Industrial and mining sectors are amongst the largest single energy consumers in South Africa, making them a primary focus for implementing energy saving initiatives. Refrigeration systems on mines are responsible for consuming up to25 % of the electrical energy consumption on a typical South African deep level mine. Ample opportunities to reduce the energy consumption of these systems exists, as many of the current systems rely on old technology and function under partial or inadequate control management. In compiling this thesis, various energy saving strategies on deep level mines were investigated. In specific, the effects of controlling and improving the cooling auxiliaries. Scenarios were investigated and simulated, where after an optimum solution was implemented. Implementations, such as the ones covered in this dissertation, form part of the IDM (Integrated Demand Management) energy efficiency incentive introduced by Eskom, where funding is made available based on actual power saving; ensuring that the projects will be financially viable to the clients. Reduced electrical energy consumption realised from the abovementioned projects were measured, captured and compared to the consumption before project implementation to determine the achieved savings. Savings of up to 30 % of the plant installed capacity were realised, providing average savings of up to 2.3 MW per day. / Thesis (MIng (Mechanical Engineering))--North-West University, Potchefstroom Campus, 2013

Industrial and mining sectors are amongst the largest single energy consumers in South Africa, making them a primary focus for implementing energy saving initiatives. Refrigeration systems on mines are responsible for consuming up to25 % of the electrical energy consumption on a typical South African deep level mine. Ample opportunities to reduce the energy consumption of these systems exists, as many of the current systems rely on old technology and function under partial or inadequate control management. In compiling this thesis, various energy saving strategies on deep level mines were investigated. In specific, the effects of controlling and improving the cooling auxiliaries. Scenarios were investigated and simulated, where after an optimum solution was implemented. Implementations, such as the ones covered in this dissertation, form part of the IDM (Integrated Demand Management) energy efficiency incentive introduced by Eskom, where funding is made available based on actual power saving; ensuring that the projects will be financially viable to the clients. Reduced electrical energy consumption realised from the abovementioned projects were measured, captured and compared to the consumption before project implementation to determine the achieved savings. Savings of up to 30 % of the plant installed capacity were realised, providing average savings of up to 2.3 MW per day. / Thesis (MIng (Mechanical Engineering))--North-West University, Potchefstroom Campus, 2013

Computational simulations of air glow discharge plasma in the presence of supersonic flow are presented. The glow discharge model is based on a self-consistent, multi-species, continuum description of the plasma with finite-rate chemistry effects. The glow discharge model is coupled to a compressible Navier-Stokes solver to study the effect of the plasma on the flow and the counter-effect of the flow on the plasma. A finite-rate air chemistry model is presented and validated against experiments from the literature at a pressure of 600 mTorr. Computational results are compared with experimentally measured V-I characteristics, axial positive ion densities and electron temperature, and reasonably good qualitative and quantitative agreement is observed. The validated air plasma model is then used to study the effect of the surface plasma discharge on M=3 supersonic flow at freestream pressure 18 Torr and the corresponding effects of the flow on the discharge structure in two dimensions. The species concentrations and the gas temperature are examined in the absence and presence of bulk supersonic flow. The peak gas temperature from the computations is found to be 1180 K with the surface plasma alone in the absence of flow, and 830 K in the presence of supersonic flow. Results indicate that O- ions can have comparable densities to electrons in the pressure range 1-20 Torr, and that O2- ion densities are at least two orders of magnitude smaller over the pressure range considered. Different ion species are found to be dominant in the absence and presence of supersonic flow, highlighting the importance of including finite-rate chemistry effects in discharge models for understanding plasma actuator physical phenomena. Electrode polarity effects are investigated, and the cathode upstream actuation is found to be stronger than the actuation strength with the cathode downstream, which is consistent with experimental findings of several groups. A parallel computing implementation of the plasma and flow simulation tools has been developed and is used to study the three-dimensional plasma actuator configuration with circular pin electrodes. / text

