In a proposed design for a concentrated solar power tower, sand is irradiated by solar energy and transfers its energy to another fluid stream by means of a finned tube heat exchanger. To maximize heat transfer and minimize potential damage to the heat exchanger, it is desired to have a very uniform flow through the heat exchanger. However, performing full scale flow tests can be expensive, impractical, and depending upon the specific quantities of interest, unsuitable for revealing the details of what it happening inside of the flow stream. Thus, the discrete element method has been used to simulate and study particulate flows. In this project, the flow of small glass beads through a square pyramid shaped hopper and a wedge shaped hopper were studied at the lab scale. These flows were also simulated using computers running two versions of discrete element modeling software – EDEM and LIGGGHTS. The simulated results were compared against the lab scale flows and against each other. They show that, in general, the discrete element method can be used to simulate lab scale particulate flows as long as certain material properties are well known, especially the friction properties of the material. The potential for increasing the accuracy of the simulations, such as using better material property data, non-uniform particle size distributions, and non-spherical particle shapes, as well as simulating heat transfer within a granular flow are also discussed.

Doctor of Philosophy / Department of Biological & Agricultural Engineering / Mark E. Casada / Ronaldo G. Maghirang / The United States grain handling infrastructure is facing major challenges to meet worldwide customer demands for wholesome, quality, and safe grains and oilseeds for food and feed. Several challenges are maintaining grain quality during handling; reducing dust emissions for safety and health issues; growing shift from commodity-based to specialty (trait-specific) markets; proliferation of genetically modified crops for food, feed, fuel, pharmaceutical, and industrial uses; and threats from biological and chemical attacks. This study was conducted to characterize the quality of grain and feed during bucket elevator handling to meet customer demand for high quality and safe products. Specific objectives were to (1) determine the effect of repeated handling on the quality of feed pellets and corn; (2) characterize the dust generated during corn and wheat handling; (3) develop and evaluate particle models for simulating the flow of grain during elevator handling; and (4) accurately simulate grain commingling in elevator boots with discrete element method (DEM). Experiments were conducted at the research elevator of the USDA-ARS Center for Grain and Animal Health Research (CGAHR) to determine the effect of repeated handling on the quality of corn-based feed pellets and corn. Repeated handling did not significantly influence the durability indices of feed pellets and corn. The feed pellets, however, had significantly greater breakage (3.83% per transfer) than the corn (0.382% per transfer). The mass of particulate matter < 125 μm was less for feed pellets than for corn. These corn-based feed pellets can be an alternative to corn in view of their handling characteristics. Another series of experiments was conducted in the same elevator to characterize the dust generated during corn and wheat handling. Dust samples were collected from the lower and upper ducts upstream of the cyclones in the elevator. Handling corn produced more than twice as much total dust than handling wheat (185 g/t vs. 64.6 g/t). Analysis of dust samples with a laser diffraction analyzer showed that the corn samples produced smaller dust particles, and a greater proportion of small particles, than the wheat samples. Published data on material and interaction properties of selected grains and oilseeds that are relevant to DEM modeling were reviewed. Using these material and interaction properties and soybeans as the test material, the DEM fundamentals were validated by modeling the flow of soybean during handling with a commercial software package (EDEM). Soybean kernels were simulated with single- and multi-sphere particle shapes. A single-sphere particle model best simulated soybean kernels in the bulk property tests. The best particle model had a particle coefficient of restitution of 0.6; particle static friction of 0.45 for soybean-soybean contact (0.30 for soybean-steel interaction); particle rolling friction of 0.05; normal particle size distribution with standard deviation factor of 0.4; and particle shear modulus of 1.04 MPa. The single-sphere particle model for soybeans was implemented in EDEM to simulate grain commingling in a pilot-scale bucket elevator boot using 3D and quasi-2D models. Pilot-scale boot experiments of soybean commingling were performed to validate these models. Commingling was initially simulated with a full 3D model. Of the four quasi-2D boot models with reduced control volumes (4d, 5d, 6d, and 7d; i.e., control volume widths from 4 to 7 times the mean particle diameter) considered, the quasi-2D (6d) model predictions best matched those of the initial 3D model. Introduction of realistic vibration motion during the onset of clear soybeans improved the prediction capability of the quasi-2D (6d) model. The physics of the model was refined by accounting for the initial surge of particles and reducing the gap between the bucket cups and the boot wall. Inclusion of the particle surge flow and reduced gap gave the best predictions of commingling of all the tested models. This study showed that grain commingling in a bucket elevator boot system can be simulated in 3D and quasi-2D DEM models and gave results that generally agreed with experimental data. The quasi-2D (6d) models reduced simulation run time by 29% compared to the 3D model. Results of this study will be used to accurately predict impurity levels and improve grain handling, which can help farmers and grain handlers reduce costs during transport and export of grains and make the U.S. grain more competitive in the world market.

