Dissertations / Theses on the topic 'Laboratory modelling'

Collapsible soils covers naturally over 10% of the earth’s surface. This makes it a global problem and it is essential that engineers identify and control collapsibility prior to construction. Hence in this thesis, a study on identification, evaluation and control of soil collapsibility is undertaken. Four geologically different soils have been tested at five compactive variables from optimum moisture content (OMC). The soils tested include: Brown inorganic silty clay of low plasticity (A); White inorganic silt with slight plasticity (B); Red inorganic clay of intermediate plasticity (C); and Brown sand-clay mixtures with inorganic clay of low plasticity (D). The soils were each compacted at moisture variations 60% - 80%, 80% - 95%, 95% - 105%, 110% - 125% and 125% - 150% respectively representing ‘Low Dry OMC’, ‘High Dry OMC’, ‘At OMC’, ‘Low Wet OMC’ and ‘High Wet OMC’. The major causes of collapsibility of soil and the geomorphological processes that gives the pedogenesis of collapsible soils, is highlighted and great emphasis is placed on the adverse effect of collapsible soils. The experimental results from particle size distribution, Atterberg, compaction, triaxial and double oedometer tests showed that the soil’s percentage fine with the fines material (silt or clay), coefficient of uniformity, optimum moisture content, Atterberg limits, and stress-strain properties affect the metastability of the soils and they can be compared to the soil’s collapse potential when pressures and moisture content are applied on the soils. Results obtained showed that the soil’s collapse potential is directly proportional to 1) percentage fines, 2) the difference between the silt and clay percentage, 3) the Atterberg limits (liquid limit, plastic limit and plasticity index), and 4) internal friction angle; and inversely proportional to 1) coefficient of uniformity, 2) initial moisture content, 3) cohesion and finally 4) peak deviator stress. Each soil’s geological property proved to have an adverse effect on the metastability of the soils especially the dry of optimum moisture content. The most interesting results were obtained from the oedometer test. Results of the critical pressure varied with each soil and their compactive variable; Most of the soils at their ‘dry OMC’ had the highest collapse potential. In general, the lower the critical pressure the higher the collapse potential of the soil. The experimental data obtained herein were checked with the past research collapse indexes and found the results agreeing with just two research work out of eighteen examinations. Finally models for identifying soil collapsibility are generated with relationship between parameters from sieve, Atterberg, proctor compaction and triaxial. Laboratory data and data from twelve research work were used to verify the models and they show that the models work. After the verification of these formulas with past research data collected, the best models were three compactive variable models. The models give a collapsibility index in terms of percentage fines, initial moisture content, initial degree of saturation and initial dry density.

Thesis (Ph.D.) -- University of Maryland, College Park, 2009.
Thesis research directed by: Dept. of Physics. Title from t.p. of PDF. Includes bibliographical references. Published by UMI Dissertation Services, Ann Arbor, Mich. Also available in paper.

Le travail présenté dans cette thèse s’inscrit dans le domaine de l'astrophysique de laboratoire, qui consiste à étudier en laboratoire des processus physiques qui se produisent dans des objets astrophysiques. Les principaux avantages ici sont que les processus peuvent être étudiés de manière contrôlée et que leur dynamique complète peut être étudiée. Présentement, nous avons profité des installations laser à haute intensité pour effectuer nos études.Pour cela, dans ce manuscrit, seront traitées les questions liées à l'astrophysique de laboratoire qui impliquent l'interaction d’un plasma en détente dans le vide en présence d’un champ magnétique ambiant. La présence d'un champ magnétique dans une variété de phénomènes astrophysiques rend l’introduction de cette composante magnétique dans le laboratoire nécessaire afin que ces études soient pertinentes. Pour ce faire, en collaboration avec Laboratoire National des Champs Magnétiques Intenses -LNCMI, une bobine Helmholtz, spécialement conçue pour travailler dans un environnement laser a été développée, permettant d'atteindre une force de champ magnétique jusqu'à 30 T.Les objets astrophysiques sur lesquels cette étude est centrée sont les étoiles jeunes ou « Young Stellar Objects » (YSOs). Plusieurs étapes du processus de formation de ces étoiles seront ici étudiées : (i) la génération de jets collimatés à très grande échelle, (ii) la dynamique d'accrétion impliquant, dans la représentation standard, des flux de matière tombant sur la surface de l’étoile sous forme de colonnes magnétiquement confinées, et (iii) des canaux d'accrétion plus exotiques, comme l'accrétion équatoriale qui implique la propagation du plasma perpendiculairement aux lignes de champ magnétique.Plus précisément, dans un premier chapitre, la dynamique de formation des jets sera discutée. Une première partie est dédiée au mécanisme de formation de jet dans un champ magnétique poloïdal (aligné par rapport à l'axe principal d’expansion du plasma). Une seconde partie traite de la distorsion d'une telle formation de jet par l'interaction du même plasma en expansion avec un champ magnétique désaligné (c'est-à-dire présentant un angle par rapport à l'axe d’expansion du plasma). Enfin, une troisième partie détaille la propagation du plasma dans un champ magnétique perpendiculaire. Cette dernière partie nous permet d'étudier des canaux exotiques d'accumulation de matière sur les étoiles, consistant en une accrétion du disque d’accrétion directement vers l'étoile, c’est-à-dire sur le plan équatorial, impliquant une propagation orthogonale aux lignes de champ magnétiques. Le deuxième chapitre aborde le thème de la dynamique d'accrétion par l'intermédiaire de colonnes de matière magnétiquement confinées, tombant sur la surface stellaire. En utilisant la même configuration expérimentale que dans le premier chapitre, le jet formé (dans le cas du champ magnétique parfaitement aligné) est utilisé pour imiter la colonne d'accrétion et est lancé sur une cible secondaire qui agit comme la surface stellaire. La dynamique de choc à l'emplacement de l'obstacle est soigneusement étudiée et des liens avec les observations de phénomènes d’accrétion astrophysique sont construits. Un cocon de plasma, formé autour de la région d'impact via l'interaction avec le champ magnétique, est observé être similaire à celui trouvé dans les simulations astrophysiques. Ce cocon est un élément important en tant que milieu potentiel d'absorption des émissions X. Ce milieu permettrait en effet d'expliquer les écarts observés entre les émissions UV / optiques et les émissions X provenant des étoiles lors des phases d’accrétion
The work that is presented here has been performed in the frame of laboratory astrophysics, which consists in studying in the laboratory physical processes occurring in astrophysical objects. The main advantages in doing so are that the processes can be studied in a controlled way and that their full dynamics can be investigated. Here, we have been taking advantage of high-intensity laser facilities to perform our studies.In this manuscript, will be treated issues that include the interaction of a plasma expanding into vacuum with an ambient magnetic field. The presence of a magnetic field in a variety of astrophysical phenomena makes the inclusion of this component in the laboratory of great interest. We have used for our study a split Helmholtz coil, specifically designed in order to work in a laser environment, that allows for reaching a magnetic field strength up to 30 T.The astrophysical objects on which this study is focused are Young Stellar Objects (YSOs). Several steps of the star formation process are here investigated: (i) the generation of very long range, bright jets, (ii) the accretion dynamic involving, in the standard representation, matter falling down on the star in the shape of magnetically confined columns, and (iii) more exotic accretion channels, as the equatorial accretion that implies propagation of plasma perpendicularly to magnetic field lines.More precisely, in a first chapter, the jet formation dynamic will be discussed. A first part is dedicated to the jet formation mechanism in a poloidal magnetic field (aligned with respect to the main plasma expansion axis). A second part is dealing with the distortion of such jet formation via the interaction of the same expanding plasma with a misaligned magnetic field (i.e. presenting an angle with respect to the plasma expansion axis). Finally, a third part details the propagation of the plasma within a perpendicular magnetic field. This last part allows us to investigate exotic channels of matter accretion onto the stars, consisting of equatorial accretion from the disk to the star, through orthogonal magnetic field lines. The second chapter addresses the topic of the standard accretion dynamic via magnetically confined columns of matter, falling down onto the stellar surface. Using the same experimental setup as in the first chapter, the formed jet (in the case of the perfectly aligned magnetic field) is used to mimic the accretion column, and is launched onto a secondary target that acts as the stellar surface. The shock dynamic at the obstacle location is carefully studied and links with astrophysical accretion observations are built. A plasma cocoon, shaped around the impact region via the interaction with the magnetic field, is found to be similar to the one found in astrophysical simulations. This cocoon is an important element as a potential X-ray absorptive medium in order to explain observed discrepancies, between observed UV/Optical and X-ray emissions emitted from accreting stars

In order to investigate the kinetics and selectivity of the liquid phase nitration of phenol, reliable procedures for the production of the nitrating agents N₂O₅ and CINO₂ were established. Production of N₂O₅ was achieved via on-line mixing of NO₂ (1%) with O₃(~5%) in the 2:1 ratio in a darkened reaction vessel. CINO₂ was produced by conversion of N₂O₅ using NaCI solution (4 M). The concentration of nitrating agents was determined by bubbling the gas stream through water to yield nitrate which was analysed using Ion Chromatography and also by off-line reduction to nitrite followed by analysis using the sulphanilamide/UV method. A method for quantitative solid phase extraction with phenol in the aqueous phase. Nitration experiments were conducted under both acidic and basic conditions over a temperature range relevant to environmental processes. Results indicate the formation of 4-nitrosophenol (additionally characterised using GC-MS) as well as 2- and 4-nitrophenol (additionally characterised using GC-MS) as well as 2- and 4-nitrophenol. The product ratio of the products species has been shown to be dependent upon both the nitrating agent on pH. Finally, a chemical kinetic model has been developed to quantify the relative nitration pathways of aromatics in the troposphere. A box model, coded using FACSIMILE software, has been used to partitioning of benzene and phenol into the liquid phase and to assess the relative importance of the gas and liquid phases. The model includes the phase partitioning of 21 species and focuses on the conversion of benzene to phenol and finally nitrophenol in both the gas and liquid phase. Results indicate that the liquid phase contributes significantly to the production of nitrophenols in the troposphere. The system was shown to be sensitive to the assumed liquid water content over the range 3x10^-9 to 3x10^-6 as well as the droplet diameter and temperature. The accuracy of the liquid phase rate coefficients for the phenol + OH and phenol + NO₃ reactions were also shown to be very significant although altering the liquid phase benzene rate coefficient had little impact on the system as a whole. The model was also extending to include the partitioning of the nitrophenol products. This allowed an estimation to be made regarding the fate of the product species. At the benchmark Lc value of 3x10^-7, used to describe tropospheric cloud conditions, some 58% of the nitrophenols are produced by liquid phase processes whereas less than 2% of the nitrophenols remains in the liquid. This suggests that a great deal of the nitrophenol that may be observed in the gas phase is actually produced through liquid phase pathways.