The effects of active flow control by oscillatory blowing at the leading edge of a nonslender delta wing with a Λ=50° sweep angle have been investigated. Pressure measurements and Particle Image Velocimetry measurements were conducted on a half wing to investigate the formation of leading edge vortices for oscillatory blowing, compared to the stalled flow for the no blowing case. Stall has been delayed by up to 8, and significant increases in the upper surface suction force have been observed. Velocity measurements show that shear layer reattachment is promoted with forcing, and a vortex flow pattern develops. The time averaged location of the centre of the vortical region moves outboard with increased excitation. The near-surface flow pattern obtained from the PIV measurements shows reattachment in the forward part of the wing. There is no measurable jet-like axial flow in the vortex core, which seems to break down at or very near the apex. This highlights that unlike slender delta wings, vortex breakdown is not a limiting factor in the generation of lift for nonslender delta wings. Phase averaged measurements reveal the perturbation due to the pulsed blowing, its interaction with the shear layer and vortex, apparent displacement of the vortex core, and relaxation of the reattachment region. The flow in a phase averaged sense is highly three dimensional. Experiments indicate that unsteady blowing at Strouhal numbers in the region of St=0.5 to St=0.75, and in the region of St=1.25 to St=1.5 can be a highly effective. Reattached flow can develop from stalled flow after pulsing has been initiated with a time constant of tU/c=5 for unsteady blowing at St=0.75, and tU/c=7 for St=1.5. Experiments with excitation from finite span slots located in the forward half of the wing show that partial blowing can be more effective at low momentum coefficients. Force measurements of a full delta wing confirmed that the effectiveness of this method of flow control was not only confined to half delta wings.

The work presented in this thesis concerns the application of concepts that are widely used in metamaterial research to the control of fluid flow. In particular surface structuring and resonance were investigated. The initial work focussed on Stokes flow over structured surfaces. The effective boundary conditions that the structuring creates, analogous to the impedance boundary condition encountered in electromagnetism and acoustics, were examined. Exact solutions for the flow and slip length along the grooves of a family of surfaces were derived. These were compared to Finite Element Method (FEM) models and previous work valid for arbitrary structured surfaces, which was based on a perturbation expansion. Good agreement was found for all available surfaces. The previously presented solution was then also compared to results for a sinusoidal surface, finding good agreement for low aspect ratios but diverging at intermediate aspect ratios. Extending the perturbation theory beyond first order was found to improve the agreement. To explore the concept of resonance in fluid dynamics laminar flow around a circular bluff body with an attached flexible tail was considered, investigating how the resonant behaviour of the elastic tail modified the drag and vortex shedding frequency of the body. The results were compared against the no tail case as well as a rigid tail. For short tail lengths the average drag was reduced compared to both reference cases, whilst the vortex shedding could be either enhanced or reduced. When one of the resonant frequencies of the tail matched the vortex shedding frequency of the body, the resonance motion of the tail resulted in in sharp changes to both the drag and vortex shedding frequency. In the finally section of the thesis I describe the Particle Image Velocimetry experiments that were set up to verify the resonant flexible tail behaviour. The process by which the initial set up was upgraded is given. Results are shown for a circular bluff body being towed through the fluid. This is then extended to a circular bluff body with an attached rigid tail. Preliminary results for the flexible tail case are then presented.