Thesis (MScEng)--Stellenbosch University, 2012. / ENGLISH ABSTRACT: In this study, tool forces from rock cutting tests were numerically simulated through a discrete element method (DEM) in association with PFC3D™. Tribological interactions such as contact, shearing, fracturing, friction and wear were presented during these cutting simulations. Particle assemblies, representing Paarl granite and Sandstone-2, were created in PFC3D™ through a material-genesis procedure. The macro-properties of these particle assemblies, namely Young’s modulus, Poisson’s ratio, uniaxial and triaxial compressive strength and Brazilian tensile strength, were calibrated by modelling the uniaxial and triaxial compressive strength test and the Brazilian tensile strength test. The calibration was done through adjustment of the micro-properties of the assembly, namely the stiffness and strength parameters of the particles and bonds. The influence of particle size on the calibration was also investigated. These assemblies were used in the rock cutting tests. Results suggested that DEM can reproduce the damage formation during calibration tests successfully. From the results obtained from the calibration tests, it was also concluded that particle size is not a free parameter but influences the macro-properties greatly. Different rock cutting tools were simulated, namely point-attack (conical) picks, chisel-shaped tools and button-shaped tools. The numerical cutting tools were treated as rigid walls to simplify the simulation and the tool forces were not influenced by wear. In each simulation the cutting tools advanced at a constant velocity. The tool forces acting on the cutting tool, in three orthogonal directions, were recorded during the numerical simulations and the peak cutting forces were predicted by theoretical equations. The damage to the Paarl granite and Sandstone-2 assemblies was revealed as broken bonds, which merge into microscopic fractures. The mean peak cutting forces of sharp cutting tools obtained from numerical, theoretical and experimental models (from the literature) were compared. Finally the influence of factors, including wear on the tool and depth of cut, on the value of tool forces was also investigated. The results from the rock cutting tests revealed that the correlation between the numerical and the experimental models as well as the theoretical and experimental models was not strong when using sharp point-attack and chisel-shaped picks. It was concluded that the influence of wear plays a substantial part in the cutting process and it has to be included during the numerical simulation for the results to be accurate and verifiable. This study also found that there is a non-linear increase in tool forces with an increase in depth of cut, since the contact area increases. At larger cutting depths, chip formation also generally increased and therefore damage to the sample as well as wear on the cutting tool will be minimized at shallow cutting depths. Overall this study concludes that DEM are capable of simulating calibration methods and rock cutting processes with different cutting tools and producing results which are verifiable with experimental data. Therefore numerical prediction of tool forces will allow the design of efficient cutting systems and the operational parameters as well as the performance prediction of excavation machines. / AFRIKAANSE OPSOMMING: In hierdie studie is die kragte wat tydens rotssny-toetse op die sny gereedskap inwerk, numeries gesimuleer met behulp van ‘n diskrete element metode (DEM) in samewerking met PFC3D™. Tribologiese interaksies soos kontak, skeer, breking, wrywing en slytasie is gedurende hiersie snytoetse voorgestel. Partikel versamelings, wat Paarl graniet en Sandsteen-2 verteenwoordig, is in PFC3D™ geskep deur middel van ‘n materiaal-skeppings prosedure. Die makro-eienskappe van die partikel versamelings, naamlik Young se modulus, Poisson se verhouding, eenassige en drie-assige druksterkte en Brasiliaanse treksterkte, is gekalibreer deur modellering van die eenassige en drie-assige druksterkte toets en die Brasiliaanse treksterkte toets. Die kalibrasie is gedoen deur aanpassing van die mikro-eienskappe, naamlik die styfheid en die sterkte parameters van die partikels en bindings. Die invloed van partikelgrootte is ook ondersoek. Daarna is hierdie versamelings in die rotssny-toetse gebruik. Resultate het daarop gedui dat DEM die kraakvorming gedurende kalibrasie toetse suksesvol kan reproduseer. Vanuit die kalibrasie is ook gevind dat die partikelgrootte nie ‘n vrye parameter is nie, maar die makro-eienskappe grotendeels beïnvloed. Verskillende rotssny gereedskap is gesimuleer, naamlik koniese, beitel-vormige