Research Doctorate - Doctor of Philosophy (PhD) / Determining the operating parameters and design considerations for dense phase (non-suspension) conveying of fine powders in pneumatic systems typically use empirical, steady-state modelling techniques, as the mechanisms of the flow behaviour are still not fully understood. However, this necessary simplification in the modelling of the dense phase flow still presents significant challenges in ensuring that the predicted outcomes adequately reflect the physical nature of the flow, and therefore provide good design guidance. This thesis represents an examination and development of techniques required for designing dense phase systems of fine powders in three specific areas; prediction of a materials potential to dense phase convey, solids friction correlations and their subsequent effect on pressure drop prediction, and modelling the solids flow from a local perspective. The dense phase capability analysis was conducted by reviewing the current predictive techniques utilising known dense phase material data. It was apparent in the thesis that there were distinct strong predictive regions in all the diagrams; however some diagrams showed areas with weak predictive regions. This work also illustrated the difficulties in comparing different de-aeration rate techniques and significantly, a new mode of flow predictive chart was developed which eliminated the need to determine de-aeration rates while still maintaining distinctly strong dense phase predictive capability. Solids friction based pressure models invariably use a power law which require determination of co-efficient/s and exponent/s. Detailed in this thesis is the research which shows why solutions do not always occur in these power law based friction models and defines a method of determining stable and meaningful values for the exponents. Furthermore, a generic air/particle parameter based solids friction model was developed which is a clear advancement in defining the frictional resistance of dense phase pneumatic conveying of powder. This thesis also proposes a new continuum model which calculates the force balance between the conveying air flow, the resistance of the particles and geometrical effects, like bends. The solution to this model provides qualitative information on fine powder dense phase flow velocity from a solids flow perspective and represents a major step in advancing dense phase modelling from a particle flow basis.

Research Doctorate - Doctor of Philosophy (PhD) / Determining the operating parameters and design considerations for dense phase (non-suspension) conveying of fine powders in pneumatic systems typically use empirical, steady-state modelling techniques, as the mechanisms of the flow behaviour are still not fully understood. However, this necessary simplification in the modelling of the dense phase flow still presents significant challenges in ensuring that the predicted outcomes adequately reflect the physical nature of the flow, and therefore provide good design guidance. This thesis represents an examination and development of techniques required for designing dense phase systems of fine powders in three specific areas; prediction of a materials potential to dense phase convey, solids friction correlations and their subsequent effect on pressure drop prediction, and modelling the solids flow from a local perspective. The dense phase capability analysis was conducted by reviewing the current predictive techniques utilising known dense phase material data. It was apparent in the thesis that there were distinct strong predictive regions in all the diagrams; however some diagrams showed areas with weak predictive regions. This work also illustrated the difficulties in comparing different de-aeration rate techniques and significantly, a new mode of flow predictive chart was developed which eliminated the need to determine de-aeration rates while still maintaining distinctly strong dense phase predictive capability. Solids friction based pressure models invariably use a power law which require determination of co-efficient/s and exponent/s. Detailed in this thesis is the research which shows why solutions do not always occur in these power law based friction models and defines a method of determining stable and meaningful values for the exponents. Furthermore, a generic air/particle parameter based solids friction model was developed which is a clear advancement in defining the frictional resistance of dense phase pneumatic conveying of powder. This thesis also proposes a new continuum model which calculates the force balance between the conveying air flow, the resistance of the particles and geometrical effects, like bends. The solution to this model provides qualitative information on fine powder dense phase flow velocity from a solids flow perspective and represents a major step in advancing dense phase modelling from a particle flow basis.

Thesis (Ph.D.), University of Adelaide, Dept. of Botany, 1979.