The work reflected in this thesis includes a detailed study of co-flow jet (CFJ) technologies as they are applied to a typical thin airfoil, NACA 6415, at take-off and landing speeds. Numerical analysis and experimental testing were conducted on baseline and co-flow jet airfoils of the same plan form. The CFJ mechanism employs high pressure air injected along the span at the leading edge while a low pressure source removes the same amount of air along the span at the trailing edge. Hence, the net mass flux of the system is zero energy loss is minimized. The jet produced along the upper surface of the airfoil mixes with and excites the free stream flow resulting in increased lift, augmented stall margin, and decreased drag. At certain angles of attack the decreased drag is negative and thrust is produced. The research was comprised of four phases including computational fluid dynamics (CFD) simulations, design and manufacturing of a transformable baseline and adjustable slot size CFJ airfoil, implementation of a CFJ Wind Tunnel Laboratory, and wind tunnel testing. A computational fluid dynamics code, developed at the University of Miami, was used to study flow fields and to obtain analytical results of aerodynamic properties for the baseline and CFJ airfoils. Modeling of both wing shapes utilized the baseline ordinates of a cambered NACA 6415 airfoil. The free stream steady state flow was set to Mach=0.1 to simulate take-off and landing speeds where the co-flow jet mechanism would demonstrate its largest increase in performance. CFD simulations of both models provided aerodynamic coefficients as well as mass flow and jet effect data specifically useful to the CFJ airfoil. The NACA 6415 model used for wind tunnel testing was designed and produced to provide both baseline and CFJ results with adjustable injection and suction slot sizes. Connections for a side-mounted force balance and an air delivery system for the co-flow jet were included in the airfoil model. The design and manufacturing of a wind tunnel test section extension was necessary to provide support for the additional aerodynamic loads induced by the CFJ airfoil and to house various air connections and test sensors. A CFJ Wind Tunnel Laboratory was designed and constructed during the course of the research and included selection of proper air delivery apparatus for the injection and suction air for the CFJ jet. All testing controls and sensor equipment were acquired and installed to obtain various data needed for experimental analysis. Finally, a data acquisition system was designed to consolidate all testing information for ease of use. Wind tunnel testing of the baseline and CFJ airfoils provided the aerodynamic loads and coefficients needed to demonstrate the performance enhancements of the co-flow jet flow control method. Experimental and numerical results were examined to understand the benefits of the co-flow jet as it compares to a similar baseline airfoil. The CFD simulations and experimental measurements agree fairly well. All results indicate that the CFJ flow control method is very effective for a typical thin airfoil with 15% maximum thickness.

Microfluidics refers to devices and methods for controlling and manipulating fluid flows at length scales less than a millimeter. Miniaturization of a laboratory to a small device, usually termed as lab-on-a-chip, is an advanced technology that integrates a microfluidic system including channels, mixers, reservoirs, pumps and valves on a micro scale chip and can manipulate very small sample volumes of fluids. While several flow control concepts for microfluidic devices have been developed to date, here flow control concepts based on thermally responsive polymer solutions are presented. In particular, flow control concepts base on the thermally triggered reversible phase change of aqueous solutions of the polymer Pluronic will be discussed. Selective heating of small regions of microfluidic channels, which leads to localized gel formation in these channels and reversible channel blockage, will be used to control a membrane valve that controls flow in a separate channel. This new technology will allow generating inexpensive portable bioanalysis tools where microvalve actuation occurs simply through heaters at a constant pressure source without a need for large external pressure control systems as is currently the case. Furthermore, a concept for controlled cross-channel transport of particles and potentially cells is presented that relies on the continuous regeneration of a gel wall at the diffusive interface of two co-streaming fluids in a microfluidic channel.

No description available.