en knopie-vormige instrumente. Die numeriese sny gereedskap is gesimuleer as rigiede mure om simulasies te vereenvoudig en die gereedskap-kragte is dus nie deur slytasie beïnvloed nie. Tydens elke simulasie is die sny gereedskap vorentoe beweeg teen ‘n konstante snelheid. Die gereedskap-kragte, in drie ortogonale rigtings, is aangeteken gedurende die numeriese simulasies en die piek snykragte is ook voorspel deur teoretiese vergelykings. Die skade aan die Paarl graniet en Sandsteen-2 versamelings, is voorgestel as gebreekte bindings, wat saamsmelt tot mikroskopiese frakture. Die gemiddelde piek snykragte van skerp sny gereedskap van numeriese, teoretiese en eksperimentele modelle (uit die literatuur) is vergelyk. Ten slotte is die invloed wat faktore, onder andere die slytasie van gereedskap en die snydiepte, op die grootte van die kragte het ondersoek. Die resultate van die rotssny-toetse het aan die lig gebring dat die korrelasie tussen die numeriese en eksperimentale modelle sowel as die teoretiese en eksperimentele modelle nie sterk is tydens die gebruik van skerp koniese en beitel-vormige instrumente nie. Die gevolgtrekking is gemaak dat die invloed van slytasie van sny gereedskap ‘n wesenlike rol speel in die snyproses en dat dit in die numeriese simulasie ingesluit moet word sodat die resultate akkuraat en virifieerbaar is. Hierdie studie het ook gevind dat daar ‘n nie-lineêre toename in die gereedskap-kragte is met ‘n toename in snydiepte aangesien die kontak-area toeneem met ‘n toename in die snydiepte. By groter snydieptes, het die formasie van afsplinterings verhoog en dus sal skade aan die partikel versamelings en die slytasie van die gereedskap geminimeer word by vlakker snydieptes. Algeheel het die studie tot die gevolgtrekking gekom dat DEM in staat is om kalibrasie metodes en rotssny-toetse met verskillende sny gereedskap te simuleer asook om resultate te produseer wat verifieerbaar is met eksperimentele data. Numeriese voorspellings van die gereedskap-kragte sal dus toelaat om doeltreffende sny prosesse en operasionele parameters te ontwerp sowel as om die werkverrigting van uitgrawings masjiene te voorspel.

Thesis (MScEng)--Stellenbosch University, 2013. / ENGLISH ABSTRACT: Production of graphite dust inside the Pebble Bed Modular Reactor (PBMR) influences the reactor operation negatively. Graphite is used as a moderator in the reactor core and the formation and transportation of graphite dust away from the reactor core decreases the amount of moderator which in turn has a negative impact on the reactor operation. High levels of radioactive dust may also contaminate reactor components which may pose a health risk to maintenance personnel. In this study a pressure vessel was designed and used to measure the wear of a graphite pebble in helium at elevated temperatures. By means of a multi-linear regression analysis a proper mathematical function was established in order to relate graphite wear to certain tribological parameters. These parameters were identified through a literature study. Discrete Element Modelling (DEM) was used to simulate the gravitational flow of graphite pebbles through the reactor core. The experimentally determined mathematical function was incorporated into the DEM simulation to estimate the annual mass of graphite dust to be produced by the PBMR pebble bed as a result of pebble-pebble interaction and pebble-wall interaction during refuelling. / AFRIKAANSE OPSOMMING: Die vorming van grafiet stof binne die korrelbed-modulêre reaktor (PBMR) beïnvloed die werking daarvan negatief. Grafiet word gebruik as 'n moderator in die reaktor kern en die vorming en vervoer van grafietstof weg van die reaktor kern lei tot 'n afname in die hoeveelheid moderator en dit het 'n negatiewe impak op die werking van die reaktor. Hoë vlakke van radioaktiewe grafietstof kontamineer ook reaktorkomponente wat 'n gesondheidsrisiko vir onderhoudspersoneel inhou. In hierdie studie was 'n drukvat ontwerp en gebruik om die slytasie van 'n grafietkorrel in helium by verhoogde temperature te meet. 'n Multi-lineêre regressie analise is dan gebruik om 'n wiskundige funksie daar te stel wat die verband tussen grafietslytasie en die eksperimentele parameters vas stel. Hierdie parameters was met behulp van 'n literatuurstudie geïdentifiseer. Diskrete Element Modellering (DEM) was gebruik om die gravitasionele vloei van grafietkorrels in die reaktor te modelleer. Die eksperimenteel bepaalde wiskundige funksie word in die DEM simulasie ge-inkorporeer om 'n skatting te maak van die jaarlikse massa grafietstof wat gevorm sal word in die PBMR korrelbed as 'n gevolg van korrel-korrel interaksie en korrel-wand interaksie gedurende hersirkulasie.