Tidal meandering channels are ubiquitous features of tidal landscapes and play a fundamental role on the eco-morphodynamic evolution of these environments. However, only a handful of papers provide details on tidal meander planimetric shape, morphometric characteristics and morphodynamic evolution, and the internal achitecture of tidal meanders has not been explored in detail. Moreover, the morphodynamic evolution of tidal meanders and the related sedimentary products have often been interpreted on the basis of the well developed models and theories existing for their fluvial counterparts, despite a number of differences were a priori identifiable. Toward the goal of improving current understanding of the morphodynamic evolution of tidal meanders, five main issues have been investigated in the present work: i) rates of migration and evolutionary dynamics of tidal meanders; ii) assessment and quantification of differences and analogies existing between the planform features of tidal and fluvial meanders; iii) variations of tidal meander hydrodynamics in response to different tidal phases, and the role that these variations exert on tidal meander sedimentary products; iv) role played by bidirectional flows, tidal asymmetries and lateral tributaries; v) assessment of influence of tide amplitude, basin slope and initial shoreline configuration on tidal channel network ontogeny and evolution via laboratory experiment. A multidisciplinary approach has been adopted, with different methodologies encompassing remote sensing techniques, field observations, numerical modelling and physical-laboratory experiments. Activities have been carried out in parallel with sedimentological studies, in order to provide a comprehensive framework. The main results from this work highlighted that: I) once conveniently scaled with channel width, tidal meander migration rates are very similar to those displayed by fluvial meanders, thus challenging the paradigm of tidal meanders as a stable landscape features; II) differences and analogies between tidal and fluvial meander planforms can be addressed in a quantitative way, and different metrics exist thta allow one to successfully quantify these differences; III) strong asymmetries exist between different tidal phases, exerting a crucial role on the depositional patterns of tidal meanders; IV) under certain conditions, lateral tributaries can strongly influence the evolution of bends modifying local mechanisms of flow and sediment distribution; V) tidal channel network features evolve differently in response to different tidal ranges, basin slopes and relative sea level changes, whereas the number of breaches along the initial shoreline seems to have little effect on the evolution of the network itself.
Le reti di canali meandriformi costituiscono una delle principali componenti dei sistemi mareali, e giocano un ruolo di fondamentale importanza nell’evoluzione eco-morfodinamica di questi ambienti. Tuttavia, solo un numero limitato di studi scientifici ne ha analizzato le configurazioni planimetriche, le caratteristiche morfometriche e l’evoluzione morfodinamica. Inoltre, l’evoluzione morfodinamica e i prodotti sedimentari dei meandri a marea sono spesso stati interpretati sulla base di teorie e modelli sviluppati per i loro omologhi fluviali, nonostante numerose differenze tra le due tipologie siano identificabili a priori. Nell’intento di comprendere più approfonditamente l’evoluzione morfodinamica dei meandri a marea, nel presente lavoro sono stati studiati 5 differenti argomenti: i) tassi di migrazione e dinamiche evolutiove dei meandri a marea; ii) stima e quantificazione delle differenze planimetriche esistenti tra meandri fluviali e tidali; iii) variazioni dell’idrodinamica dei meandri a marea in risposta all’alternanza delle fasi mareali, e influenza di queste variazioni sui prodotti sedimentari propri dei meandri a marea; iv) ruolo della bidirezionalità del flusso, delle asimmetrie mareali e dei tributari laterali; v) stima dell’influenza dell’ampiezza di marea, delle pendenze topografiche del bacino tidale e della configurazione iniziale della linea di costa sulla nascita ed evoluzione morfologica delle reti di canali a marea. Nelle suddette analisi é stato utilizzato un approccio di tipo multidisciplinare, combinando metodologie quali remote-sensing, osservazioni in situ, modellazione numerica ed esperimenti su modelli fisici. Le attività sono state condotte in parallelo con studi sedimentologici, così da fornire un quadro che fosse il più esaustivo possibile. I principali risultati ottenuti evidenziano che: I) se convenientemente normnalizzati con la larghezza del canale, i tassi di migrazione dei meandri a marea sono molto simili a quelli dei loro corrispettivi fluviali, inficiando così il paradigma che vede i meandri tidali come un’entità morfologica essenzialmente stabile; II) le differenze tra meandri tidali e fluvali non sono solo qualitative, e diverse sono le metriche che possono essere utilizzate per quantificare queste differenze; III) le asimmetrie tra le diverse fasi di marea sono significative, e influenzano i patterns deposizionali in modo determinante; IV) gli affluenti laterali possono influenzare fortemente l’evoluzione dei meandri, modificando i meccanismi locali di distrubuzione dei flussi e dei sedimenti; V) le reti di canali a marea evolvono in modo diverso in risposta a differenti ampiezze di marea, pendenze del bacino tidale e cambiamenti del livello relativo del medio mare, mentre la configurazione iniziale della linea di costa non sembra avere effetti significativi sull’evoluzione della rete stessa.

An understanding of arsenic interactions with mineral surfaces in aqueous environments is important if arsenic transport in groundwater is to be predicted. This study contributes to the understanding of arsenic (As(V)) adsorption onto goethite (alpha-FeOOH) by linking equilibrium batch experiments and dynamic column experiments to the use of atomistic modelling. In this way the mechanisms of arsenic adsorption onto goethite surfaces ore explored. The equilibrium batch experiments illustrate the declining affinity of As(V) as a negatively charged oxy-anion to sorption on goethite, with increasing solution pH. These experiments also demonstrate that non-linear, Freundlich or Langmuir adsorption is favoured. Partition coefficients obtained are consistent with previously published values. Column experiments using goethite distributed in quartz grains under a range of conditions of flow velocity, influent arsenic concentration and goethite mass concentration, being more representative of natural conditions of groundwater flow, have provided new observations on As(V) adsorption. The onedimensional solute transport code BIO1D has been used to simulate the advection and sorption processes and to fit partition coefficients to the experimental breakthrough curves. Linear and Freundlich isotherms both provide a close representation of the experimental data. The results demonstrate that the partition coefficients under dynamic flow conditions are up to three orders of magnitude less than those derived from the equilibrium batch experiments. Also, the partition coefficients show an inverse relationship with flow velocity, indicating a kinetic effect. Experimental variation of the column redox conditions has enabled the determination of a pseudo-partition coefficient for sorption under anaerobic conditions, which is up to a quarter of that derived under aerobic conditions. Atomistic modelling shows that the (011) and (111) surfaces of goethite ore preferential for binding As(V) as AsO(OH)3. The model simulations also indicate that monodentate-mononuclear complexes are favoured over bidentate-binucleor formations. The results show a preferential orientation for the adsorbed As(V) molecule, with the hydroxyl groups lying parallel and the oxygen anion perpendicular to the goethite surface. The preferred complexation mechanisms and the relative energetics calculated using the atomistic models are good indicators of the non-linear adsorption observed in the experimental work.

Over the last few decades, the specifications for wrought aluminium products have become increasingly strict. Not only are the dimensional tolerances important, but also the material properties (strength, earing performance, corrosion resistance) must meet specified levels. The material properties are controlled by the microstructure. Hence the necessity for a modern aluminium plant to control the microstructure of the finished product through the processing parameters. This microstructure strongly depends on the microstructures produced during each of the processing steps. Therefore, it is necessary to control the microstructure throughout the production process. lt is thus imperative to know and model the influence of the processing conditions at each step. The present work focuses on one processing step: break-down rolling. During this step, the thickness of the ingot is reduced from 500 to 25 mm on a reversing mill. Compared with the other production steps, break-down rolling has not been studied extensively. One of the reasons for this is the absence of a laboratory technique that simulates this process accurately. During this work the Sheffield Mill for Aluminium Roughing at Temperature (SMART) was developed and it was proven that SMART can be used to simulate industrial break-down rolling. Furthermore, the data generated from SMART have been used to validate and refine a model from the literature. This model (developed at NTNU in Norway) predicts the evolution of the recrystallised fraction, the grain size and certain texture components throughout a multi-pass rolling operation. lt is shown here that the model predictions show a reasonable agreement with the results from SMART. Using the present experimental data, a set of recommendations to improve the model has been derived. Apart from the microstructural data, the experiments on SMART were also used to model the lateral spread that occurs during laboratory rolling. A new model is proposed that shows a better performance compared with the models that are available from the literature. The present work was carried out on AA3104 (AI-1Mn-1Mg) which is mostly used for the production of beverage cans.

Surface water is one of the fundamental parts of the environment and needs to be protected from all pollution sources for human survival. Urban development and human activities have increased the contamination of coastal and estuarine water due to insufficiently treated sewage, runoff from fertilised agricultural area and lawns and releasing industrial pollutant directly into river and estuaries. Cohesive sediment in estuaries can act as either a source or a sink of many pollutants, such as nutrients and heavy metals. Understanding the interactions between sediment and nutrients in water bodies is important because high input rates of nutrients can negatively affect water quality. The prediction of both deposition and the resuspension of cohesive sediment in estuaries supports the understanding of their turbidity, which is important in terms of the biomass of these water bodies and many of the occurring biochemical processes and the morphological processes, which determine the suitability and maintenance of fairways and harbour basins. The complex ways in which hydrodynamic and biochemical parameters affect cohesive sediment are primary reasons for the poor representation of the deposition, erosion and settling of cohesive sediment processes in modelling tools within water estuaries. However, our current understanding regarding the accurate prediction of cohesive sediment transport processes is insufficient because of flocculation processes which occur under certain circumstances (e.g. the increase of salinity in the brackish zone of rivers, which leads to the formation of flocs that are both larger and less dense than individual particles). The phenomenon of flocculation is known to play a significant role in the sediment transport processes of settling, deposition and erosion of cohesive sediment. There is no unique equation that can be universally used to predict the deposition and suspension rates of cohesive sediment because each estuary is dynamically and physically different from another and this is particularly true for the highly dynamic estuary e.g. Severn Estuary. Therefore, this study focuses on gaining a better understanding of the transport processes of cohesive sediment, including a better inclusion of the flocculation processes by developing a new settling velocity equation for cohesive sediment including flocculation processes as a function of hydrodynamic parameters. It also aims to apply this equation to a numerical model IV and to test this refined model by simulating the flocculation phenomenon in the highly dynamic Severn Estuary. This study employed an extensive experimental setup using a small scale particle image velocimetry (PIV) system. The experimental research was carried out using suspended sediment samples from the Severn Estuary in the UK. A PIV system and an image processing routine were used to measure both floc size distribution and settling velocity. The experimental results indicate that both the floc size and the settling velocity are controlled by the interaction between turbulence and salinity at salinities of less than 10 ppt. At a salinity either equal to or more than 10 ppt, both the floc size and the settling velocity were functions of only turbulent shear stress. The new equations were successfully applied in the Delft3D model; the model results show that they aptly were able to match with observed suspended sediment distributions throughout the estuary. Overall, the developed model can be regarded as a basic tool for being applied to help manage the suspended sediment processes in the Severn Estuary and for assessing the potential impact of climate change and human interference such as tidal renewable energy schemes in this water body.

Aquatic sediments are the recipients of a continual rain of organic debris from the water column. The decomposition reactions within the sediment and the rates of material exchange between the sediment and water column are critically moderated by the transport processes within the sediment. The sediment and solute movement induced by burrowing animals – bioturbation and bioirrigation – far exceed abiotic transport processes such as sedimentation burial and molecular diffusion. Thalassinidean shrimp are particularly abundant burrowing animals. Living in high density populations along coastlines around the world, these shrimp build complex burrow networks which they actively maintain and irrigate.¶ I used a laser scanner to map thalassinidean shrimp (Trypaea australiensis) mound formation. These experiments measured rapid two-way exchange between the sediment and depth. Subduction from the sediment surface proved to be just as important as sediment expulsion from depth, yet this is not detected by conventional direct entrapment techniques. The experiments demonstrated that a daily sampling frequency was needed to capture the extent of the two-way exchange.¶ I derived a one-dimensional non-local model accounting for the excavation, infill and collapse (EIC) of burrows. Maximum likelihood analyses were used to test the model against 210Pb and 228Th profiles taken from sediment cores in Port Phillip Bay, Melbourne. The maximum likelihood approach proved to be a useful technique for quantifying parameter confidence bounds and allowing formal comparison with a comparable biodiffusion model. The EIC model generally outperformed the biodiffusion model, and in all cases best EIC model parameter estimates required some level of burrow infill with surface material. The EIC model was expanded to two and three dimensions, which allowed the representation of lateral heterogeneity resulting from the excavation, infill and collapse of burrow structures. A synthetic dataset generated by the two-dimensional model was used to demonstrate the effects of heterogeneity and core sampling on the mixing information that can be extracted from one-dimensional sediment core data.¶ Burrow irrigation brings oxygenated water into burrow depths, and can affect the nitrogen cycle by increasing the rates of coupled nitrification and denitrification reactions. I modelled the nitrogen chemistry in the annulus of sediment surrounding an irrigated burrow using a radially-symmetrical diffusion model. The model was applied to three published case studies involving thalassinidean shrimp experiments and to field data from Port Phillip Bay. The results highlighted divergences between current theoretical understanding and laboratory and field measurements. The model further demonstrated potential limitations of measurements of burrow characteristics and animal behaviour in narrow laboratory tanks. Activities of burrowing animals had been hypothesised to contribute to high denitrification rates within Port Phillip Bay. Modelling work in this thesis suggests that the model burrow density required to explain these high denitrification rates is not consistent with the sampled density of thalassinidean shrimp in the Bay, although dense burrows of other animals are likely to be important. Limitations of one-dimensional representations of nitrogen diagenesis were explored via comparisons between one-dimensional models and the full cylinder model.