Droplet based microfluidics is a developing technology with great potential towards improving large scale combinatorial studies that require high throughput and accurate metering of reagents. Each droplet can be thought of as a miniature microreactor where complex reactions can be performed on the micro-scale by mixing, splitting and combining droplets. This thesis investigates the operation and control of droplet microfluidic devices operating using constant pressure sources to pump fluids where feedback from the droplets influences the overall performance of the device. For this purpose, a model system consisting of a single T-junction droplet generator and a single network node is used to understand how pressure source control effects droplet generation and transport through microfluidic networks. The first part of this thesis focuses on the generation of Newtonian liquid-liquid droplets from a microfluidic T-junction operating within the squeezing-to-transition regime with stable flow rates. An experimental study was performed to characterize the effects of geometry (height/width ratio, channel width ratio) and flow parameters (Capillary number, flow rate ratio, viscosity ratio) on the droplet size, spacing and rate of production. Three stages of droplet formation were identified (lag, filling and necking), including the newly defined lag stage that appears at the beginning of the formation cycle once the interface pulls back after a droplet detaches. Based on the experimental observations, a model was developed to describe the formation process which incorporates a detailed geometric description of the drop shape with a force balance in the filling stage and a control volume analysis of the necking stage. The model matches well with the experimental results as data falls within 10% of the predicted values. Subsequently, the effect of surfactants on the formation process was investigated. Surfactant transport occurs on a timescale comparable to the production rate of droplets resulting in dynamic interfacial tension effects. This causes strong coupling between the mass transport of surfactants and the drop production process. Using the previously defined force balance, the apparent interfacial tension at the end of the filling stage was measured. The results show that there is a significant deviation from the equilibrium interfacial tension at normal operating conditions for the T-junction generators due to the rapid expansion of the interface. A model was developed to calculate the dynamic interfacial tension for pre and post micellar solutions, which was then incorporated into the overall model for droplet formation in T-junction generators. Next, the behaviour of microfluidic droplet generators operating under pressure source control was investigated. Coupling between the changing interface and hydrodynamic resistance of the droplets and the flow rate of the two phases creates fluctuations in the output of the droplet generator. Oscillations were found to occur over the short term (one droplet formation cycle) and long term (many formation cycles). Two metrics were developed to quantify these oscillations. Short term oscillations were quantified by tracking droplet speed in the output channel and long term oscillation were quantified by measuring changes in droplet spacing. Analysis of experimental and numerical data shows that long term oscillations have a periodicity that matches the residence time of droplets in the system. A simple model is developed to determine the influence of Laplace pressure, droplet resistance, T-junction generator design and network architecture on the magnitude of these oscillations. From the model a set of design rules are developed to improve the overall operation of T-junction generators using pressure driven flow. The final part of this thesis studies the transport of droplets through a single microchannel junction under various geometric and flow conditions applied to the two outlet channels. Droplets alter the hydrodynamic resistance of the channel they travel within which creates a feedback effect where the decision of preceding droplets influences the trajectory of subsequent droplets. Multifaceted behaviour occurs where sometimes the trajectory of droplets follows a repeatable pattern and other times it is chaotic. As part of this work, a discrete analytical model was developed that predicts droplet transport through the junction including transitions between filtering and sorting, bifurcation in the patterns, composition of the patterns, and an estimation of when patterns will disintegrate into chaos. The model was validated by comparing it to compact numerical simulations and microfluidic experiments with good agreement.The complex behaviour of a simple junction emphasizes the challenge that remains for more highly integrated droplet microfluidic networks operating with pressure driven flow.

One of the fastest growing segments of Internet applications are real-time mul- timedia applications, like Voice over Internet Protocol (VoIP). Real-time multimedia applications use the User Datagram Protocol (UDP) as the transport protocol because of the inherent conservative nature of the congestion avoidance schemes of Transmis- sion Control Protocol (TCP). The e®ects of uncontrolled °ows on the Internet have not yet been felt because UDP tra±c frequently constitutes only » 20% of the total Internet tra±c. It is pertinent that real-time multimedia applications become better citizens of the Internet, while at the same time deliver acceptable Quality of Service (QoS). Traditionally, packet losses and the increase in the end-to-end delay experienced by some of the packets characterizes congestion in the network. These two signals have been used to develop most known °ow control schemes. The current research considers the °ow accumulation in the network as the signal for use in °ow control. The most signi¯cant contribution of the current research is to propose novel end- to-end °ow control schemes for unicast real-time multimedia °ows transmitting over best-e®ort networks. These control schemes are based on predictive control of the accumulation signal. The end-to-end control schemes available in the literature are based on reactive control that do not take into account the feedback delay existing between the sender and the receiver nor the forward delay in the °ow dynamics. The performance of the proposed control schemes has been evaluated using the ns-2 simulation environment. The research concludes that active control of hard real- time °ows delivers the same or somewhat better QoS as High Bit Rate (HBR, no control), but with a lower average bit rate. Consequently, it helps reduce bandwidth use of controlled real-time °ows by anywhere between 31:43% to 43:96%. Proposed reactive control schemes deliver good QoS. However, they do not scale up as well as the predictive control schemes. Proposed predictive control schemes are e®ective in delivering good quality QoS while using up less bandwidth than even the reactive con- trol schemes. They scale up well as more real-time multimedia °ows start employing them.