The Distinct Element Method (DEM) and Discontinuous Deformation Analysis (DDA) are the two most commonly used discrete element methods in rock mechanics. Discrete element approaches are computationally expensive as they involve the interaction of multiple discrete bodies with continuously changing contacts. Therefore, it is very important to ensure that the method selected for the analysis is computationally efficient. In this research, a general assessment of DDA and DEM is performed from a computational efficiency perspective, and relevant enhancements to DDA are developed. The computational speed of DDA is observed to be considerably slower than DEM. In order to identify reasons affecting the computational efficiency of DDA, fundamental aspects of DDA and DEM are compared which suggests that they mainly differ in the contact mechanics, and the time integration scheme used. An in-depth evaluation of these aspects revealed that the openclose iterative procedure used in DDA which exhibits highly nonlinear behavior is one of the main reasons causing DDA to slow down. In order to improve the computational efficiency of DDA, an alternative approach based on a more realistic rock joint behavior is developed in this research. In this approach, contacts are assumed to be deformable, i.e., interpenetrations of the blocks in contact are permitted. This eliminated the computationally expensive open-close iterative procedure adopted in DDA-Shi and enhanced its speed up to four times. In order to consider deformability of the blocks in DDA, several approaches are reported. The hybrid DDA-FEM approach is one of them, although this approach captures the block deformability quite effectively, it becomes computationally expensive for large-scale problems. An alternative simplified uncoupled DDA-FEM approach is developed in this research. The main idea of this approach is to model rigid body movement and the block internal deformation separately. Efficiency and simplicity of this approach lie in keeping the DDA and the FEM algorithms separate and solving FEM equations individually for each block. Based on a number of numerical examples presented in this dissertation, it is concluded that from a computational efficiency standpoint, the implicit solution scheme may not be appropriate for discrete element modelling. Although for quasi-static problems where inertia effects are insignificant, implicit schemes have been successfully used for linear analyses, they do not prove to be advantageous for contact-type problems even in quasi-static mode due to the highly nonlinear behavior of contacts.

Computer simulation using the Discrete Element Method (DEM) has emerged as a powerful tool in studying the behaviour of particulate systems during powder flow and compaction. Contact law between particles is the most important input to the Discrete Element simulation. However, most of the present simulations employ over-simplistic contact laws which cannot capture the real behaviour of particulate systems. For example, plastic yielding, material brittleness, sophisticated particle geometry, surface roughness, and particle adhesion are all vitally important factors affecting the behaviour of particle interactions, but have been largely ignored in most of the DEM simulations. This is because it is very difficult to consider these factors in an analytical contact law which has been the characteristic approach in DEM simulations. This thesis presents a strategy for obtaining the contact laws numerically and a comprehensive study of all these factors using the numerical approach. A numerical method, named as the Material Point Method (MPM) in the literature, is selected and shown to be ideal to study the particle interactions. The method is further developed in this work in order to take into account all the factors listed above. For example, to study the brittle failure during particle impact, Weibull’s theory is incorporated into the material point method; to study the effect of particle adhesion, inter-atomic forces are borrowed from the Molecular Dynamic model and incorporated into the method. These developments themselves represent a major progress in the numerical technique, enabling the method to be applied to a much wider range of problems. The focus of the thesis is however on the contact laws between extremely fine particles. Using the numerical technique as a tool, the entire existing theoretical framework for particle contact is re-examined. It is shown that, whilst the analytical framework is difficult to capture the real particle behaviour, numerical contact laws should be used in its place.