Thermoacoustic instabilities are a mayor problem in industrial combustors, where they can lead to catastrophic hardware damage. An industrial gas turbine combustion chamber is a very complex and expensive system. Thus, a laboratory burner has been built for research purposes, where a large number of parameters can be varied. This study is part of the Marie Curie research network LIMOUSINE, which was set up to model thermoacoustic instabilities in the combustor chamber of gas turbines. The objective of the present thesis is to theoretically model and analyze thermoacoustic instabilities in the LIMOUSINE laboratory burner. A mathematical model of the laboratory burner has been developed. A more general form of the wave equation has been derived in the time-domain, in which the mean temperature gradient was taken into account. The governing differential equation has been solved by applying the Green’s function approach, which allows separating the effects of the unexcited burner and the fluctuating heat-release. Using perturbation techniques general solutions are given for the cases when the temperature increase is either small or large. Conclusions have been drawn about the necessary complexity of thermoacoustic models by comparing increasingly complex configurations. The forcing term of the wave equation is studied by investigating the kinematics of ducted premixed flames theoretically, and a new heat-release law is derived. Instability criterion has been derived by applying the non-linear source term. The stability parameter map of the burner has been also investigated. Expressions for the limit-cycle amplitudes and frequencies were derived using weakly non-linear theory. The predictions of the mathematical model have been compared to measurements.

Functional Magnetic Resonance Imaging (fMRI) is a common technique for imaging brain activity in humans. However, the fMRI signal stems from local changes in oxygen level rather than from neuronal excitation. The change in oxygen level is referred to as the Blood Oxygen Level Dependent (BOLD) response, and is connected to neuronal excitation and the BOLD response are connected by the neurovascular coupling. The neurons affect the oxygen metabolism, blood volume and blood flow, and this in turn controls the shape of the BOLD response. This interplay is complex, and therefore fMRI analysis often relies on models. However, none of the previously existing models are based on the intracellular mechanisms of the neurovascular coupling. Systems biology is a relatively new field where mechanistic models are used to integrate data from many different parts of a system in order to holistically analyze and predict system properties. This thesis presents a new framework for analysis of fMRI data, based on mechanistic modelling of the neurovascular coupling, using systems biology methods.  Paper I presents the development of the first intracellular signaling model of the neurovascular coupling. Using models, a feed-forward and a feedback hypothesis are tested against each other. The resulting model can mechanistically explain both the initial dip, the main response and the post-peak undershoot of the BOLD response. It is also fitted to estimation data from the visual cortex and validated against variations in frequency and intensity of the stimulus. In Paper II, I present a framework for separating activity from noise by investigating the influence of the astrocytes on the blood vessels via release of vasoactive sub- stances, using observability analysis. This new method can recognize activity in both measured and simulated data, and separate differences in stimulus strength in simulated data. Paper III investigates the effects of the positive allosteric GABA modulator diazepam on working memory in healthy adults. Both positive and negative BOLD was measured during a working memory task, and activation in the cingulate cortex was negatively correlated to the plasma concentration of diazepam. In this area, the BOLD response had decreased below baseline in test subjects with >0.01 mg/L diazepam in the blood. Paper IV expands the model presented in Paper I with a GABA mechanism so that it can describe neuronal inhibition and the negative BOLD response. Sensitization of the GABA receptors by diazepam was added, which enabled the model to explain how changes to the BOLD response described in Paper III could occur without a change in the balance between the GABA and glutamate concentrations. The framework presented herein may serve as the basis for a new method for identification of both brain activity and useful potential biomarkers for brain diseases and disorders, which will bring us a deeper understanding of the functioning of the human brain.

The dehydration behaviour of several crystalline materials has been investigated, by means of lab-based studies, using X-ray diffraction and Differential Scanning . Calorimetry, and molecular modelling techniques. A review of several groups of materials, namely sugars, amino acids, and pharmaceutically important materials was undertaken. This showed how difficult it is to make predictions about dehydration behaviour based solely on examination of the structure of hydrated forms. Few patterns could be seen, other than the size of the molecules is important in determining whether or not a material will form crystalline structures at all levels of hydration. Amino acids are largely affected by the nature of the side chains. The nature of the structure of some of the pharmaceuticals meant that the pattern of dehydration, and the means by which the water molecules left the structure was easy to predict. The dehydration characteristics of a-lactose monohydrate, citric acid monohydrate, and calcium sulphate dihydrate have been studied using experimental methods and molecular simulation methods. The kinetics of dehydration of a-lactose monohydrate and citrjc acid monohydrate have also been studied, using a cell 'designed for this study which enabled the progress of the phase change to be followed using X-ray diffraction. The morphologies of these materials were also investigated, and the surfaces of the hydrated forms studied in order'to establish whether or not there may be a preferential route by which water leaves the structure. For a-lactose monohydrate, a further study was undertaken to try to establish the cause of the tomahawk morphology seen when it is grown from solution. Whilst the results presented are not clear-cut, they do indicate that both polar effects and the presence of lactose in the crystailising solution both have an influence on the shape of the crystal.

The thesis is related to laboratory astrophysics, and investigates with this technique, the launching mechanism for young stellar object jets and the interaction of two supernovae remnant in the Sedov-Taylor regime. Recent experiments performed at Imperial College on the pulsed-power magpie facility have successfully shown the formation of magnetically driven radiatively cooled plasmas jets formed from radial wire arrays, which are relevant to studying the launching mechanisms of astrophysical jets [A. Ciardi, et al. Phys. Plasmas 14, p056501 (2007)]. The experiments have been now extended to study episodic mass ejection ( 25 ns [F. A. Suzuki-Vidal, et al. 49th Annual Meeting of the Division of Plasma Physics, UO4.00007 (2007)]) and the interaction of jets and magnetic bubbles with an ambient gas. The dynamics of the interaction is investigated through three-dimensional resistive magneto-hydrodynamic simulations using the code gorgon [A. Ciardi, et al. Phys. Plasmas 14, p056501 (2007) – J.P. Chittenden, et al. Plasma Phys. Control. Fusion 46 B457 (2004)]. In particular ablation of the cathode is investigated numerically to explain the periodicity and subsequent formation of multiple bubbles. Comparison with experiments is offered to validate the results. The complex structure of the magnetic field is investigated, the conservation of the magnetic flux is explained and the consequent confinement offered to the central jet. Furthermore the interaction of the plasma outflows with an ambient gas is investigated. The formation of shocks in the ambient gas, as well as the formation of three-dimensional Mach stems is analyzed. In addition, recent experiment at Imperial College performed by the QOLS group, by laser-heating a medium of atomic clusters [R. A. Smith, et al. 2007 Plasma Phys. Control. Fusion 49 B117-B124 (2007)], shows the capability to create plasmas with sufficiently high energy densities to launch strong shocks. Interactions between high-Mach number shock waves are believed to be responsible for many of the complex, turbulent structures seen in astrophysical objects including supernova remnants. The experiment of two colliding Sedov-Taylor regime blast-waves is modelled. Detailed 3D numerical modeling is performed in order to study the importance of thermal conduction, rarefaction waves, refractive shock waves and complex three-dimensional mach stem formation. The simulated data are benchmark against a three-dimensional tomography image (newly developed experimental technique). The collision of two blast-waves should reproduce the non uniform interstellar medium where supernovas normally expand.

ABSTRACT MONITORING AND STATISTICAL MODELLING OF DUST CONCENTRATION OF SOME TURKISH LIGNITES UNDER LABORATORY CONDITIONS Karakas, Ahmet Ph.D., Department of Mining Engineering Supervisor: Prof. Dr. Tevfik Gü
yagü
ler September 2010, 171 pages Although technological developments enable maximum safety, high dust formation is still a crucial problem in coal mining sector. This study is aimed to investigate the relationship between amount of coal dust produced during cutting operation and some important coal properties together with cutting parameters for different particle size ranges in laboratory conditions. For this purpose, six Turkish lignite samples were used in the experiments. Two experiment sets were designed to generate coal dust by using cutting action of the saw. First group of experiments were conducted in large scale saw system by using saws having three different diameters and dust concentration measurements were done for three group of particle size namely: 0-2.5 &mu
m, 0-5 &mu
m and 0-10 &mu
m. Second group of experiments were done in small scale saw system by changing the table advancing speed and tip speed of the system. For this group of experiments, only one type of lignite samples were used. These measurements were carried out only 0-10 &mu
m particle size range by using a saw with 30 cm in diamater. In this study, to characterize the lignite samples
proximate, petrographic and grindability analysis (HGI) were made. During the experiments, dust concentrations were measured by using Microdust Pro real time dust monitoring equipment. At the end of the study, the relationship between coal dust concentration and some coal properties and cutting operating parameters were expressed by using four different regression equations. Also it has been found that tip speed of saw, fixed carbon, ash and huminite content, vitrinite reflectance and hardgrove grindability index are very important parameters in coal dust generation. Keywords: Respirable Coal Dust, Dust Generation, Real Time Dust Sampler, Coal Dust Diseases, Coal Cutting Operation.

Weaver is a hybrid knowledge discovery environment which fills a current gap in Artificial Intelligence (AI) applications, namely tools designed for the development and exploration of existing knowledge in complex, knowledge and data-poor domains. Such domains are typified by incomplete and conflicting knowledge, and data which are very hard to collect. Without the support of robust domain theory, many experimental and modelling assumptions have to be made whose impact on field work and model design are uncertain or simply unknown. Compositional modelling, experimental simulation, inductive learning, and experimental reformulation tools are integrated within a methodology analogous to Popper's scientific method of critical discussion. The purpose of Weaver is to provide a 'laboratory' environment in which a scientist can develop domain theory through an iterative process of in silico experimentation, theory proposal, criticism, and theory refinement. After refinement within Weaver, this domain theory may be used to guide field work and model design. Weaver is a pragmatic response to tool development in complex, knowledge- and data- poor domains. In the compositional modelling tool, a domain-independent algorithm for dynamic multiple scale bridging has been developed. The multiple perspective simulation tool provides an object class library for the construction of multiple simulations that can be flexibly and easily altered. The experimental reformulator uses a simple domain-independent heuristic search to help guide the scientist in selecting the experimental simulations that need to be carried out in order to critically test and refine the domain theory. An example of Weaver's use in an ecological domain is provided in the exploration of the possible causes of population cycles in red grouse (Lagopus, lagopus scoticus). The problem of AI tool validation in complex, knowledge- and data-poor domains is also discussed.

The Biot-Gassmann theory of poroelasticity forms the basis of most in­vestigations of wave propagation in fluid saturated rock. In recent years the need to incorporate the concept of squirt flow into the theoretical framework has been recognised. Microstructural models which contain squirt flow give inconsistent predictions which contradict rigorous results from poroelastic theory. I derive a microstructural poroelastic model which incorporates the squirt flow mechanism. The model is consistent in its limiting forms with the stan­dard results of poroelasticity and effective medium theory. An important feature of the model is that it is relatively independent of assumptions about the aspect ratio spectrum. I describe how the various parameters which occur in my model may be derived or estimated from experimental data, and proceed to a preliminary calibration of the model using published resonant bar data. Although I show that the data can be fit satisfactorily, significant ambiguity remains in the interpretation of the results. A number of ultrasonic tests of P- and S- velocity, in rock similar to the resonant bar, as a function of both effective stress and pore fluid type show results which are at variance with the predictions of published poroelastic theories. I demonstrate that the anomaly can be explained with reference to physical effects predicted by my model. Moreover, the requirement to explain the ultrasonic results places constraints upon the modelling of the resonant bar data, removing much of the ambiguity from the analysis. I present a calibration which gives a consistent qualitative explanation of both the resonant bar and ultrasonic data. The calibrated model makes a number of predictions concerning the ef­fect of changing pore fluid viscosity, sample permeability and frequency. In principle experiments could be carried out to test these predictions.