The X-ray emission from a plasma focus has been studied using time integrated and streak photography. The plasma focus, a small volume of very dense and hot plasma was created in a coaxial plasma gun driven by a fast current pulse of period T ~ 2 μsec which was produced by discharging a condensor bank of V = 12-15 kV, and C = 84 μf. Measurements have shown that a diffuse X-ray emitting plasma column is formed in the 'early' pinch stage, which extends a few centimeters in the axial direction, has expansion velocities of between 2-6 x 10⁷ cm/sec. and lasts for 30-60 nsec. In the following 10-30 nsec. X-ray emission occurs from small plasma regions which have little or no axial velocity. The distance between these "hot" spots are of the order of half a centimeter. These measurements and observations of the X-ray emitting regions are consistent with results obtained by Peacock and Mather. By comparing their results with those of this experiment it is concluded that the appearance of the isolated X-ray sources is associated with the m = 0 instability. / Science, Faculty of / Physics and Astronomy, Department of / Graduate

Desalination plants are becoming essential due to the limited water resources in order to reduce the pressure of high demand of freshwater in many countries in recent decades. A concerning problem associated with desalination plants is the high concentration brine which has high risk to marine environments. Inclined dense jets are commonly used to treat brine produced by desalination plants or in industrial outfall discharges. They are produced when the brine is discharged at an upwardly inclined angle through a pipe or a diffuser system. Previous studies have mainly focused on jets on a horizontal bottom. In the present study, the influence of sloped bottom is investigated by numerical simulations using a modified solver in OpenFOAM (pisoFoam). Four different Reynolds Averaged Navier Stokes (RANS) turbulence models (Realizable k-ε, Standard k-ε, RNG k-ε and Nonlinear k-ε were employed to assess the accuracy of the selected turbulence models in predicting the jet behavior. Jets of inclination angle of 30° with four different initial conditions (Froude number=15, 20, 25, 30) on three different bed slope angles (2°, 5°, 10°) in stagnant water were conducted. Although inclined dense jets of the discharge angel of 60° are more common in discharge systems, sometimes they cannot be used in shallow waters in order to prevent surface pollution. In such cases, a relatively small jet inclining angle can be used to prevent the surface pollution and as shown in this thesis, bed slope can enhance the brine mixing and dilution. The results showed that Realizable k-ε model is more accurate among the turbulence models studied herein. The dilution at the impact point can be estimated based on the Froude number and initial conditions. After the impact point, the slope did enhance the dilution of the plume compared to the horizontal bed. The dilution was thus affected by the slope and the dilution after the impact point on the slope appeared to be linearly related to the distance to the source. Besides, the slope could enhance the jet dilution up to 20% compared with the horizontal bed after the impact point.