Lunar soil, defined as the finest part of the lunar regolith which covers the entire surface of the Moon, has shown to have remarkable shear strength properties, highlighted by the clearly visible effects of soil cohesion. The main objective of this thesis is to unveil the physical explanations causing this unusual soil behaviour in a waterless, airless, lunar environment. Ultra-High Vacuum (UHV), in particular, is considered responsible for increasing the strength of surface energy forces due to lunar soil outgassing. In turn, the presence of surface energy forces, arising from van der Waals intermolecular forces, is thought to alter the mechanical properties of lunar soil. A particle-based microscopic approach by means of the Discrete Element Method (DEM) was utilised to investigate the effects of surface energy forces on the macroscopic soil be- haviour. A micro-mechanical contact model, based on the JKR theory, was selected to describe the inter-granular behaviour between lunar soil particles. Physical and geometrical parameters typical of lunar soil were employed. Several triaxial tests were run to identify a link, if any, between the microscopic surface energy parameter and the macroscopic soil cohesion, which was interpreted as a true soil cohesion. In addition, very low stress levels and high soil densities were simulated in order to take into account the low gravitational field and the high state of soil compaction caused by continuous meteorite impacts on the Moon. Results from triaxial tests were analysed at both the peak and critical state. It was found that in the ideal case of perfectly spherical grains, the presence of adhesion is a source of noticeable macroscopic soil cohesion. However, no influence was observed in terms of macroscopic friction angle. Furthermore, a brittle macroscopic soil behaviour was revealed, owing to the simulated inter-granular chemical bonds and the very low stress conditions applied. Finally, similar to the behaviour of cemented sands, very little cohesion was recorded at the critical state. Subsequently, particle shape effects were investigated by complementing the numerical model with a simple form of inter-particle rolling resistance. Simulations were also run with non-convex grains of increasing geometrical complexity in order to simulate more realistically the irregular shapes of lunar soil grains. In both cases, the interplay of surface energy forces with particle shape effects resulted in even higher shear strength, with predictions similar to the estimates of shear strength for real lunar soil. Once again, the peak strength was dominated by macroscopic cohesion which, on the other hand, was hardly observable at the critical state, confirming the tendency observed from spherical grains. Finally, the practical implications of the above findings were discussed in terms of bearing capacity, trafficability and slope stability on the lunar surface. In particular an analytical approach, based on the bearing capacity problem, was devised to study the performance of a rigid wheel rotating on a lunar terrain and operating under different dynamic conditions.

Les structures de sphères creuses appartiennent à la famille des matériaux cellulaires qui ont récemment été étudiés pour leurs multiples propriétés. Dans le cas de cette thèse, le but des sphères creuses est de dissiper l’énergie d’impact d’un oiseau sur un cockpit d’avion. Elles sont développées dans le cadre du projet SAMBA (Shock Absorber Material for Birdshield Application) afin d’optimiser leur énergie spécifique absorbée (J/kg).Dans un premier temps, des essais quasi-statiques (v = 5 mm/min) et dynamiques (v = 2 m/s) de compression uni-axiale sont conduits à température ambiante sur une seule sphère creuse de diamètre 30 mm. Une propagation rapide de fissures macroscopiques est observée. Le formalisme de la Mécanique Élastique Linéaire de la Rupture (MELR) est utilisé pour estimer le taux de restitution d’énergie critique dynamique GIdc du matériau constitutif. La position du sommet de fissure est mesurée pendant la propagation de fissure à l’aide d’une caméra rapide. La Méthode des Éléments Discrets (DEM) permet de simuler la rupture dynamique en implémentant une technique de relâchement des nœuds. Le taux de restitution d’énergie GIdc peut être estimé à partir de l’histoire (position et temps) du sommet de fissure. Le modèle numérique montre que les structures sphériques dissipent une proportion importante de l’énergie par des effets dynamiques. A une même vitesse de propagation, plus l’épaisseur de coque est fine, plus les effets inertiels générés par la rupture sont importants et ce pour une même vitesse de propagation.Le modèle numérique DEM est ensuite employé pour reproduire la rupture dynamique sur une sphère creuse à l’aide d’un critère en contrainte seule ou un critère mixte en contrainte – énergie. Les bons résultats obtenus démontrent la capacité de la DEM à représenter la propagation de fissures en régime dynamique.Finalement, des essais numériques et expérimentaux multi-sphères sont réalisés afin évaluer le comportement des sphères creuses au sein d’un assemblage. / Hollow sphere structure (HSS) belongs to cellular solids that have been studied recently for its multiples properties. In our case, HSS aims to absorb soft impacts energy on an airliner cockpit. HSS is investigated through the SAMBA (Shock Absorber Material for Bird-shield Application) project because of its promises in term of specific energy dissipated (J/kg) during impact.First of all, quasi-static and dynamic (v = 5 mm/min to v = 2 m/s) uniaxial compression tests are conducted at room temperature on a single sphere (D = 30 mm). Rapid crack propagation (RCP) is observed to be predominant at macroscopic scale. The formalism of Linear Elastic Fracture Mechanics (L.E.F.M.) is therefore used to estimate the dynamic energy release rate GIdc . The crack tip location is measured during the crack propagation using a high speed camera. The Discrete Element Method (DEM) is used to simulate the dynamic fracture by implementing the node release technique. The dynamic energy release rate can be determined using an experimentally measured crack history. In spherical structures the numerical results reveal a high proportion of energy dissipated through inertial effects as well as a dependence of the thickness of the hollow sphere over the range of 0.04 mm to 1.2 mm.The DEM model Is then employed to reproduce the RCP according to two failure criterions: a stress criterion and a coupled stress-energy criterion. It reveals to be an interesting way to model the mechanical behavior of brittle materials.Eventually, experimental and numerical multi-spheres tests are performed to evaluate the behavior of brittle hollow spheres within an assembly.