La contribution des parcs éoliens offshore en termes d’énergie renouvelable ne cesse de croître. L’électricité produite notamment par les éoliennes en mer du Nord représente de ce fait une part importante de l’énergie consommée en Europe. Pourtant, les objectifs de la Commission Européenne en termes de transition énergétique à l’horizon de 2020 sont loin d’être réalisés sans l’optimisation des techniques existantes et le développement de technologies de plus en plus innovantes. Les monopieux sont les fondations les plus populaires dans l’industrie des éoliennes offshores. Ce sont des pieux larges supportant des éoliennes de fortes puissances et situées dans des eaux plus profondes. L’objectif général de cette thèse de doctorat est d’étudier les interactions solstructure des monopieux sous sollicitation latérale, fondés dans les sols argileux. Une attention particulière est donnée aux comportement du sol dans le but de modéliser finement le comportement des sols argileux
Nowadays, offshore wind energy industry is developing exponentially, due to the significant contribution of the North Sea wind turbines energy production to the total consumed energy in Europe. Given that the EU's energy target is to increase the share of renewable energy by 2020, there is a great potential of the offshore wind energy applications towards this direction. In order to achieve this, the support of scientific research is crucial. Monopiles have been by far the most support structure for offshore turbines, nowadays becoming applicable also for complex site conditions. The main objective of this PhD thesis is to study the soil-foundation interaction problem for offshore wind turbines monopiles embedded in clays. We focus therefore on the numerical modelling of natural clay behavior. We aim to develop a constitutive model for clay soils, which allow developing new p-y curves that could be widely applied in offshore wind turbines monopiles

Polar regions have recently been subject to warming at almost double the global average, causing the retreat of glacier and ice cover and exposing large expanses of terrestrial habitats. These pioneer ecosystems are colonised by microbes. Microbial activity is thought to be responsible for the initial build-up of biomass, organic carbon and nutrients, facilitating the growth of higher order plants in developed soils. However, the reliance of the microbial community on ancient carbon pools, external sources (e.g. deposition) of nutrients, photosynthesis and nitrogen fixation is unknown. Furthermore, the seasonal dynamics are largely unexplored. Thus, SHIMMER (Soil biogeocHemIcal Model for Microbial Ecosystem Response) was developed. SHIMMER is a new numerical modelling framework designed to simulate pioneer ecosystem dynamics in glacier forefields. Initial performance and sensitivity tests showed model dynamics are highly dependent on microbial growth and death rate constants, growth efficiency, and 010 values. The model was applied to the Midtre Lovembreen forefield (Svalbard) and key model parameters were refined based on laboratory measurements. Model output and field data showed that in situ bacterial production is responsible for the observed accumulation of biomass and nutrients over several decades. Carbon fluxes were also shown to increase with soil age. Additionally, modelling showed that microbial dynamics in Arctic soils are susceptible to future anthropogenically-induced climate changes. Aspects of these glacier forefield systems that require further empirical research are: entire catchment nutrient budgets (in particular estimating allochthonous deposition), increasing data availability (especially during the winter), and improving mathematical formulation of biological processes (e.g. cell death rates and nitrogen fixation). SHIMMER is a new quantitative and process-focussed approach that improves the general understanding of glacier forefield soils, and the significance of glacier forefield soils in local to global scale biogeochemical cycles.

Muskeg or peat deposits cover large areas in northern Manitoba. Test sections of a newly constructed highway on peat were instrumented to investigate their performance and to develop more economical means of construction method. Test Section ‘A’ was constructed with geotextile base layer while Section ‘B’ was with geotextile and corduroys (timber logs). The test sections were constructed during winter for ease in mobilizing construction equipment at the site when the ground was frozen and were instrumented to observe its behaviour and performance. Settlements were measured using monitoring plates and pins. Ground temperatures were measured using thermistors. Porewater pressures were measured using vibrating wire piezometers. Peat in the study area has an average thickness of 4m, with the upper layer classified as fibrous and the lower layer as amorphous with strong to complete decomposition. Standard laboratory tests were conducted on bored samples from the site. Hydraulic conductivity tests were carried out at different vertical pressures to determine its permeability. Thermal conductivity was determined at frozen and unfrozen state of peat. Conventional incremental oedometer tests were conducted to determine the compressibility parameters and secondary compression indices of the peat layers. Constant-rate-of-strain (CRS) tests were also performed to supplement the results obtained from the conventional method. Isotropically-Consolidated Undrained (CIŪ) triaxial tests were carried out to determine the shear strength of peat. A commercially-available computer program was used in the numerical modelling to simulate the field performance of the instrumented sections. The results from numerical modelling were reasonably close to the measured values in the field. Laboratory-scale physical modelling was undertaken to understand further the operating mechanisms involved in the performance of the two test sections under a more controlled environment. Artificial transparent clay that has similar deformation properties with most of the natural clays and peats was used as foundation material. It allows determination of spatial deformations beneath the embankment using Particle Image Velocimetry (PIV) technique. The load-settlement behaviour in the field was also reasonably simulated in the laboratory-scaled physical model. Deformation patterns from PIV indicate that embankment with geotextile layer and corduroy has smaller settlements and lateral movements in the foundation compared to that of the embankment with only geotextile layer.

Braided rivers are complex, fascinating fluvial pattern, which represent the natural state of many gravel and sand bed rivers. Both natural and human causes may force a change in the boundary conditions, and consequently impact the river functionality. Detailed knowledge on the consequent morphological response is important in order to define management strategies which combine different needs, from protection of human activities and infrastructures to preservation of the ecological and biological richness. During the last decades, research has made significant advance to the description of this complex system, thanks to flume investigations, development of new survey techniques and, to a lesser extent, numerical and analytical solutions of mathematical models (e.g. Ashmore_2013). Despite that, many relevant questions, concerning the braided morphodynamics at different spatial and temporal scales (from the unit process scale, to the reach scale, and eventually to the catchment scale) remain unanswered. For example, quantitative analysis of the morphological response to varying external controls still requires investigation and needs the definition of suitable, stage-independent braiding indicators. In addition, the morphodynamics of the fundamental processes, such as bifurcations, also needs further analysis of the driving mechanisms. General aim of the present study is to develop new methods to exploit, in an integrated way, the potential of the new possibilities offered by advanced monitoring techniques, laboratory models, numerical schemes and analytical solutions. The final goal is to fill some gaps in the present knowledge, which could ultimately provide scientific support to river management policies. We adopted analytical perturbation approaches to solve the two-dimensional shallow water model; we performed laboratory simulations on a large, mobile-bed flume; we analysed existing topographic measurements from LiDAR and Terrestrial Laser scanning Devices; and we simulated numerically the river hydrodynamics. Within each of the six, independent, research chapters, we interconnected results from the different approaches and methodologies, in order to take advantage of their potential. Summarising, the more relevant and novel outcomes of the present work can be listed as follows: 1) We explored the morphological changes during a sequence of flood events in a natural braided river (Rees River, NZ)and we proposed a morphological method to assess the sediment transport rate. In particular we propose a semi-automatic method for estimating the particles path-length (Ashmore and Church, 1998) on the basis of the size of the deposition patches, which can be identified on the basis of DEM of differences. Comparison with results of numerical simulation confirmed that such an approach can reproduce the response of the bedload rate to floods of different duration and magnitude. 2) We developed a new indicator of the reach-scale morphology and, on the basis of existing laboratory experiments, we explored its dependence, under regime conditions, to the controlling factors: slope, discharge, confinement width, grain size. In spite of its synthetic nature, this simple indicator embeds the information needed to estimate the variability of the Shield stress throughout the braided network, and consequently enables to assess the transport-rate and its variation with the driving discharge. 3) We investigated, through flume experiments, the effect of the flow unsteadiness on the sediment transport in a braided river. This is possible only by following a statistical approach based on multiple repetitions of the same flow hydrograph. Results revealed that for confined network an hysteresis of the bedload response occurs, which leads to higher sediment transport during increasing flow, whereas relatively unconfined networks always show quasi-equilibrium transport rates. 4) A second set of laboratory experiments provided information on the morphodynamics of a braided network subject to variations of the sediment supply. We proposed a simple diffusive model to quantify the evolution of the one-dimensional bed elevation profile. Such simple approach, albeit having a limited range of practical applications, represents the first attempt to quantify this process and enables to study the relevant temporal and spatial scales of the phenomenon. 5) We solved analytically the two-dimensional morphodynamic model for a gravel-bed river bifurcation. This furnishes a rigorous proof to the idea proposed by Bertoldi and Tubino (2007) to interpret the morphological response of bifurcation in light of the theory of the morphodynamic influence. The analytical approach enables to investigate the fundamental mechanics which leads to balance, and unbalance, configurations and, from a more practical point of view, allows for a better prediction of the instability point than the existing 1D models (e.g. Bolla Pittaluga et al., 2003).

Braided rivers are complex, fascinating fluvial pattern, which represent the natural state of many gravel and sand bed rivers. Both natural and human causes may force a change in the boundary conditions, and consequently impact the river functionality. Detailed knowledge on the consequent morphological response is important in order to define management strategies which combine different needs, from protection of human activities and infrastructures to preservation of the ecological and biological richness. During the last decades, research has made significant advance to the description of this complex system, thanks to flume investigations, development of new survey techniques and, to a lesser extent, numerical and analytical solutions of mathematical models (e.g. Ashmore_2013). Despite that, many relevant questions, concerning the braided morphodynamics at different spatial and temporal scales (from the unit process scale, to the reach scale, and eventually to the catchment scale) remain unanswered. For example, quantitative analysis of the morphological response to varying external controls still requires investigation and needs the definition of suitable, stage-independent braiding indicators. In addition, the morphodynamics of the fundamental processes, such as bifurcations, also needs further analysis of the driving mechanisms. General aim of the present study is to develop new methods to exploit, in an integrated way, the potential of the new possibilities offered by advanced monitoring techniques, laboratory models, numerical schemes and analytical solutions. The final goal is to fill some gaps in the present knowledge, which could ultimately provide scientific support to river management policies. We adopted analytical perturbation approaches to solve the two-dimensional shallow water model; we performed laboratory simulations on a large, mobile-bed flume; we analysed existing topographic measurements from LiDAR and Terrestrial Laser scanning Devices; and we simulated numerically the river hydrodynamics. Within each of the six, independent, research chapters, we interconnected results from the different approaches and methodologies, in order to take advantage of their potential. Summarising, the more relevant and novel outcomes of the present work can be listed as follows: 1) We explored the morphological changes during a sequence of flood events in a natural braided river (Rees River, NZ)and we proposed a morphological method to assess the sediment transport rate. In particular we propose a semi-automatic method for estimating the particles path-length (Ashmore and Church, 1998) on the basis of the size of the deposition patches, which can be identified on the basis of DEM of differences. Comparison with results of numerical simulation confirmed that such an approach can reproduce the response of the bedload rate to floods of different duration and magnitude. 2) We developed a new indicator of the reach-scale morphology and, on the basis of existing laboratory experiments, we explored its dependence, under regime conditions, to the controlling factors: slope, discharge, confinement width, grain size. In spite of its synthetic nature, this simple indicator embeds the information needed to estimate the variability of the Shield stress throughout the braided network, and consequently enables to assess the transport-rate and its variation with the driving discharge. 3) We investigated, through flume experiments, the effect of the flow unsteadiness on the sediment transport in a braided river. This is possible only by following a statistical approach based on multiple repetitions of the same flow hydrograph. Results revealed that for confined network an hysteresis of the bedload response occurs, which leads to higher sediment transport during increasing flow, whereas relatively unconfined networks always show quasi-equilibrium transport rates. 4) A second set of laboratory experiments provided information on the morphodynamics of a braided network subject to variations of the sediment supply. We proposed a simple diffusive model to quantify the evolution of the one-dimensional bed elevation profile. Such simple approach, albeit having a limited range of practical applications, represents the first attempt to quantify this process and enables to study the relevant temporal and spatial scales of the phenomenon. 5) We solved analytically the two-dimensional morphodynamic model for a gravel-bed river bifurcation. This furnishes a rigorous proof to the idea proposed by Bertoldi and Tubino (2007) to interpret the morphological response of bifurcation in light of the theory of the morphodynamic influence. The analytical approach enables to investigate the fundamental mechanics which leads to balance, and unbalance, configurations and, from a more practical point of view, allows for a better prediction of the instability point than the existing 1D models (e.g. Bolla Pittaluga et al., 2003).

The link suspension is the most prevailing suspension system for freight wagons in central and western Europe. The system design is simple and has existed for more than 100 years. However, still its characteristics are not fully understood. This thesis emphasizes freight wagon dynamics and comprises three parts:

In the first part a review of freight wagon running gear is made. The different suspension systems are described and their advantages and disadvantages are discussed. The review covers the running gear standardized by UIC and the conventional so-called three-piece bogie. Additionally five improved three-piece bogies and twelve novel running gear designs are presented.