An analysis of waveforms, power spectral density and array responses was performed using geophones and broadband seismometers, co-deployed as part of a geologically motivated study. Broadband seismometers record excellent waveforms but, due to cost and deployment effort, wavefields are usually spatially aliased above ~0.1 Hz. Industry rapidly deploys many thousands of inexpensive, passive geophones to record full, unaliased seismic wavefields; however, waveform quality is limited below the instrument's natural frequency of ≥2 Hz. In 2012, coincident passive and controlled-source seismic surveys were deployed to investigate tectonics in Idaho and Oregon. Broadband stations were deployed at quiet sites every 15 km, taking experienced professionals >1 person-days per station. Fifty 4.5 Hz geophones and "Texan" seismographs at 200-m spacing were deployed per person-day by inexperienced students. Geophone data were continuously recorded for 3 nights and 1 day, while broadband seismometers were deployed for ~2 years. The spectral and array responses of these real deployments were compared. For a M7.7 teleseismic event, the broadband seismometer and geophone recorded nearly identical waveforms down to <0.03 Hz (32 s) and matching power spectral density down to 0.02 Hz (50 s). For quiet ambient noise, the waveforms strongly correlate down to <0.25 Hz (4 s) and the power spectral density match to the low-frequency side of the microseismic peak at ~0.15 Hz (~7 s). By deploying a much larger number of geophones, waveforms can be stacked to reduce instrument self-noise and beamforming can be used to identify wavefield azimuth and apparent velocity. Geophones can be an effective tool in ambient noise seismology down to ~7 seconds and can be used to record large seismic events effectively down to tens of seconds, well below the natural frequency of the instruments. A well-designed deployment of broadbands and geophones can enable full wavefield studies from long period to short period. Scientific and societal applications that could benefit from the improved unaliased wavefield bandwidth include local to regional seismicity, strong ground motion, magma migration, nuclear source discrimination, and crustal studies. / Master of Science / An analysis of seismic responses was performed using common seismology sensors, codeployed as part of a geologically motivated study. Broadband seismometers record seismic activity extremely well, however, due to cost and deployment effort, are less effective above ~0.1 Hz. Industry rapidly deploys many thousands of inexpensive, geophones, to record effectively above ~2 Hz; however, quality of the signal is limited below 2 Hz. In 2012, coincident seismic surveys were deployed to investigate earth structures in Idaho and Oregon. Broadband stations were deployed at every 15 km, taking experienced professionals >1 person-days per station. Fifty geophones and “Texan” seismographs at 200-m spacing were deployed per person-day by inexperienced students. Geophone data were continuously recorded for 3 nights and 1 day, while broadband seismometers were deployed for ~2 years. The seismic responses of these real deployments were compared. For a M7.7 earthquake, the broadband seismometer and geophone recorded nearly identical waveforms down to <0.03 Hz (32 s) and had similar characteristics down to 0.02 Hz (50 s). For low energy seismic signal, the waveforms were comparable down to <0.25 Hz (4 s) and had similar characteristics at ~0.15 Hz (~7 s). By deploying a much larger number of geophones, waveforms can be added together to improve signal quality and determine where the seismic source is located. Geophones can be an effective tool for low energy seismic signal down to ~7 seconds in period and can be used to record large seismic events effectively down to tens of seconds in period. A well-designed deployment of broadbands and geophones can enable full seismic studies from low and high frequencies which would have many scientific and societal benefits.

L'innovation détermine en grande partie la compétitivité des entreprises car elle facilite leur distinction dans l'économie du savoir et qu'elle induit une flexibilité organisationnelle face un environnement turbulent. Selon une perspective évolutionniste, toutefois, les organisations poursuivent avec le temps une trajectoire technologique de plus en plus étroite car les routines qui sous-tendent la création et l'accumulation de connaissances à la base de l'innovation rigidifient les organisations. L'exploration technologique des organisations devient plus locale, plus proche de ce qu'elles connaissent déjà. Dans un processus de vieillissement, les organisations perdent graduellement leur flexibilité concurrentielle. Le problème est que l'innovation a besoin d'être alimentée en nouvelles connaissances, et donc, en information riche pour poursuivre sa contribution à la flexibilité des organisations. Le réseau dense, à cause de la densité des interactions entre partenaires, représente une source particulièrement féconde d'information riche pour innover. Nombre de travaux identifient d'ailleurs le potentiel du réseau dense pour innover dans un milieu turbulent. Le comment demeure néanmoins peu connu. Cette recherche vise à mieux comprendre comment le réseau dense alimente l'innovation en information riche. L'étude longitudinale d'un cas unique a servi à répondre aux questions spécifiques sur les processus d'apprentissage, d'absorption de l'information et sur des conditions de faisabilité de l'innovation en réseau dense, ou codéveloppement. L'unité d'analyse, constituée de cinq partenaires en codéveloppement, a représenté un contexte très complexe qui a justifié une étude empirique en profondeur. L'observation des chercheurs à quatre réunions de travail, la tenue d'un journal de bord, des entrevues semi-dirigées auprès des partenaires et des documents sur la préhistoire, la période active et la période passive du projet jusqu'à son abandon ont servi à la collecte des données. La technique du modèle logique a servi à valider des modèles proposés sur l'apprentissage et l'absorption d'information riche à la base du codéveloppement. Les comportements des partenaires ont aussi été comparés aux comportements considérés nécessaires à la faisabilité du codéveloppement. Malgré l'abandon du projet, les chaînes d'événements observés ont supporté les modèles logiques proposés dans l'explication de l'échec. Premièrement, l'apprentissage repose sur le déploiement d'une infrastructure qui est stabilisée par la coopération, activée par l'interaction et permettant l'appropriation des nouveaux savoirs. Cependant, les nouvelles connaissances produisent des effets rétroactifs sur les conditions d'apprentissage qui, dans le cas étudié, ont eu l'effet de déstabiliser la coopération et de créer un désinvestissement menant à l'échec du partenariat. Deuxièmement, des informations riches ont été absorbées à travers des activités d'interprétation, de création de savoir et de décisions. Les partenaires ont utilisé des signaux faibles et forts mis à leur disposition par des liens faibles et forts du réseau. La proximité socio-technique entre les partenaires a facilité la prise de décision mais en contrepartie, un mode décisionnel centralisé a inhibé les efforts d'action collective. Dans le cas étudié, les comportements ont permis de constater la corrélation entre la faiblesse des conditions de faisabilité et l'échec du codéveloppement. Ces conditions se composaient d'anticipations positives, de l'autonomie des individus, de leur ouverture, de mécanismes de réflexivité, de la légitimité du réseau, d'un leadership transformationnel, de coévôlutions fortes, de cultures alignées, d'un langage commun et de temps approprié pour déployer le codéveloppement. La recherche tient son originalité de l'autopsie d'un échec en codéveloppement à partir d'une étude in tiempo de sa perte. En plus de donner une réponse à l'apport d'information riche au codéveloppement, elle montre l'intérêt des chaînes logiques pour s'adapter aux mouvances d'un contexte complexe. La recherche fournit des ancrages théoriques à la recherche-action ou par systèmes souples. Ces résultats, de généralisation analytique, seront applicables à d'autres situations en considérant la richesse des constats de ce cas longitudinal unique.

