic theses and dissertations. This service is provided by the Networked Digital Library of Theses and Dissertations.
    Our metadata is collected from universities around the world. If you manage a university/consortium/country archive and want to be added, details can be found on the NDLTD website.
1

Multiscale carbon fibre composites with epoxy-graphite nanoplatelet matrices

Bin Junid, Ramli January 2017 (has links)
This thesis reports the effects of incorporating graphite nanoplatelets (GNPs) to epoxy-carbon fibre (CF) laminates to produce multiscale composites. A grade of epoxy resin typical for the application in aerospace engineering, triglycidyl-p-aminophenol (TGPAP), was used in this work cured with 4,4'-diaminodiphenyl sulfone (DDS). To improve the processability of TGPAP, a diluent, the diglycidyl ether of bisphenol F (DGEBF), was added to formulations. Compositions of TGPAP/DGEBF/DDS were optimised using response surface methodology (RSM) with the target response being to obtain high glass transition temperature (Tg) and low resin viscosity. From RSM, the optimum values were obtained at 55.6 wt. % of DGEBF and a stoichiometric ratio of 0.60. Before addition into epoxy, GNPs were treated either covalently using 3-aminopropyltriethoxysilane (APTS) or non-covalently using a commercial surfactant, Triton X-100 (abbreviated as A-GNPs and T-GNPs, respectively). After treatment, XPS analysis showed a new peak at 100 eV for A-GNPs indicating silicon and the C/O ratio increased from 11.0 to 26.2 for T-GNPs relative to unmodified GNPs (U-GNPs), suggesting attachment of the modifier molecules had occurred. Nanocomposites (NCs) were prepared by incorporate GNPs into epoxy using mechanical mixing. Rheological percolation threshold of GNP-epoxy suspensions were determined using oscillatory-shear rheometry as 3.9 wt. % for AR-GNPs, 3.6 wt. % for U-GNPs, 3.2 wt. % for A-GNPs and 3.5 wt. % for T-GNPs, suggesting surface treatment improved dispersion. At 4 wt. % of GNPs, flexural strain of NCs was decreased relative to neat epoxy by 46% for AR-GNPs, 48.6% for U-GNPs, 4.6% for A-GNPs and 30.8% for T-GNPs but flexural moduli showed small increases of 6.1-7.4%. Fracture toughness (K1C) also improved. For example, the K1C increased from 0.80 ± 0.04 MPa.m1/2 for neat epoxy to 1.32 ± 0.01 MPa.m1/2 for NCs containing 6 wt. % of U-GNPs possibly due to the branching of cracks resulting from the embedded GNPs. Due to their mechanical performance, A-GNPs were used to fabricate epoxy/CF/A-GNPs multiscale composites. Multiscale composites showed inferior properties relative to a comparable conventional composite in flexural testing, interlaminar shear strength (ILSS) and interlaminar fracture toughness mode II (G11C) due to weaker bonding at the matrix-CF interface. However, multiscale composites showed ~40% higher capability than conventional composite to absorb energy during impact due to greater interfaces formed by the inclusion of A-GNPs into the system.
2

Nouvelles frontières pour le cadre Arlequin en élastodynamique HF localisées - Application à la propagation des fissures / New frontiers for the Arlequin framework in localized HF elastodynamics - Application to crack propagation