The second part focuses on the lateral force-displacement characteristics in the link suspension. Results from stationary measurements on freight wagons and laboratory tests of the link suspension characteristics are presented. To improve understanding of the various mechanisms and phenomena in link suspension systems a simulation model is developed. Link suspension systems have strongly nonlinear characteristics including a hysteresis loop. The loop exhibits usually three characteristic sections with different tangential stiffnesses. The actual contact geometry of the links and end bearings has a significant influence on the characteristics. By wear in ordinary service - as well as by geometric tolerances on new components - the contact geometry may deviate considerably from nominal geometry. Further, it seems that elastic deformation in the contact surfaces has considerable effects on the suspension characteristics, in particular on the initial rolling stiffness for small displacements. Also, flexibilities in links and end bearings influence the characteristics. It is also observed that new components after a short period of dynamic testing can exhibit a very low amount of energy dissipation, a phenomenon that is also indicated in some stationary measurements on wagons.

To summarize the second part, it appears that the link suspension characteristics are very sensitive to several factors being hard to control in the real world of freight wagon operations. The various stiffnesses and hysteresis loops have a considerable variation and may have a strong influence on the ride qualities of vehicles. As long as the characteristics can not be controlled within closer limits than found in this study, there is a strong need for sensitivity analysis to be made, both in predictive multibody simulations of vehicle dynamics, as well as in verification and acceptance tests.

In the third part a study on the possibility to improve ride qualities of freight wagons with link suspensions is presented. Parametric studies with multibody dynamic simulations on freight wagons equipped with link-suspension bogies are performed. The effect of supplementary friction and hydraulic damping is investigated under various running conditions: speed, loading, tangent and curved track, wheel-rail contact geometry, track gauge and track irregularities. Substantial improvements of the lateral running behaviour of wagons with link suspension bogies can be achieved - both at ordinary speeds and at increased speeds - by using a proper combination of supplementary hydraulic dampers. Speeds up to 160 km/h could be realistic.

Infill materials found in natural rock joints may cause a reduction in joint shear strength, influencing rock mass stability. The shear strength of rock mass, already reduced by these discontinuities, will further diminish if they are filled with sediments, thereby posing significant concerns for any construction or excavation carried out in rock. These concerns invite accurate quantification of the shear strength of infilled joints and proper understanding of the basic mechanics of discontinua and the principles involved in their shear deformation. The practical application of any models developed through such studies will be of immense help to mining, tunnelling, and all other underground construction works. The geotechnical research work carried out by the University of Wollongong in the late 90’s included infilled joint modelling using hyperbolic techniques. A new shear strength model was developed in these studies for predicting unfilled and infilled joint strength based on the Fourier transform method, energy balance principle and the hyperbolic stress-strain simulation. Taking into account the field conditions frequently encountered, the diversity observed in joint shear response and the occasional inadequacy of data (for the estimation of Fourier coefficients and the hyperbolic constants), this study was undertaken to develop a semi-empirical methodology for predicting the shear strength of infilled joints. In this research study joint shear behaviour was studied under CNS and CNL conditions and also the effect of joint orientation and confinement. The study aimed to develop a methodology which includes joint surface characteristics, joint properties, and infill materials. A new model for predicting the shear strength of infilled joints based on a series of tests carried out on two types of model joint surfaces (with asperity angles of 9.50 and 18.50) is presented. Graphite, bentonite and clayey sand were used as infill materials. All tests were carried out in a large-scale shear apparatus under constant normal stiffness (CNS) conditions. The results indicate that at low infill thickness to asperity height ratio (t/a), the combined effect of the basic friction angle (ϕb) and the joint asperity angle (i) is pronounced, but diminishes with increasing t/a ratio so that the shear strength converges towards the infill alone. This decrease in shear strength with increasing t/a ratio is represented in a normalised manner by dividing the peak shear stress by the corresponding normal stress. Summation of two algebraic functions (A and B) that represent the joint and infill characteristics, correctly model the decay of normalised shear strength with increasing t/a ratio. The new model successfully describes the shear strength of the graphite, clay (bentonite) and clayey sand filled model joints.

Fluctuating water table conditions influence capillary-held LNAPL(Light Non-Aqueous Phase Liquids) mass above and below the water table. Risks posed by such dynamic source zones vary over time as water tables oscillate from tidal effects, seasonality or anthropogenic interferences. In this study, the first automated multiphase flow dynamic water table experimental system comprising both hardware and software, was developed to: i) automatically implement programmable cyclic water table fluctuations via Raspberry Pi\(^T\)\(^M\) based inexpensive electronics; ii) dynamically monitor the real-time saturation distributions of all fluids (red-dyed-LNAPL, blue-dyed-water and air) in 2-D sand tank, using high-temporal-and-spatial resolution automated multi-spectral photography; and iii) accurately interpret large detailed datasets via advanced multi-spectral imaging. Such automated data acquisition and processing permit LNAPL releases and their redistributions under oscillating water table to be demonstrated in videos of photographic records, interpreted 2-D saturation contours and 1-D profiles. Eight experimental scenarios were undertaken to discern the influencing mechanisms of cyclic fluctuations incorporating with other influential factors including aquifer media and heterogeneities, volume and timing of releases, etc. Applicability of standard modelling by NAPL simulator was exercised, which provided a good general match of overall features of the release and oscillation dynamics. The high-resolution-and-frequency detailed quantitative dataset harvested was expected to supplement and expand the theories of multiphase flow distribution in porous media, where owing to the realization of the automated system, unprecedented processes were captured; and serve as a robust validation source of numerical and conceptual models which are essential tools in contamination site characterization, prediction, and remediation.

Proton exchange membrane fuel cells (PEMFCs) are promising candidates as power sources due to their high energy conversion efficiency, power density and low pollutants emission. Water management is of vital importance to achieve maximum performance and durability from PEMFCs. The main object of this work was to develop a mathematic model to better understand the water transport in PEMFCs under practical conditions. The aim is to enhance the output power of fuel cells by establishing effective water removal and distribution strategies. A single-phase flow, along the channel, isothermal model of a PEMFC is developed and validated against experimental data. Reactant flow and diffusion are simulated using the Navier-Stokes equation and Maxwell-Stefan equation, respectively. Water transport through the membrane is described by the combinational mechanism in which electro-osmotic drag, back diffusion and hydraulic permeation are all included. Agglomerate assumption is applied for the catalyst layer structure. This model is used to study the effects of the catalyst layer properties on cell performance. The model indicates that the rapid decrease in current density at lower cell voltage is due to an increased oxygen diffusion resistance through the ionomer film. A two-phase flow, across the channel, isothermal model is developed. The water phase transfers between water vapour, dissolved water and liquid water are taken into account and liquid water formation and transport are introduced. Liquid water occupies the secondary pores of the cathode catalyst layer to form a liquid water film on the outer boundary of the ionomer film. This model is used to study the influence of catalyst layer parameters and operating conditions on the cell performance. The model provides useful guidance for optimisation of the ionomer volume fraction in the cathode catalyst layer and the relative humidity of the cathode gas inlet. A two-phase flow, across the channel, non-isothermal model is developed. The model considered the non-uniform temperature distribution within the fuel cell. The modelling results show that heat accumulates within the cathode catalyst layer under the channel. Higher operating temperatures improved the fuel cell performance by increasing the kinetic rate, reducing the liquid water saturation on the cathode and increasing the water carrying capacity of the anode gas. Applying higher temperature on the anode and enlarging the width ratio of the channel/rib could improve the cell performance. A multi-variable optimisation of the cathode catalyst layer composition is represented by a surrogate modelling. Five design parameters, platinum loading, platinum mass ratio, ionomer volume fraction, catalyst layer thickness and agglomerate radius, are optimised by a multiple surrogate model and their sensitivities are analysed by a Monte Carlo method based approach. Two optimisation strategies, maximising the current density at a fixed cell voltage and within a specific range, are implemented for the optima prediction. At higher current densities, cell performance is improved by reducing the ionomer volume fraction and increasing the catalyst layer porosity. The one-dimensional, isothermal, time dependent and steady state models for the anode of a direct methanol fuel cell (DMFC) are developed. The two models are based on the dual-site mechanism, in which the coverage of intermediate species of methanol, OH and CO on the surface of platinum and ruthenium are included. Both the effect of operating conditions and electrode parameters are investigated. The distributions of methanol concentration and overpotential inside the electrode are represented and the current densities predicted by the intrinsic and macro kinetics are compared. From the analysis of the different models developed in this thesis, the main results can be summarised as: (1) Mass transport resistance resulted from the oxygen diffusion through the ionomer film surrounding the agglomerate is the main reason for the rapid fall of current density at lower cell voltage. (2) Ionomer swelling has a significant effect on fuel cell performance because it resulted in a decrease in the porosity and an increase in the ionomer film thickness, leading to an increase in the oxygen transport resistance. (3) Catalyst layer composition has a vital impact on the utilisation of the platinum catalyst and cell performance. (4) Heat accumulates within the cathode catalyst layer under the channel. Applying higher temperatures on the anode optimises the temperature distribution within the MEA and improves the cell performance. (5) Cell performance is improved by enlarging the width ratio of channel/rib. However, the improvement is limited by the sluggish oxygen reduction reaction. (6) For the methanol oxidation reaction in a Pt-Ru anode, the intrinsic current density is determined by the coverage ratios of the intermediate species. The structure and property of the electrode also play an important role in determining the anode performance of a DMFC.

The fatigue performance of asphalt is addressed in the mechanistic-empirical pavement thickness design procedure used in Australia. The aim of this research program was to investigate the laboratory fatigue performance of a typical asphalt mix used in Western Australia as a structural layer, including the development of new testing protocols, a new fatigue life prediction model, an accelerated testing approach, and master curves that can be input into the M-E design procedures used in Australia.

Afin de gérer au mieux le pilotage des unités d'ozonation lors de la potabilisation des eaux, un modèle prédictif a été développé. L'objectif du modèle était de pouvoir prédire, sur une installation industrielle, les concentrations en ozone, en bromates et en différents micropolluants. Le modèle chimique proposé est mécanistique et peut être subdivisé en plusieurs parties : auto-décomposition de l'ozone, influence de l'alcalinité, formation de bromates, influence de la MON (Matière Organique Naturelle). Le modèle d'influence de la MON comporte 12 paramètres ajustables, le modèle pour la formation des bromates comporte un paramètre ajustable, les valeurs des autres paramètres sont fixées d'après la littérature. Le modèle hydraulique est de type systémique et comprend des réacteurs idéaux (parfaitement agités et piston). L'identifiabilité du jeu de paramètres a été conduite par une analyse de sensibilité (eFAST). La procédure d'optimisation par méthode de Nelder-Mead a été testée. Le modèle proposé permet de rendre bien compte des variations de pH, de température, de concentration de MON, de doses d'ozone sur la décomposition de l'ozone et la génération de radicaux. Les essais sur la formation de bromates ont montré que le modèle donne de bons résultats pour des concentrations inférieures à 20 µg. L-1, ce qui est particulièrement intéressant dans le cas d'une application industrielle. Enfin, une étude sur une unité industrielle a montré que des modèles calibrés en laboratoire (chimie, hydraulique) peuvent être appliqués directement sur site. Le modèle de formation des bromates est néanmoins instable dans le temps et doit être périodiquement réajusté
Facing major challenges, management of ozonation process will increasingly need prediction tools based on modelling. Dealing with different types of waters (chemistry) and different types of tanks (hydraulics), modelling of ozonation units has to adapt to site-specific conditions. The main objective of this work was to develop an integrated modelling procedure for industrial ozonation processes for predicting concentration profiles of: ozone, bromate and specific micropollutants. Two types of chemical models were considered: semi-empirical models with adjustable kinetics (for the role of Natural Organic Matter, NOM) and mechanistic models with predetermined kinetics (for other phenomena related to ozonation). Hydraulic flow conditions were modelled by systematic networks (patterns of ideal reactors). Ozone decomposition, hydroxyl radical generation and bromate formation were studied at lab-scale with a specially developed apparatus. Various experimental conditions were tested: pH, temperature, ozone doses, initial bromide concentrations, concentration of NOM and nature of NOM. The model was able to adequately reproduce experimental measurements for nine of the eleven water samples studied, covering a wide domain: 6. 1 ≤pH ≤8. 15; 1. 02 meq. L-1≤ AT ≤6. 02 meq. L-1; 0-. 0. 5 mg. L-1≤TOC≤3. 1 mg. L-1. For bromate, considering the crucial zone between the quantification limit and 20 µg. L-1, a large majority of the simulated concentrations corresponded to experimental concentrations or were located in the experimental uncertainty interval. A full-scale study showed that models calibrated at lab-scale may be used directly on-site to predict the formation of bromate and ozone profiles. Readjustments of the model for bromate formation should however account for seasonal changes

This research provides a multiscale geomechanical characterization workflow for ultra-tight and anisotropic Goldwyer gas shales by integrating field appraisal, laboratory deformation and ultrasonic testing, viscoelastic modelling, and hydraulic fracture simulation. The outcome of this work addresses few of the practical challenges in unconventional reservoirs including but not limited to (i) microstructure & compositional control on rock mechanical properties, (ii) robust estimation of elastic anisotropy, (iii) viscous stress relaxation to predict the least principal stress Shmin at depth from creep, (iv) influence of specific surface area on creep, and (v) impact of stress layering on hydraulic fracturing design.