This thesis demonstrates an efficient parallel method of solving the generalized eigenvalue problem, KΦ = M ΦΛ, where K is symmetric and M is symmetric positive-definite, by first converting it to a standard eigenvalue problem, solving the standard eigenvalue problem, and back-transforming the results. An abstraction for parallel dense linear algebra is introduced along with a new algorithm for forming A := U⁻ᵀ K U⁻¹ , where U is the Cholesky factor of M , that is up to twice as fast as the ScaLAPACK implementation. Additionally, large improvements over the PBLAS implementations of general matrix-matrix multiplication and triangular solves with many right-hand sides are shown. Significant performance gains are also demonstrated for Cholesky factorizations, and a case is made for using 2D-cyclic distributions with a distribution blocksize of one. / text

A novel approach to study electron thermal transport in dense plasmas was successfully implemented to measure the temperature-dependent conductivity and test the currently available dense plasma model by Lee and More. Intense, femtosecond laser pulses with energy up to 7 mJ per pulse were used to heat free-standing 170-370 nm aluminum foils. We carried-out a new approach to study the plasma transport properties of electron and thermal conduction. In this new approach, rather than probing the front (laser-heated) surface, probing was done on the back surface of a thicker metallic foil heated by a thermal conduction wave generated from a laser-heated front surface. Frequency-domain interferometry with chirped probe pulses allowed us to simultaneously measure the time-dependence of the optical reflectivity and phase-shift in a single shot with subpicosecond resolution. In addition, solid heating was observed to be dominated by the thermal conduction wave prior to the shock-wave breakout at the back surface when laser energy was directly deposited in a thin metallic foil. As a result we were able to estimate the optical conductivity of a dense aluminum plasma in the range of 0.1 – 1.5 eV. The optical parameters were calculated using the output of a hydrodynamic simulation along with the published models of bound electron contributions to the conductivity and were found to be in reasonable agreement with the measurement. We found that the Lee and More model of a dense plasma’s conductivity predicts the real and imaginary part of the measured optical conductivity to within 20%. The simulation results were then used to examine the temperature dependence of the conductivity for 170 and 230 nm aluminum foils heated with the 2-5 mJ pulses. In all cases the same conductivity was obtained, though the arrival of the heat wave and subsequent shock waves varied with the choice of intensity and target thickness. This consistency in the data gave us good confidence in the validity of this technique for deriving conductivity as a function of temperature. / text

No description available.