No description available.

This thesis investigates the physics of statistically unsteady axisymmetric turbulent jets and plumes using theory and direct numerical simulation. The focus is on understanding and modelling the physics that govern the behaviour of radially integrated quantities, such as the integral scalar flux, momentum flux and buoyancy flux. To this end, a framework is developed that generalises previous approaches, making no assumption about the longitudinal velocity profile, turbulence transport or pressure. The framework is used to develop well-posed integral models that exhibit a good agreement with simulation data. In the case of passive scalar transport, shear-flow dispersion is observed to be dominant in comparison with longitudinal turbulent mixing. A dispersion closure for free-shear flows based on the classical work of Taylor (Proc. R. Soc. Lond. A, vol. 219 1954b, pp. 186-203) is therefore developed. In the analysis of jets whose source momentum flux undergoes an instantaneous step change, it is demonstrated that a momentum-energy framework, of the kind used by Priestley & Ball (Q. J. R. Meteorol. Soc., vol. 81 1955, pp. 144-157), is the natural choice for unsteady free-shear flows. The framework is used to demonstrate why existing top-hat models of unsteady jets and plumes are ill-posed and that jets and plumes with Gaussian velocity profiles remain approximately straight-sided and are insensitive to source perturbations. Contrary to the view that the unsteady jet and plume equations are parabolic, it is shown that the generalised system of equations is hyperbolic. In unsteady plumes, the relative orientation of three independent families of characteristic curves determines whether propagating waves are lazy, forced or pure. To relate findings that are based on the momentum-energy framework to the classical mass-momentum framework, an unsteady entrainment coefficient is defined that generalises the decomposition proposed by Kaminski et al. (2005, J. Fluid Mech., vol. 526, pp. 361-376).

No description available.

Quantum turbulence is a conceptually simple form of turbulence, consisting of a tangle of quantised vortex lines. It provides a model system, through which it may be possible to understand features of the complex and not yet fully understood classical turbulence. A novel detector made from arrays of custom-designed tuning forks was developed and used to investigate properties of excitation beams and quantum turbulence in superfluid 3He-B at temperatures below 200 µK. The detector was constructed from 5 arrays of 5 tuning forks mounted in a copper block to create a 25 pixel square detector of excitation flux. The detector was situated in a cell such that it could be illuminated with a beam of thermal excitations, and that turbulence could be generated in the path of the beam, which will cast a shadow on the face. Characterisation of the detector response to beams generated by the black-body radiator and source wire were performed. We observe that the beam generated by a black-body radiator appears approximately symmetric, consistent with being emitted from a point-source of excitations. In addition to this we find that the profile of the beam generated by the black-body radiator was independent of the power applied to it, an important assumption of the properties of the black-body radiator that had not been previously observed. The beam emitted from a vibrating wire was found to be much narrower and has a angular profile that changes as the velocity of the wire is increased. Probing a turbulent tangle generated by a vibrating wire with this beam showed that the turbulence appears to fill all of the volume in between the radiator and detector. The vortex line density of the tangle appeared the be greatest in the vicinity of the wire, consistent with previous measurements of the vortex tangle generated by vibrating wires. In addition, we find that there are reproducible features in the development of the shadow as a function of the wire velocity, indicating that there is some structure in the development. The shadow is independent of the power of the beam used to probe the tangle. Measurements of fluctuations in the shadow cast by the vortex tangle show that the turbulence has a spectrum reminiscent of the Kolmogorov spectrum. In addition to this the resolution of the detector is such that it is possible in principle to measure the shadow cast by a single vortex line, and we examine candidate events for such a measurement. This represents the first such measurement of the motion of vortex lines in 3He-B.