Evolution of the traditional consumer in a power system to a prosumer has posed many problems in the traditional uni-directional grid. This evolution in the grid model has made it important to study the behaviour of microgrids. This thesis deals with the laboratory microgrid setup at the Munich School of Engineering, built to assist researchers in studying microgrids. The model is built in Dymola which is a tool for the OpenModelica language. Models for the different components were derived, suiting the purpose of this study. The equivalent parameters were derived from data sheets and other simulation programs such as PSCAD. The parameters were entered into the model grid and tested at steady state, firstly. This yielded satisfactory results that were similar to the reference results from MATPOWER power flow. Furthermore, fault conditions at several buses were simulated to observe the behaviour of the grid under these conditions. Recommendations for further developing this model to include more detailed models for components, such as power electronic converters, were made at the end of the thesis.

Dynamic simulation programs were created in the Python programming language, to describe a laboratory scale, sub-zero distillation column, that is used to separate mixtures of tetrafluoroethylene (TFE), hexafluoropropylene (HFP) and octafluorocyclobutane (OFCB). Both the equilibrium and rate-based modelling approaches were taken, to generate a comparison between the efficiency and simulation time of both models. A physical properties data bank for the three components had to be created, as the main and many of the sub-models require physical or thermodynamic properties for evaluation. The different physical property models, found in literature, were programmed into functions that could easily return the wanted property, given a set of required inputs. The applicable mixing rules for each property type was also programmed into functions, to allow for easy retrieval. The vapour-liquid equilibrium (VLE) model used, is also one that comes from literature and is based on parameters for the three binary systems. The VLE model consists of the Peng-Robinson equation of state, that utilises the Mathias-Copeman alpha function and the Wong-Sandler mixing rules, to describe the vapour phase. The liquid phase is described by the non-random two liquid (NRTL) activity coefficient model. Furthermore, the γ-Φ VLE formulation was used to put the thermodynamic model together. These models were also written into functions to serve as simulation building blocks. Mass and energy transfer on packed sections in the rate-based model was described by the Maxwell-Stefan diffusion model. The form of this model that was utilised, is the matrix-based, exact solution of the Maxwell-Stefan equations, under the two-film theory. This model was slightly simplified by assuming that the corrective flux matrix reduces to the identity matrix- an assumption that is regularly made in distillation modelling. Emphasis was laid in documenting how the models are put together to build the simulations. Dynamic simulation algorithms rarely accompany distillation models reported in literature, or authors make use of commercial software to order the modelling equations for them. One of the objectives of the research presented here was, therefore, to report on the process developed to solve the problem. Both simulation programs delivered typical responses that can be expected of distillation systems. The actual change in the magnitude of the values, however, proved to be significantly small. The cause of this, being the large liquid molar hold-up values that were produced by the model initialisation. The feed flow rate, in comparison, is too small to bring about a significant effect when suddenly increased. This could mean that the system is not capable of reaching the steady-state produced by the initialisation (as the feed cylinder may be too small to contain the required amount of feed gas) and that the column may have to be run in a continuous dynamic state. To be sure of this, however, the model will first need to be validated against experimental data. Furthermore, the simulation programs proved to progress very slowly, particularly the simulation built around the rate-based model. A time step-size of 0.5 resulted in an integration time around 1 minute and 20 seconds for the equilibrium model, while the rate model ran for over 19 minutes, both for a timespan of 300 s. It is recommended that future research focuses on building start-up simulations for the models, to provide better initial results and to give more insight into the operation of the column. Experimental validation of the models is also important, to establish their accuracies. Finally, work has to be done to improve the simulation speeds, especially if it is required that one of the models are integrated into the column's control system.
Dissertation (MEng)--University of Pretoria, 2019.
Fluorochemical Expansion Initiative
Department of Science and Technology
Department of Trade and Industry
Chemical Engineering
MEng
Unrestricted

Assessment of the potentially acidic, heavy metal-ladenleachates that leave deposits of sulfide ore mill tailings andevaluation of various possible options for mill tailingremediation are scientific problems of increasing practicalimportance. High costs may be associated with the mill tailingremediation, not least after recent changes in Swedish andEuropean environmental legislation. This thesis presents amethodology for studying and quantifying geochemical processesthat contribute to generation of so-called acid mine drainage(AMD). The methodology builds from first principles regardinggeochemical processes, and is based on geochemicalcharacterisation of the mill tailings combined with explicitmodel quantification of the effect of factors, such astemperature, pH, and mineral (BET) surface area, that influencemineral weathering rates. Application of the modellingmethodology to a case study site, Impoundment 1, Kristineberg,northern Sweden, including quantification of slow processesthrough literature rate laws, successfully reproduced the pHand relative concentrations of major ions in the impoundmentgroundwater. Absolute concentrations of most major ions, withthe exception of Zn, were 1-2 orders of magnitude higher in themodel than in the field, which is consistent with the commonlyobserved scale dependence of mineral weathering rates; however,application of a single calibration factor, X_r=10^-2, to all weathering rate expressions, sufficed toaccount for this apparent scale dependence.

Subsequent laboratory determination of mineral weatheringrates in Impoundment 1 tailings indicated that rates for themajor minerals pyrite (FeS₂) and aluminosilicates were in fact 1-2 orders ofmagnitude lower in the ~50-year-old tailings than ratesreported in the literature. Weathering rates of chalcopyrite(CuFeS₂) and sphalerite (ZnS) were by contrast 1-3 ordersof magnitude greater than predicted by the literature rate lawsthat were used in the modelling study. While the mechanism ofZn release requires further investigation for improved forwardmodel prediction, the underestimation of Zn concentration inImpoundment 1 by the model was resolved. The laboratory studyfurthermore indicated that the weathering rates of most majorminerals exhibited the same dependence on pH, temperature andsurface area as reported in the literature, and therebysupported the use of literature rate laws for model assessmentof dominant geochemical processes in tailings deposits, onceallowance is made for lower rates in older tailingsmaterial.

Analysis of the dominant geochemical processes in the modelof Impoundment 1 indicated that slow weathering ofaluminosilicate minerals provided the bulk of protonattenuation and, as a result, considerably affected the rate ofdepletion of fast-reacting pH-buffering minerals. Inclusion ofthe kinetics of aluminosilicate dissolution and of thefeedbacks between slow and fast processes is thus potentiallycrucial for prediction of pH and its long-term evolution. Thesensitivity of modelled groundwater composition and pH to ironredox reactions, such as may be accelerated by acidophilicbacteria, indicated that, while iron redox cycling was low atthe present case study site, quantification of microbialmediation of these reactions may be necessary for predictingAMD quality under other conditions. The laboratory studies alsoindicated that application of common sterilisation techniques,such as is necessary for study of relative contributions ofabiotic and biotic weathering processes, had little effect onthe long-term (>30 days) abiotic element release rates inthe tailings.

This study suggests that within certain limits, which appearnarrower than currently recognised in industrial predictionpractices, it is possible to predict the weathering behaviourof major minerals, and hence proton release and attenuation, inbase metal tailings under field conditions.

Integrated geophysical, geochemical and hydrogeological modelling techniques have been used in this PhD research study to characterise the ground contamination at four study sites. The primary aim of the research is to study the efficiency of using an integrated approach to study hydrocarbon and heavy metal contamination in the subsurface soil, geology and groundwater systems. Each of the four sites have different industrial legacies (hence types of contamination) and periods of operation. The sites in increasing age are: Bromsgrove landfill, Cathays Railway Depot, Barry Docks and Weston-Super-Mare gasworks. The geophysical methods used in this study are Electrical Resistivity imaging (IRIS) and multifrequency ground conductivity (Gem-2). These techniques can map the electrical properties of the sub-surface in both 2D and 3D. The electrical properties are then compared with independent professional geochemical sampling programmes for both soil and groundwater. The results of this project show that the geophysical methods provide a clear indication of the zones of ground contamination (hydrocarbons and heavy metals) in both 2D and 3D. Electrical resistivity/conductivity mapping using both independent techniques show consistent results. The geophysical results also show a good correlation with the geochemical sampling programmes for soil, water and gas. The results are encouraging in that geophysics could be used more widely in professional contaminated land site investigations to supplement conventional invasive techniques and sampling programmes. The research has also investigated whether geophysical parameters such as electrical conductivity could be used as a groundwater modelling parameter in the Visual Modflow software. The advantage being that geophysics can provide a high intensity of data in 3D and even temporal 4D at a site. The geophysical results can define an accurate spatial distribution of the contaminated area. This can be used to assign a ‘concentration recharge’ in the modelling compared to the normal practice of using chemical data from a small number of discrete sampling points. However this approach will still require chemical data for validation.