Dans les études expérimentales de la micromécanique des matériaux granulaires, la mesure des forces de contact entre particules est de nos jours toujours un challenge en comparaison avec les outils et techniques bien mieux établis pour la caractérisation cinématique à l'échelle des particules. Cette thèse de doctorat s'attaque à cet ambitieux problème. L'approche proposée implique deux aspects : (i) la caractérisation expérimentale du réseau de contact et de la cinématique à l'échelle des particules, qui peut être réalisée avec des techniques d'imagerie standards ; (ii) une approche numérique capable d'exploiter ces mesures afin de déduire les forces de contact.L'une des contraintes qu'on s'était imposée était de ne s'appuyer que sur la connaissance de la géométrie des particules ainsi que du réseau de contacts pour réaliser la déduction des forces de contact. Trois techniques numériques différentes ont été proposées à cet effet : une méthode basée sur l'élasticité des contacts (CEM), une méthode basée sur la dynamique de contact (CDM) et une méthode basée sur l'équilibre élasto-plastique de l'assemblage granulaire (QSM). Chacune de ces techniques repose sur une approche de la famille des méthodes en éléments discrets ; il s'agit respectivement de le DEM de type Cundall, la dynamique des contacts non régulière, et une approche de calcul statique élastoplastique. La non-unicité de la solution est le principal problème avec les techniques choisies, et elles sont étroitement liées à l’indétermination des forces dans le système.Les trois méthodes sont d'abord présentées et validées en les appliquant à l'estimation des forces dans les systèmes granulaires 2D générés au moyen de simulations DEM explicites. Nous prenons ces simulations comme des expériences "idéales" dans le sens où elles fournissent des données similaires à celles extraites des expériences, mais dépourvues d'erreurs de mesure. Un avantage évident de cette stratégie est d’obtenir des ensembles de forces faisant office de référence faisant foi. Sur cette base, les principaux aspects affectant la détermination des forces peuvent être étudiés. En particulier, le rôle crucial de l'histoire du chargement est mis en évidence et certaines solutions pour les prendre en compte dans la détermination des forces ont été prospectées. Une évaluation de l'influence de l'erreur de mesure a également été réalisée pour prédire l'applicabilité de chaque méthode à des expériences réelles. Une brève analyse de la variabilité des solutions est également fournie.Finalement, des tentatives ont été faites pour déduire des forces issues d’expériences effectuées dans le dispositif 1gamma2epsilon. La cinématique des particules et la connectivité ont été évaluées au moyen de la technique de corrélation d'image numérique. Les avantages et inconvénients des trois méthodes ont été éclaircis. Ils nous conduisent à envisager une utilisation combinée des trois méthodes pour tirer parti de leurs atouts respectifs. À l'avenir, il conviendra de réfléchir à la prise en compte de la stabilité de la solution -- dans l'algorithme de convergence vers une solution -- avec l'espoir de limiter la variabilité des solutions. / In the experimental study of the micro-mechanics of granular materials, measuring inter-particle contact forces is still a challenging task, if compared to the well-established tools and techniques for the kinematic characterisation at particle scale. This doctoral thesis addresses this problem. The proposed approach consists of two parts: an experimental characterisation of the granular network geometry and of particle-scale kinematics, which can be carried out with common imaging techniques such as Digital Image Correlation; a numerical approach aiming to exploit these measurements for the estimation of forces.One imposed constraint was to only make use of the rigid motions of particles, together with the knowledge of the contact network, to infer contact forces. Three different numerical techniques have been proposed to this purpose, referred to as Contact Elasticity Method (CEM), Contact Dynamics-based Method (CDM) and Quasi-Static Method (QSM). Each of these techniques is based on the formulation of common approaches in the family of Discrete Element Methods, respectively the classical Cundall-like DEM, the Non Smooth Contact Dynamics and a quasi-static approach accounting for both contact elasticity and plasticity. It is shown that memory of the history of the packing is the main concern with all the chosen techniques.The three methods are first presented and validated by applying them to the estimation of