Abben, Khalil 03 July 2019 (has links)
L’objectif principal de ce travail de thèse est de consolider et rendre opérationnelle l’extension du cadre multi-modèles et multi-échelles Arlequin à la modélisation et la simulation (par éléments finis) en élasto-dynamique, en faisant l’hypothèse de localisation des ondes hautes fréquences, tenant ainsi compte de phénomènes physiques dissipatifs divers.Parmi les applications visées par ces travaux, citons i) la propagation dans le sol d’ondes sismiques et leurs impacts sur des infrastructures critiques, ii) l’analyse multi-résolutions du comportement dynamique d’une structure impactée ou encore la propagation de fissures en dynamique dans des matériaux.Les contraintes que l’on s’impose dans ce travail sont doubles. La première est que l’on s’interdit de polluer la ou les zones critiques localisées. La seconde est que l’on souhaite aussi approcher le plus correctement possible le comportement des champs mécaniques dans les zones approchées plus grossièrement. Une étude de l’ensemble des paramètres Arlequin est menée. Des préconisations pratiques sont fournies, en étant étayées par des simulations 1D et une simulation 2D. Une attention toute particulière est portée à l’opérateur de couplage Arlequin en volume (dont on rappelle et souligne le caractère incontournable pour les problèmes de dynamique multi-échelle ; les couplages en surfaces étant inopérants, pour ces problèmes). Sur ce sujet, un des faits saillants de ces travaux de thèse est le développement d’un nouvel opérateur de couplage Arlequin réduit : tirant profit d’une représentation modale des champs de multiplicateurs de Lagrange, définis dans la zone de couplage, d’une notion de (1-epsilon) Compatibilité de modèles (initiée dans [Ben01b]) et du caractère multi-résolution des champs primaux du problème, dans la même zone de superposition Arlequin, cet opérateur permet de réduire considérablement les coûts des calculs des problèmes dynamiques multi-échelles abordés ici, par rapport à un couplage classique, tout en assurant des transmissions plus précises que celles données par deux autres méthodes de réduction, rappelées et mises en oeuvres dans cette thèse. Ces avantages sont étayés pour une barre élastique, en statique et en dynamique.Les approches développées sont utilisées et validées, par comparaison avec des résultats de la littérature, pour l’application phare de ce travail, consistant à simuler le comportement dynamique d’une structure fissurée, dans le cas d’une fissure fixe et celui d’une fissure propagative, en utilisant l’enrichissement par la fonction Level-Set à la X-Fem dans le modèle grossier et des éléments finis fins au voisinage du fond de fissure. / The main objective of this thesis work is to consolidate and make operational the extension of the multi-model and multi-scale Arlequin framework to modeling and simulation (using finite element) in elastodynamic, by making the hypothesis of localization of high frequency waves, thus taking into account various dissipative physical phenomena. Applications targeted by this work include i) ground propagation of seismic waves and their impact on critical infrastructures, ii) multi-resolution analysis of the dynamic behavior of an impacted structure or iii) the dynamic propagation of cracks.The constraints imposed on this work are twofold. The first is that one is prohibited from polluting the localized or critical areas. The second is that we also want to approach as accurately as possible the behavior of the mechanical fields in the coarsly approximated areas. A study of all dynamic Arlequin parameters is conducted. Practical recommendations are provided and supported by 1D and 2D simulations. Particular attention is paid to the volume Arlequin coupling operator (whose essential character for coupling in multi-scale dynamic problems is recalled and underlined; surface couplings being inoperative in this context). On this subject, one of the highlights of these thesis works is the development of a new reduced Arlequin coupling operator: taking advantage of a modal representation of the Lagrange multiplier fields defined in the coupling zone, a concept of (1- epsilon)-Compatibility of models (initiated in [Ben01b]) and the multi-resolution character of the overlayed primal fields, this operator makes it possible to reduce considerably the computational costs of the multiscale dynamic problem discussed here (when compared to a classical coupling) while ensuring transmissions more accurately than those given by two other reduction methods, recalled and implemented in this thesis. These benefits are supported by an elastic bar test, both in static and dynamic regimes.The developed approaches are used and validated, in comparison with results of the literature, for the flagship application of this work consisting of simulating the dynamic behavior of a cracked structure in the case of a fixed crack and that of a propagative crack using enrichment by the Level-Set function à la X-Fem in the coarse model and fine finite elements near the crack tip.
3

Design de polyuréthanes thermoplastiques (TPU) et étude des morphologies multi-échelles de mélanges bitume / TPU / Design of thermoplastic polyurethanes (TPU) and study of multi-scale morphlogies of bitumen/TPU blends