A number of studies have focused on urban trees to understand their mitigation capacity of air pollution. Present PhD project aims to describe Particulate Matter (PM) accumulation on tree leaves in a specific urban environment as case study, both in terms of quantity and quality. Furthermore, experimental data obtained from laboratory analysis are crossed with an air pollution model designed for urban environments, in order to relate PM deposition to main polluting sources. PM deposition on Quercus ilex leaves was quantitatively analyzed in four districts of the City of Terni (Italy) for three periods of the year. Fine (between 0.2 and 2.5 µm) and Large (between 2.5 and 10 µm) PM fractions were analyzed. Mean PM deposition value on Quercus ilex leaves was 20.6 µg cm-2. Variations in PM deposition correlated with distance to main roads and downwind position relatively to industrial area. Epicuticular waxes were measured and related to accumulated PM. For Fine PM deposited in waxes we observed a higher value (40% of total Fine PM) than Large PM (4% of total Large PM). Model chosen to perform the air quality simulation on the study area was the TAPM (The Air Pollution Model) and ran for 1 year period on study area (Jan 2012 – 2013). This model allowed describing PM dispersion pattern within the city, according to the main polluting sources position. TAMP implementation evidenced areas potentially more polluted in the case study city and the local effect of air pollution sources. Qualitative analysis is obtained through the application of a monitoring and characterization protocol for PM deposited on Q.ilex leaves. Four trees were selected as representative of urban environments based on their proximity to a steel factory and a street. PM from quantitative analysis was characterized using scanning electron microscopy and energy dispersive X-ray spectroscopy, inferring the associations between particle sizes, chemical composition, and sampling location. Modeling of particle size distributions showed a tri-modal fingerprint, with the three modes centered at 0.6 (factory related), 1.2 (urban background), and 2.6 μm (traffic related). Chemical detection identified 23 elements abundant in the PM samples. Principal component analysis recognized iron and copper as source-specific PM markers, attributed mainly to industrial and heavy traffic pollution respectively. A similar technique implemented for qualitative analysis is being used in the “European Sampling Campaign” a huge PM characterization project, in collaboration with Antwerp University within COST FP1204. Platanus sp. leaves were collected in 20 different cities and analyzed. SEM/EDX was supported by SIRM (Saturation Isothermal Remnant Magnetization) analysis. First results are reached, nevertheless the work is still in progress: the largest part of analyzed particles is fine PM; not direct association between site and PM dimension or density on leaves is found. PM quality is different between upper or lower leaf side. Iron is confirmed as “technology” marker. There is a relation between Fe concentration and SIRM results. Results from this study allow to increase our understanding about air pollution interactions with urban vegetation and could be hopefully taken into account when guidelines for local urban green management are realized. Upscale of qualitative results on leaf area basis provided a useful indicator for strategic evaluation of harmful PM pollutants using tree leaves.
Numerosi studi sono stati svolti sul ruolo degli alberi in ambiente urbano nella mitigazione dell’inquinamento atmosferico. Il progetto qui presentato ha come obiettivo di descrivere l’accumulo del Particolato Sottile (PM) in uno specifico ambiente urbano, sia da un punto di vista quantitativo che qualitativo. I dati di laboratorio sono stati confrontati con gli output di un modello progettato per ambienti urbani, al fine di collegare la deposizione del PM alle principali sorgenti di inquinanti. La quantità di PM depositata su foglie di Quercus ilex (leccio) è stata analizzata in quattro quartieri della città di Terni (Italia), per tre periodi dell’anno. È stato analizzato sia il PM fine (0.2 e 2.5µm) che il grossolano (2.5 e 10µm). Il valore medio di deposizione sulle foglie è 20.6 µg cm-2. La variabilità di deposizione è legata a 2 fattori: distanza dalla strada e posizione sottovento relativamente all’area industriale. Le cere cuticolari sono state quantificate. Per il PM fine è stato osservato un valore maggiore di accumulo nelle cere (40% del fine totale) rispetto al grossolano (4% del grossolano totale). Il modello utilizzato nell’area di studio è il TAPM (The Air Pollution Model; Australia). È stato applicato su un anno (Gen’12 – Gen’13). L’analisi ha permesso di descrivere la modalità di diffusione del PM nell’area di studio, in base alla posizione delle principali sorgenti inquinanti. L’applicazione di TAPM ha evidenziato le aree potenzialmente più inquinate e l’effetto delle sorgenti locali. Per l’analisi qualitativa sono stati selezionati quattro alberi, come rappresentativi di differenti ambienti urbani. La scelta è basata su: distanza dalla strada, vicinanza all’acciaieria. Il PM raccolto tramite analisi quantitativa è stato analizzato attraverso microscopia elettronica a scansione e spettroscopia a raggi-X. La relazione tra dimensione del PM, composizione chimica e localizzazione dell’albero è stata evidenziata. La distribuzione delle dimensioni delle particelle è descritta tramite una struttura tri-modale: 0,6 (acciaieria), 1,2 (fondo urbano) e 2,6 (strada). Sono stati individuati 23 elementi come componenti del PM. L’analisi delle componenti principali (PCA) ha identificato il ferro e il rame quali traccianti di sorgenti specifiche: area industriale e traffico. Una simile analisi qualitativa è ancora in uso nella “Campagna Europea di Campionamento”, ampio progetto di caratterizzazione del PM depositato su foglie di Platanus sp., in collaborazione con l’Università di Anversa. I campioni sono stati raccolti in 20 città. L’analisi SEM/EDX è supportata da quella SIRM (Saturazione Isotermica Magnetizzazione Rimanente). Il lavoro è ancora in corso, i primi risultati vengono presentati. C’è una generale omogeneità di PM all’interno di ogni città, e si può apprezzare una variazione dimensionale e di densità di particelle solo tra pagina inferiore e superiore della foglia. Il ferro è un probabile tracciante del traffico. È stata trovata una buona relazione tra % di ferro e dati SIRM. I risultati di tale studio hanno permesso di approfondire la comprensione riguardo le interazioni tra vegetazione urbana e inquinanti atmosferici. È auspicabile che venga preso in considerazione per la realizzazione di linee guida per la gestione del verde urbano. L’upscale dei risultati qualitativi fornisce degli utili indicatori per eventuali progetti di mitigazione del PM tramite foresta urbana.

The introduction of computer-aided tools into the product development process allows improving the quality of the product, evaluating different variants of the same product in a faster way and reducing time and costs. They can play a meaningful role also in designing custom-fit products (especially, those characterized by a tight interaction with the human body), increasing the comfort and improving people’s quality of life. This thesis concerns a specific custom-fit product, the lower limb prosthesis. It is part of a research project that aims at developing a new design platform centred on the digital model of the patient and his/her characteristics. The platform, named Prosthesis Virtual Laboratory (PVL), is being developed by the V&K Research Group (University of Bergamo) and integrates ICT tools and product-process knowledge. It provides two environments: one for prosthesis design (named Prosthesis Modelling Lab), both transfemoral and transtibial, and one for the prosthesis testing (named Virtual Testing Lab). The main objective has been to embed within the Virtual Testing Environment numerical simulation tools to analyse the interaction between the socket and the residual limb under different loading conditions, allowing the prosthetist to automatically run the simulation and optimize socket shape. Simulation tools, such as Finite Element Analysis (FEA), permit to predict the pressures at the interface socket-residual limb, evaluate the comfort of socket and validate the socket design before manufacturing phase. However, the diffusion of simulation tools in orthopaedic laboratories is strongly limited by the high level of competence required to use them. Furthermore, the implementation of the simulation model is time consuming and requires expensive resources, both humans and technological, especially onerous for small orthopaedic labs. To effectively employ the numerical analysis in prosthesis design, the simulation process has been automated and embedded within the virtual design platform. Therefore, in such a context, the specific scientific objectives have been to: • Critically analyse the state of the art with regard to methods and tools to evaluate socket-residual limb interaction. • Identify the key issues to automate the simulation activities. • Define a set of simulation rules and the Finite Element Analysis model. • Implement and integrate within the new design platform the automatic simulation procedure. • Test the integrated design platform with a case study. • Identify future development trends. Research activities have been organized into four main activities as follows. The first activity consisted in an extensive analysis of the last two decades State of the Art on numerical models adopted to study residual lower-limb and prosthetic socket interaction. Starting from literature, the key issues of the simulation process (e.g., geometric models reconstruction, materials characterization, simulation steps, and boundary conditions), the methodologies and procedures have been identified. Particular attention has been also paid to the parameters commonly adopted to evaluate socket comfort. This phase played a fundamental role since it constituted the basis for the implementation of the embedded simulation procedure. It also permitted to highlight that current finite element models are stand-alone and not integrated with prosthetic CAD or Digital Human Modelling (DHM) systems. In the second activity the tools and methods necessary to develop the embedded simulation module have been selected. By using these tools, it was possible to identify the simulation rules and the best practice procedures, which are fundamental to implement an automatic simulation module. Initially, the modelling tools have been considered since they provide the geometric models for the numerical analysis of the socket-residuum interaction and for the virtual gait analysis of the patient’s avatar. Then, particular attention has been paid on the choice of the FE solver, that has been made according to the results of preliminary FE models. They were implemented using two different solvers: Abaqus (commercial) and CalculiX (open-source). The latter has been experimented to verify the possibility to develop a design platform totally independent from commercial tools. However, according to the results, Abaqus has been chosen because it allows managing adequately simulation problems characterized by large deformations and difficult contact conditions, its results are comparable with those found in literature, and its scripting code does not require specific customization. The last considered tool was the Digital Human Modelling system (LifeMOD) since it permits to enhance the accuracy of the numerical analysis. By performing the gait simulation of the patient’s avatar, it provides the directions and the magnitude of forces and moments that act on the socket. The third activity consisted in defining the architecture of the simulation module, implementing the module and the interfaces with the socket CAD tool (namely Socket Modelling Assistant-SMA) to get the geometric models of the involved parts (socket and residual limb) and with the DHM system to acquire forces acting on the socket during patient’s walking. The simulation module has been implemented using the Python language and the integrated environment works as follows. Once the prosthetist has created the 3D socket model, SMA acquires the input for the analysis (e.g., residual limb length, patient’s weight, friction coefficient, material properties), and produces the files required to generate the FE model. Abaqus automatically generates the FE model without any human intervention, solves the analysis and generates the output file containing the pressure values. Results are imported in SMA and visualized with a colour map. SMA evaluates pressure distribution and highlights the areas that should be modified. Geometry modifications are needed in the areas where pressure exceeds the maximum value and are carried out automatically by the system or by the prosthetist using the virtual tools available in SMA. Then, the system re-executes the simulation. Through this iterative process of adjustments, the socket shape is modified and optimized in order to eliminate undercuts, minimize weight and, especially, distribute loads in the appropriate way so that they can be tolerated for the longest period of time. The fourth and last activity concerned the test and validation of the simulation module integrated within the new design platform, by considering a transfemoral patient. The new virtual process and the key issues of the simulation procedure have been tested starting from the patient’s data acquisition to the release of the socket using also data coming from the gait simulation with the DHM system. The geometric model of the residual limb has been reconstructed from MRI images and the socket has been modelled using SMA. Through an iterative process, the socket shape has been optimized until the pressure distribution on the residuum was consistent. Preliminary activity concerning the FE model validation has been performed comparing the pressure distribution experimentally acquired with pressure transducers over the residuum with the simulation results. To accomplish this task, the geometric model of the real socket has been acquired using reverse engineering techniques. Two numerical simulations have been implemented, they differ for the residuum geometric models adopted: from MRI and from 3D scanning. Preliminary results have been considered positive but improvements are necessary. As an example, some geometric inconsistencies, occurred during the acquisition of the geometric model of the residual limb, have reduced the accuracy of the final results. To complete the evaluation of the simulation model, a new residuum geometric model is needed and a refinement of the material model characterization is desirable. To conclude, the simulation module embedded within Virtual Testing Laboratory has improved the prosthesis development process with the goal of assessing and validating the socket shape under different load conditions (static or dynamic) before the manufacturing phase. The testing phase of the new procedure has demonstrated the feasibility of the virtual approach for lower limb prosthesis design. The tests carried out permitted to highlight necessary improvements and future developments, such as the definition of a protocol to acquire the residual limb through MRI and 3D scan, refinement of the FE model (e.g., non-linear viscoelastic behaviour for soft tissues, friction coefficients), parallel computing to improve simulation performances, open-source solvers to implement a design platform totally independent from commercial systems, and a massive test campaign involving transtibial and transfemoral patients to fully validate the FE model and the design platform.
L’introduzione di strumenti informatizzati nel processo di sviluppo del prodotto permette di migliorarne la qualità, nonché di valutare diverse varianti del prodotto stesso in modo più veloce, riducendo in tal modo il tempo ed i costi relativi alla progettazione. Per queste motivazioni, tali strumenti possono giocare un ruolo rilevante anche nella realizzazione di prodotti personalizzati (specialmente quelli caratterizzati da una stretta interazione con il corpo umano), aumentandone il comfort e migliorando la qualità di vita delle persone. Il presente lavoro di tesi si concentra nello specifico sull’applicazione di tali strumenti informatizzati nella creazione di protesi per arti inferiori, inserendosi in un progetto di ricerca che ha come obiettivo quello di sviluppare una nuova piattaforma di progettazione centrata sul modello digitale del paziente e sulle sue caratteristiche. La piattaforma, chiamata Prosthesis Virtual Laboratory (PVL), è stata sviluppata dal gruppo di ricerca V&K dell’Università degli Studi di Bergamo nell’ottica di integrare gli strumenti informatici con la conoscenza del prodotto e del processo. La piattaforma è strutturata in modo da offrire due ambienti di lavoro: uno dedicato alla progettazione della protesi (chiamato Prosthesis Modelling Lab), sia transfemorale che transtibiale, e l’altro destinato alla fase di verifica della stessa (chiamato Virtual Testing Lab). L’obiettivo principale del lavoro di tesi è stato quello di integrare, all’interno dell’ambiente virtuale di verifica, gli strumenti di simulazione numerica che consentono di analizzare l’interazione tra l’invaso e l’arto residuo sotto diverse condizioni di carico, permettendo al tecnico protesico di effettuare la simulazione in automatico e di ottimizzare la forma dell’invaso. Gli strumenti di simulazione, come l’analisi agli elementi finiti (FEA), permettono di predire la pressione all’interfaccia tra invaso e moncone, di valutare il comfort dell’invaso e di validare la progettazione dello stesso prima della fase di manifattura. Tuttavia, la diffusione degli strumenti di simulazione nei laboratori ortopedici è fortemente limitata dall’elevato livello di competenze richieste per ottenere risultati significativi. Inoltre, l’implementazione di un modello di simulazione numerica richiede tempo e costose risorse, sia umane che tecnologiche, particolarmente onerose per i piccoli laboratori ortopedici. Affinché l’analisi numerica sia utilizzata nella progettazione delle protesi, è necessario che il processo di simulazione sia automatico ed integrato all’interno di una piattaforma virtuale di progettazione. In questo contesto, gli obiettivi scientifici specifici sono stati: • Analizzare criticamente lo stato dell'arte riguardante i metodi e gli strumenti per valutare l'interazione tra invaso ed arto residuo. • Identificare le questioni chiave per automatizzare le attività di simulazione. • Definire un insieme di regole di simulazione ed il modello per l’analisi ad elementi finiti. • Implementare ed integrare nella nuova piattaforma di progettazione la procedura di simulazione automatica. • Verificare la piattaforma di progettazione integrata con un caso studio. • Identificare le tendenze di sviluppo futuro. Le attività di ricerca sono state organizzate in quattro attività principali, come di seguito presentato nello specifico. La prima attività è consistita in un'analisi approfondita dello stato dell’arte negli ultimi due decenni relativamente ai modelli numerici adottati per studiare l’interazione tra invaso ed arto residuo. Partendo dalla letteratura, sono stati individuati i temi chiave del processo di simulazione (ad esempio la ricostruzione dei modelli geometrici, la caratterizzazione dei materiali, le fasi di simulazione e le condizioni al contorno), nonché le metodologie e le procedure di simulazione. Particolare attenzione è stata posta anche ai parametri comunemente adottati per valutare il comfort dell’invaso. Questa fase ha giocato un ruolo fondamentale in quanto costituisce la base per l’implementazione della procedura di simulazione integrata. Ha permesso altresì di evidenziare come gli attuali modelli agli elementi finiti siano indipendenti e non integrati con i sistemi CAD per protesi o di Digital Human Modelling (DHM). La seconda attività ha avuto come focus la selezione degli strumenti e dei metodi necessari allo sviluppo del modulo di simulazione, per mezzo dei quali è stato possibile identificare le regole di simulazione e le procedure di buona prassi, fondamentali per l’implementazione di un modulo di simulazione automatica. Inizialmente, gli strumenti di modellazione sono stati presi in considerazione in quanto forniscono i modelli geometrici sia per l’analisi numerica dell’interazione tra invaso ed arto residuo che per l’analisi della camminata virtuale dell’avatar del paziente. In seguito, particolare attenzione è stata posta sulla scelta del solutore a elementi finiti, che è stata fatta in accordo con i risultati ottenuti dai modelli preliminari implementati utilizzando due diversi solutori: Abaqus (commerciale) e CalculiX (open-souce). Quest’ultimo è stato impiegato per verificare la possibilità di sviluppare una piattaforma di progettazione totalmente indipendente dagli strumenti commerciali. Tuttavia, in base ai risultati ottenuti, la scelta si è indirizzata verso Abaqus, in quanto permette di gestire in modo adeguato i problemi di simulazione caratterizzati da grandi deformazioni e da difficili condizioni di contatto. L’utilizzo di questo solutore consente di ottenere risultati paragonabili a quelli presenti in letteratura ed inoltre il suo codice di script non richiede specifiche personalizzazioni. L’ultimo strumento utilizzato è stato il sistema DHM (Digital Human Modelling ) che permette di aumentare la precisione dell’analisi numerica. Attraverso l’analisi della camminata virtuale dell’avatar del paziente, questo strumento è in grado di fornire le direzioni e le intensità delle forze e delle coppie che agiscono sull’invaso. La terza attività ha riguardato la definizione dell’architettura del modulo di simulazione, l’implementazione del modulo stesso e del suo interfacciamento prima con lo strumento CAD per l’invaso (chiamato Socket Modelling Assistant - SMA), allo scopo di ottenere i modelli geometrici delle parti coinvolte (invaso ed arto residuo), ed in seguito con il sistema DHM, per acquisire le forze che agiscono sull’invaso durante la deambulazione del paziente. Il modulo di simulazione è stato implementato utilizzando il linguaggio Python e l’ambiente integrato prevede diverse fasi di sviluppo, come di seguito approfondito. Una volta che il tecnico protesico ha creato il modello 3D dell’invaso, lo SMA acquisisce gli input per l’analisi (come la lunghezza dell’arto residuo, il peso del paziente, il coefficiente di attrito, le proprietà dei materiali) e rilascia i file richiesti per generare il modello agli elementi finiti. Abaqus genera automaticamente il modello di simulazione senza che vi sia alcun intervento umano, risolve l’analisi e genera il file di output contenente i valori di pressione. I risultati sono importati nello SMA e visualizzati con una mappa di colore. La modifica della geometria dell’invaso, necessaria nelle aree in cui la pressione eccede i valori massimi, è eseguita in automatico dal sistema o dal tecnico protesico tramite gli strumenti virtuali presenti nello SMA. Il sistema, quindi, riesegue la simulazione. Attraverso questo processo iterativo di rettifica, la forma dell’invaso è modificata ed ottimizzata al fine di eliminare i sottosquadri, minimizzare il peso e soprattutto distribuire i carichi in modo appropriato, così che siano tollerabili per lunghi periodi di tempo. La quarta ed ultima attività ha riguardato la sperimentazione e la validazione del modulo di simulazione integrato all’interno della nuova piattaforma di progettazione considerando un paziente transfemorale. Il nuovo processo virtuale e le questioni chiave della procedura di simulazione sono state testate partendo dall’acquisizione dei dati del paziente fino al rilascio dell’invaso definitivo, utilizzando anche i dati provenenti dalla simulazione della camminata con il sistema DHM. Il modello geometrico dell’arto residuo è stato ricostruito partendo dalle immagini MRI e l’invaso è stato modellato utilizzando lo SMA. Attraverso un processo iterativo, la forma dell’invaso è stata ottimizzata fino ad avere una distribuzione appropriata della pressione sul moncone. L’attività preliminare riguardante la validazione del modello agli elementi finiti è stata eseguita comparando la distribuzione delle pressioni acquisite sperimentalmente sul moncone con i risultati della simulazione. Per realizzare questo compito, il modello geometrico dell’invaso reale è stato acquisito utilizzando tecniche di reverse engineering. Sono state implementate due diverse simulazioni numeriche che differiscono per il modello geometrico del moncone adottato: attraverso MRI nel primo caso, da scansione 3D nel secondo. I risultati preliminari possono considerarsi positivi ma ulteriori sviluppi sono necessari. Ad esempio, alcune incongruenze geometriche che si sono verificate durante l’acquisizione del modello geometrico hanno ridotto la precisione dei risultati finali. Per completare la valutazione del modello di simulazione è quindi necessario utilizzare un nuovo modello geometrico del moncone e sarebbe anche auspicabile raffinare il modello di caratterizzazione del materiale. Concludendo, il modulo di simulazione integrato all’interno del Virtual Testing Laboratory – VTL ha permesso di migliorare il processo di sviluppo della protesi con l’obiettivo di valutare e validare la forma dell’invaso sotto diverse condizioni di carico (statiche o dinamiche), prima della fase di manifattura. La fase di test del nuovo processo ha inoltre dimostrato la fattibilità del nuovo approccio virtuale per la progettazione delle protesi per arti inferiori. I test effettuati hanno indicato quali miglioramenti siano necessari ed i possibili sviluppi futuri, tra cui: la definizione di un protocollo di acquisizione dell’arto residuo attraverso MRI o scansione 3D, il calcolo parallelo per migliorare le prestazioni della simulazione, l’utilizzo di solutori open-source per implementare una piattaforma di progettazione totalmente indipendente dai sistemi commerciali, la realizzazione di una massiccia campagna sperimentale che coinvolga pazienti transtibiali e transfemorali al fine di convalidare pienamente il modello FE e la piattaforma di progettazione.