forces in 2D granular systems generated by means of explicit-time DEM simulations. We refer to these simulations as "ideal" experiments since they are meant to provide the same information that can be extracted from experiments, but without any measurement error. An obvious benefit of this strategy is to get reference force sets that are taken as ground truth. Based on this, the main aspects that affect the determination of forces can be investigated. In particular, the crucial role of history is emphasised here, and some solutions to take it into account in the force inference have been investigated.An assessment of the influence of measurement error has also been carried out, to predict the applicability of each method to real experiments. A short analysis of the variability of the solutions is also provided.Finally, some attempts have been made to infer forces from experiments carried out in the 1gamma2epsilon device. Particle kinematics and connectivity have been assessed by means of the Digital Image Correlation technique.The benefits and drawbacks of the three methods have been demonstrated. They conduct us to envision a combined usage of the three methods. In the future, studying the stability of equilibrium might help reducing the variability of the solutions.

A demanda por estruturas mais leves implica num ganho em economia, porém o aumento de esbeltez da estrutura pode tornar ela susceptível a instabilidade frente a tensões compressivas estáticas ou dinâmicas. A instabilidade acontece em várias escalas da estrutura analisada e pode interagir com outras formas de colapso como a propagação instável de fissuras, problema governado pela mecânica da fratura, pela plastificacão do material, ou por uma combinação dos efeitos citados. Neste contexto, no presente trabalho, se explora a capacidade do método dos elementos discretizados por barras (LDEM) na simulação de problemas de instabilidade estática e dinâmica devido as tensões de compressão. Este método permite simular o sólido como um arranjo de barras com rigidez equivalente ao contínuo que se quer representar. Leis constitutivas não lineares permitem modelar ruptura de forma simples. A equação de movimento resultante da discretização permite formular uma equação de movimento desacoplada que pode ser integrada no domínio do tempo com um método explícito (Método das Diferencias Finitas Centrais). O fato das barras serem rotuladas nos seus extremos e a solução do problema ser obtida de forma incremental permite capturar problemas com não linearidade geométrica, entre eles a instabilidade estrutural frente a tensões compressivas. Como último exemplo se realiza a análise de um painel sanduiche por flexão em três pontos, que é composto por um núcleo de poliuretano, com duas lâminas externas de material compósito, neste caso a instabilidade estrutural está associada a flambagem da camada da lâmina comprimida. Finalmente a potencialidade da metodologia de análise utilizada é discutida. / The demand for lighter structures implies a gain in economy, but the increase in slenderness of the structure may make it susceptible to instability against static or dynamic compressive stresses. Instability occurs at various scales of the analyzed structure and may interact with other forms of collapse such as unstable crack propagation, problem governed by fracture mechanics, plastification of the material, or a combination of the cited effects. In this context, in the present work, we explore the ability of the discrete elements methods by bars (LDEM) in the simulation of problems of static and dynamic instability due to the compression stresses. This method allows to simulate the solid as an arrangement of bars with rigidity equivalent to the continuum that one wants to represent. Constitutive non-linear laws allow simple modeling of rupture. The equation of motion resulting from the discretization allows us to formulate a decoupled motion equation that can be integrated in the time domain with an explicit method (Central Finite Differences Method). The fact that the bars are labeled at their ends and the solution of the problem is obtained in an incremental way allows to capture problems with geometric non-linearity, among them the structural instability against compressive tensions. The last example, the analysis of a sandwich panel by three-point bending, which is composed of a polyurethane core, with two external blades of composite material, in this case the structural instability is associated with buckling of the layer of the compressed blade . Finally, the potential of the analysis methodology is discussed.