Gallu, Raïssa 19 November 2018 (has links)
Des polyuréthanes thermoplastiques (TPU) contenant des segments rigides et segments souples d’architecture moléculaire variable sont synthétisés en deux étapes, dont la première fait intervenir un pré-polymère polyuréthane. La microstructure de ces polymères montre qu’une séparation de phases intervient entre segments souples et rigides selon la nature des segments utilisés. Les segments rigides peuvent s’organiser sous deux formes, l’une amorphe et l’autre organisée sous forme d’entités cristallines. La morphologie des TPU dépend de la structure chimique du segment rigide employé. L’incompatibilité entre segments souples et rigides a été mise en évidence à partir de l’analyse des paramètres de solubilité complétée des caractérisations à différentes échelles par des techniques de microscopie (électronique et AFM) et de diffusion des rayons X. Ces polyuréthanes thermoplastiques sont ensuite utilisés pour préparer des mélanges bitume –polymère. Les interactions entre segments du polymère et fractions du bitume sont étudiées en considérant les paramètres de solubilité de chacun d’eux et des mesures de gonflement afin de juger de la miscibilité entre les composés. Des huiles modèles sont employées dans le but de mimer certaines fractions huileuses du bitume, et les segments souples et rigides sont synthétisés séparément afin d’étudier les propriétés de chacune des phases en présence dans le mélange bitume – polymère. La morphologie multi-échelle des mélanges est étudiée en lien avec les propriétés rhéologiques et la structure du polymère utilisé. L’ajout de polymère dans le bitume permet de modifier les propriétés viscoélastiques du bitume au-delà de sa transition vitreuse grâce au gonflement sélectif du polymère par les fractions huileuses. Après avoir mis en évidence et analysé le gonflement sélectif grâce à la prise en considération des paramètres de solubilité et mesuré les tensions interfaciales, nous montrons que la présence d’une phase continue riche en polymère contenant des segments rigides semi-cristallins dans les mélanges conduit à retarder l’écoulement du matériau bitumineux à plus hautes températures. La composition en huile de cette phase riche en polymère dépendra de son affinité avec les fractions du bitume et donc de la structure chimique du polymère. De plus, la teneur du polymère en segments rigides semi-cristallins est aussi un levier permettant d’intervenir sur son affinité avec le bitume et par conséquent sur les propriétés rhéologiques du mélange bitume-polymère. / Thermoplastic polyurethane (TPU) containing hard and soft segments with variable molecular architecture are synthesized in two steps, the first one including a polyurethane pre-polymer. The microstructure of theses polymers shows phase separation occurring between soft and hard segments according to the nature of the segments. Hard segments can organize under two forms, either amorphous or crystalline. The morphology of TPU depends on the chemical structure of the hard segment involved. Incompatibility between soft and hard segments was highlighted from solubility parameters analysis, complete with characterization at various scales with microscopy (electron and AFM) and X-ray scattering technics. Theses thermoplastic polyurethanes are used to prepare bitumen-polymer blends. Interactions between the polymer segments and bitumen fractions are studied, considering solubility parameters of each of them and swelling measurements in order to study miscibility between the compounds. Model oils are used in the aims of mimicking some oily fractions of bitumen, and soft and hard segments are separately synthesized to study properties of each phases in the bitumen-polymer mixture. Multi-scale morphology of the blends is studied in connection with rheological properties and structure of the used polymer. The addition of polymer in bitumen allows to modify viscoelastic properties of bitumen beyond its glass transition due to the selective swelling of the polymer by the oily fractions. Having highlighted and analyzed the selective swelling by considering solubility parameters and interfacial tension measurements, we show that the presence of a continuous polymer-rich phase containing semi-crystalline hard segments in the blends leads to delay the flow of the bituminous material at highest temperatures. The oil composition of this polymer-rich phase will depend on its affinity with the fractions of bitumen and thus on the chemical structure of the polymer. In addition, semi-crystalline hard segment content of the polymer is also a key parameter allowing to adjust its affinity with bitumen and consequently on rheological properties of the bitumen-polymer mixture.
4

Multiscale Modeling of the Mechanical Behaviors and Failures of Additive Manufactured Titanium Metal Matrix Composites and Titanium Alloys Based on Microstructure Heterogeneity

Mohamed G Elkhateeb (8802758) 07 May 2020 (has links)
<p>This study is concerned with the predictive modeling of the machining and the mechanical behaviors of additive manufactured (AMed) Ti6AlV/TiC composites and Ti6Al4V, respectively, using microstructure-based hierarchical multiscale modeling. The predicted results could constitute as a basis for optimizing the parameters of machining and AM of the current materials.</p> <p>Through hierarchical flow of material behaviors from the atomistic, to the microscopic and the macroscopic scales, multiscale heterogeneous models (MHMs) coupled to the finite element method (FEM) are employed to simulate the conventional and the laser assisted machining (LAM) of Ti6AlV/TiC composites. In the atomistic level, molecular dynamics (MD) simulations are used to determine the traction-separation relationship for the cohesive zone model (CZM) describing the Ti6AlV/TiC interface. Bridging the microstructures across the scales in MHMs is achieved by representing the workpiece by macroscopic model with the microscopic heterogeneous structure including the Ti6Al4V matrix, the TiC particles, and their interfaces represented by the parameterized CZM. As a result, MHMs are capable of revealing the possible reasons of the peculiar high thrust forces behavior during conventional machining of Ti6Al4V/TiC composites, and how laser assisted machining can improve this behavior, which has not been conducted before.</p> <p>Extending MHMs to predict the mechanical behaviors of AMed Ti6Al4V would require including the heterogeneous microstructure at the grain level, which could be computational expensive. To solve this issue, the extended mechanics of structure genome (XMSG) is introduced as a novel multiscale homogenization approach to predict the mechanical behavior of AMed Ti6Al4V in a computationally efficient manner. This is realized by embedding the effects of microstructure heterogeneity, porosity growth, and crack propagation in the multiscale calculations of the mechanical behavior of the AMed Ti6Al4V using FEM. In addition, the XMSG can predict the asymmetry in the Young’s modulus of the AMed Ti6Al4V under tensile and compression loading as well as the anisotropy in the mechanical behaviors. The applicability of XMSG to fatigue life prediction with valid results is conducted by including the energy dissipations associated with cyclic loading/unloading in the calculations of the cyclic response of the material.</p>

Page generated in 0.0467 seconds