Packed beds of gas-solid systems are extensively used as reactors, separators, dryers, filters, heat exchangers and combustors. The design of packed beds requires a detailed knowledge of local dynamics of flow, composition and temperature. Unfortunately, investigations for the development of 3D modelling codes by computational fluid dynamics are still not sufficiently mature compared with those relying on 2D modelling or simplified pseudo-homogenous models. This project investigates non-uniform catalyst deactivation in packed bed reactors of low aspect ratios under steady-state and dynamic operations. Low aspect ratio packed beds were selected as they are known to generate non-uniform distribution of local flow. Detailed knowledge of flow dynamics in terms of local structure of the packed bed, pressure drops, interstitial flow, heat and mass rate distributions was examined. The discrete element method was used to generate various packing configurations and the results of profiles of porosity were in a good agreement with the semi-analytical models, especially, in the vicinity of the wall. Similar oscillation trends with damping profiles towards the centre of the packed beds were observed. Flow heterogeneity was assessed by tests of mass transfer dispersion through a Lagrangian approach. Interactions of fluid flow, mass and heat transfers, and local deactivation of alumina catalyst Al2O3 of CO oxidation were investigated under design and operating conditions. An increase in the activation energy of deactivation promoted the deactivation by accelerating the reaction rate and releasing additional thermal energy, which in turn accelerated the deactivation. The 3D modelling allowed observation of local catalyst deactivation at packing pore level which is typically not accessible by the 2D modelling or pseudo-homogeneous models. In addition, the deactivation was quite asymmetrical along axial and radial directions, leading to uneven rates of thermal expansion and contraction and causing local deactivation associated with temperature runaways.

A buoyancy-assisted microfluidic approach is introduced for facile production and collection of uniform drops within a wide range of sizes, particularly on a millimetre scale, which is not easily achievable via conventional microfluidic approach. The proposed methodology, characterised by vertical orientation and non-confined quiescent outer phase of the device used, was also applied to droplet-in-droplet and droplet-in-fibre encapsulation using a co-axial glass microcapillary arrangement, to obtain millimetric capsules and multi-compartmental fibres. The shell thickness of double emulsions was tuned, via altering flow rates and formulations, to produce millimetric ultrathin shelled capsules. Alginate fibres with different oil-encapsulate geometries were fabricated, via simultaneous oil-droplet formation and encapsulation, and characterised and analysed for their encapsulation volume, surface roughness, spillage ratio and mechanical strength. Furthermore, the size and locations of oil encapsulates were manipulated to obtain asymmetric fibres with parallel oil streams. An asymmetric encapsulation approach was designed and used to fabricate dehydration-responsive fibres, which demonstrated a benign and facile dehydration-triggered core-release mechanism. This core-release response was also demonstrated for fibres with parallel oil-encapsulates with multiple cargos. The fibre morphology was also tuned to provide an enhanced response to its mechanical failure, marked by a simultaneous release of potentially reactive components at the point of fracture. Such fibres, can behave as fibres with self-repairing properties. The buoyancy-assisted microfluidics was also used to produce microfibres containing gas encapsulates with tuneable morphology. The buoyancy force, driven by the trapped microbubbles, was utilised for stretching the gelling alginate fibres to fabricate ultrathin alginate microfibres, a feature not possible via conventional horizontally-oriented microfluidic techniques. The collected bubble-filled fibres were also morphed to produce new varieties of fibres, such as beaded fibres and fibres with segmented aqueous cores.