The sinking of dense shelf waters down the continental slope (or “cascading”) contributes to oceanic water mass formation and carbon cycling. Cascading is therefore of significant importance for the global overturning circulation and thus climate. The occurrence of cascades is highly intermittent in space and time and observations of the process itself (rather than its outcomes) are scarce. Global climate models do not typically resolve cascading owing to numerical challenges concerning turbulence, mixing and faithful representation of bottom boundary layer dynamics. This work was motivated by the need to improve the representation of cascading in numerical ocean circulation models. Typical 3-D hydrostatic ocean circulation models are employed in a series of numerical experiments to investigate the process of dense water cascading in both idealised and realistic model setups. Cascading on steep bottom topography is modelled using POLCOMS, a 3-D ocean circulation model using a terrain-following s-coordinate system. The model setup is based on a laboratory experiment of a continuous dense water flow from a central source on a conical slope in a rotating tank. The descent of the dense flow as characterised by the length of the plume as a function of time is studied for a range of parameters, such as density difference, speed of rotation, flow rate and (in the model) diffusivity and viscosity. Very good agreement between the model and the laboratory results is shown in dimensional and non-dimensional variables. It is confirmed that a hydrostatic model is capable of reproducing the essential physics of cascading on a very steep slope if the model correctly resolves velocity veering in the bottom boundary layer. Experiments changing the height of the bottom Ekman layer (by changing viscosity) and modifying the plume from a 2-layer system to a stratified regime (by enhancing diapycnal diffusion) confirm previous theories, demonstrate their limitations and offer new insights into the dynamics of cascading outside of the controlled laboratory conditions. In further numerical experiments, the idealised geometry of the conical slope is retained but up-scaled to oceanic dimensions. The NEMO-SHELF model is used to study the fate of a dense water plume of similar properties to the overflow of brine-enriched shelf waters from the Storfjorden in Svalbard. The overflow plume, resulting from sea ice formation in the Storfjorden polynya, cascades into the ambient stratification resembling the predominant water masses of Fram Strait. At intermediate depths between 200-500m the plume encounters a layer of warm, saline AtlanticWater. In some years the plume ‘pierces’ the Atlantic Layer and sinks into the deep Fram Strait while in other years it remains ‘arrested’ at Atlantic Layer depths. It has been unclear what parameters control whether the plume pierces the Atlantic Layer or not. In a series of experiments we vary the salinity ‘S’ and the flow rate ‘Q’ of the simulated Storfjorden overflow to investigate both strong and weak cascading conditions. Results show that the cascading regime (piercing, arrested or ‘shaving’ - an intermediate case) can be predicted from the initial values of S and Q. In those model experiments where the initial density of the overflow water is considerably greater than of the deepest ambient water mass we find that a cascade with high initial S does not necessarily reach the bottom if Q is low. Conversely, cascades with an initial density just slightly higher than the deepest ambient layer may flow to the bottom if the flow rate Q is high. A functional relationship between S/Q and the final depth level of plume waters is explained by the flux of potential energy (arising from the introduction of dense water at shallow depth) which, in our idealised setting, represents the only energy source for downslope descent and mixing. Lastly, the influence of tides on the propagation of a dense water plume is investigated using a regional NEMO-SHELF model with realistic bathymetry, atmospheric forcing, open boundary conditions and tides. The model has 3 km horizontal resolution and 50 vertical levels in the sh-coordinate system which is specially designed to resolve bottom boundary layer processes. Tidal effects are isolated by comparing results from model runs with and without tides. A hotspot of tidally-induced horizontal diffusion leading to the lateral dispersion of the plume is identified at the southernmost headland of Spitsbergen which is in close proximity to the plume path. As a result the lighter fractions in the diluted upper layer of the plume are drawn into the shallow coastal current that carries Storfjorden water onto the Western Svalbard Shelf, while the dense bottom layer continues to sink down the slope. This bifurcation of the plume into a diluted shelf branch and a dense downslope branch is enhanced by tidally-induced shear dispersion at the headland. Tidal effects at the headland are shown to cause a net reduction in the downslope flux of Storfjorden water into deep Fram Strait. This finding contrasts previous results from observations of a dense plume on a different shelf without abrupt topography. The dispersive mechanism which is induced by the tides is identified as a mechanism by which tides may cause a relative reduction in downslope transport, thus adding to existing understanding of tidal effects on dense water overflows.