The production of microbubbles is rapidly becoming of considerable global importance with many industries taking advantage of the increased mass transfer rates the bubbles can attain. Many factors have interrelated roles during bubble formation, with effects such as gas flow rate, liquid viscosity, pore size and pore orientation all imparting considerable influence during the formation process. Many of these features have been examined in detail and are relatively well understood. However, the role of surface wettability and the interactions at the gas-liquid-solid triple interface have for the most part been neglected, and it is the role of this wettability that is examined herein. Utilising the well-studied wet chemistry surface modification techniques of silanes and thiols, many substrates have been modified and the wettability tested. Contact angle goniometry has been utilised to assess the wetting characteristics of each substrate, and the role of surface roughness has been discussed in relation to both the static Young’s contact angle and the advancing and receding angles. Modified porous plates have been used to generate bubbles, with controlled single pore, multiple controlled pore, and multiple randomised pore systems being investigated. A steady flow of air was bubbled into distilled water through the various diffuser plates. It has been observed a contact angle of 90° is of vital importance, with a significant increase of bubble size above the 90° angle, defined as the hydrophobic wetting region. On the contrary, bubble size is greatly reduced below this angle, in the region defined as the hydrophilic region. The effect is seen to increase as the density of pores increases when the plate from which they are emitted is relatively smooth. Upon roughening, the effect is seen to diminish, and mechanisms for this process have been postulated. It is thought that the surface topography disrupts the modifying layers and also physically restricts the growing bubble, preventing the growth of the bubbles emitted from a hydrophobic surface. Attempts have been made to support this hypothesis both qualitatively and quantitatively. The fluidic oscillator of Zimmerman and Tesar has been examined, with numerous physical features being investigated. The oscillator was then added to the system to investigate the effect of wettability under substantial oscillation. It has been shown that the bubble size emitted from hydrophobic surfaces is significantly reduced when compared to the steady flow system. The effect is believed to be due to the ‘suction’ component of the oscillatory flow created by the oscillator. It has been seen via high speed photography that the growth rate of the growing bubble slows significantly as the flow begins to switch, before a reduction in size is seen as the gas is removed from the bubble. The opposing forces of buoyancy and suction act to elongate the bubble neck causing break off at a significantly reduced size. Although the diffuser plate is often observed to oscillate like the skin of a drum, this is not the predominant cause of the size reduction. Further experiments have been conducted using a synthetic actuator jet to create a pulsed air flow with only a positive component. Bubble size is not affected in this case, despite frequency sweeps being employed.

No description available.

The purpose of this thesis is to explain the underlying physics of turbulent particle laden flows in the context of Direct Numerical Simulations (DNS). Moreover, modelling tools, that can be used in Large Eddy Simulations (LES), are developed and validated against experimental and DNS data. DNS of homogenous and isotropic turbulence are performed in order to explain the turbulence modulation caused by the particles. In order to carry out this analysis, a novel forcing method is developed that retains the statistics of the fluid. Contrary to other forcing schemes, it is shown that this scheme does not affect the statistics of both phases. The two-way coupling spectrum reveals that the particles transfer energy from the large to the small scales, whereby some of this energy is dissipated by the fluid. Additionally, a model two-way coupling spectrum is proposed that considers the observations of the DNS study. Filtering is performed on the DNS results in order to remove the high frequency velocity components. It is evident from the results that the absence of the subgrid velocity fluctuations have an impact on the particle pair dispersion. Moreover, a novel stochastic model is proposed which reconstructs the subgrid scale turbulence characteristics. The results of the model are very promising as there is good agreement with the DNS data. LES of a turbulent horizontal particle-laden channel flow is performed in order to compare the soft-sphere and hard-sphere models and coupling strategies. A novel roughness model, used in conjunction with the soft-sphere model, is proposed. The results of the soft sphere and hard sphere models are in very good agreement with the available experimental data. Furthermore, the results, also, show that the wall roughness is an important mechanism in keeping the particles distributed across the channel despite the action of gravity.

The efficient numerical approximation of viscous, incompressible flow by families of mixed finite elements is subject to the satisfaction of a stability or inf-sup condition between the velocity and pressure approximation spaces. The present work analyses the stability of mixed hp-finite elements for planar Stokes flow on affine quadrilateral meshes comprising of regular and anisotropic elements. Firstly, a new family of mixed hp-finite elements is presented for regular elements with an inf-sup constant bounded below independently of the mesh size h and the spectral order p. In particular, the element allows continuous piecewise polynomial pressures to be used. Next, the stability of families of mixed hp-finite elements on geometrically refined anisotropic elements is considered using a macro-element technique. New results are presented for edge and corner macro-elements, in particular, for the latter case, the dependence of the inf-sup constant on the geometric refinement parameter is explicitly characterised. The families are then used in the numerical approximation of two physical Navier-Stokes problems.