This thesis reports on an experimental study of air-water mixtures flowing through idealized shell and tube, in-line and staggered heat exchangers. The measured void fractions in the maximum and minimum gaps between the tubes are reported at near atmospheric conditions, to give local variations for different tube diameters and tube bundle arrangements. The void fraction measurements were made using a gamma-ray densitometer. The pressure drops in the tube bundles are also reported. These data are compared with the correlations available in the open literatures to investigate the void fraction and pressure drop prediction methods for these heat exchangers. The in-line 38 mm tube bundle is shown to provide no significant effect on void fraction or drag force when compared with the 20 mm tube diameter bundle. A new void fraction model is therefore proposed by modifying the characteristic length of an existing slip ratio method. A new pressure drop model is presented. The acceleration pressure drop between the tubes from the separation to re-attachment is shown to be responsible for some of the frictional pressure drop with a liquid film on the tubes responsible for the remainder. The staggered bundle shows the bundle arrangement gives different void fraction and different pressure drop data when compared to the in-line bundle.

Open cell metal foams show great potential as a heat exchangers, due to their permeability to fluids and the high conductivity of the metallic network. In this study, aluminium foams were produced using the replication technique with NaCl, flour and water used to create the preform. The samples produced included both uniform pore sizes and examples where different pore sizes were created in different parts of the sample as well as these, samples made commercially by a similar technique (Corevo foams) and by an investment casting process (Duocel foams) were examined. A bespoke rig was designed, built and used to measure the thermal and fluid flow performance of all foams being investigated under forced convection conditions. Results for heat transfer coefficient and pressure drop across the sample with the comparison between each type of sample are presented. It was found that all the foams tested can have favourable heat transfer behaviour under certain conditions asymmetric behaviour can be obtained when non-uniform pore sizes are present; a factor that could be exploited in heat exchanger design.

Recent developments in hybrid rocket technology involve producing a coaxial bidirectional vortex flow field through use of tangential oxidiser injection at the base of the combustion chamber. This is found to significantly increase engine performance by providing enhanced thermal transfer at the fuel surface, resulting in increased fuel regression rates in addition to more efficient combustion. The double helical path of the flow results in reduced reactant loss at the chamber outlet whilst confining combustion to a high temperature core region defined by the inner vortex. This also results in enhanced thermal shielding as a large radial thermal gradient is established, which affords lightweight materials to used in the construction of the combustion chamber making it suitable for weight sensitive applications such as satellite propulsion. Analytical treatment of the coaxial bidirectional vortex has repeatedly found that it is theoretically possible to induce a vortex flow field consisting of multiple concentric vortices, which would further increase the benefits associated with the coaxial bidirectional vortex. Each additional vortex would further increase the length of the helical trajectory of the flow and allow for more compact combustion chambers to be designed which are both lightweight and highly efficient. However, the existence of multiple concentric vortices has yet to be confirmed and the parameters involved in producing such phenomenon are unknown, as are the parameters required to manipulate their behaviour. Therefore, the focus of this study is to investigate whether it is possible to induce multiple concentric vortices and affect their behaviour through parametric variation of geometrical constraints which are found to significantly influence vortex characteristics in similar cyclonic devices. A twofold approach is employed to investigate a range of cyclone chamber configurations both experimentally and numerically, in order to ascertain the effectiveness of each method with regards to resolving complex vortical flows and to observe and similarities in the results obtained. Due to experimental constraints a hydrocyclone was used in place of a gas cyclone which is commonly used to characterise the behaviour of the coaxial bidirectional vortex. The numerical analysis was performed using CFD, where 3D simulations were necessary to adequately resolve the spatio-temporal behaviour of the flow, with specific consideration given to solver settings applicable to the resolution of intense swirling flow. While the experimental analysis was conducted using a 2D time resolved PIV method, this was applied to the meridional and azimuthal planes of the cyclone chamber to enable comparison of the number of concentric vortices and the general characteristics of the vortex. Despite the limitations of the techniques applied it was found that it is possible to produce a confined vortex consisting of multiple concentric vortices, although the resultant flow structures are considerably more complex than initially thought. However, it was also found that the geometry of the cyclone chamber has a significant influence upon the structure of the vortex, with small chamber aspect ratios and large contraction ratios producing intense vortices which are associated with multiple concentric vortices. The position of the tangential inlets is also found to have a significant impact upon the structure of the flow, and several configurations of multiple concentric vortices were observed that are not accounted for by the analytical solutions. Another important result that was consistently observed is that the locus of zero axial velocity which defines the number of concentric vortices takes the form of modal structures, which appear to evolve in response to threshold geometrical parameters. It is thought that the underlying fluid mechanics of multiple concentric vortices are inherently linked to the harmonic characteristics of the cyclone chamber geometry, as several mechanisms have been identified which may allow for direct control over the structure of the vortex.

It is a generally accepted belief that the use of nanofluids enhances heat transfer rates in comparison with a traditional fluid and can be considered to be one of the most important energy conservation measures in many industrial applications. Despite increased interest, detailed and systematic studies of nanofluids’ flow and thermal characteristics are limited and their effect on heat transfer is often misunderstood. The concentration of a nanofluid is often chosen independently of the application conditions, nanofluid type or cost and other economic parameters such as the cost of energy and lifetime of the application. This thesis has three main objectives; the first is the measurement of the thermal properties of nanofluids and the proposal of a correlation model for nanofluid viscosity. The results of these measurements show that the nanofluid viscosity depends on the type of nanoparticles and their concentration and the fluid temperature. It is shown that the viscosity increases with increasing nanoparticle concentration and decreases with increasing temperatures over the nanoparticle concentration and temperature ranges investigated. The second objective is concerned with the measurement and evaluation of heat transfer performance and pressure drop of various nanofluids via an experiment using forced convection heat transfer within the turbulent regime. In general, it is shown that the heat transfer coefficient of nanofluids decreases with increasing nanoparticle concentration at a specific flow rate and the base fluid gives higher heat transfer with respect to the nanofluids. In the other hand, Assessing the thermal performance of nanofluids by considering the Nusselt number and its variation with Reynolds number is misleading because both Nusselt number and Reynolds number depend on the nanofluid properties (i.e. thermal conductivity, density and viscosity that are function of the volume fraction). This can lead to a false impression that some nanofluids produce an improvement in heat transfer performance. Moreover, using nanofluid will require additional pumping power to achieve the corresponding base fluid’s Reynolds number. Finally, existing single-phase liquid correlations of the heat transfer coefficient and pressure drop are compared and show good agreement in predicting nanofluid behaviour. The third objective aims to determine recommended nanoparticle concentrations of a typical nanofluid used within immersion cooling of a data centre server; this is based on two different designs for immersion cooling by Iceotope at varying Reynolds numbers. This is determined by calculating the heat transfer and pressure drop through the server for various values of the volume fraction of nanoparticles by using a model created within the finite element solver, COMSOL. The server’s total power consumption as a function of its CPU temperature and cooling system, including the increased pumping power required for varying nanofluid concentrations, are predicted and used in a proposed novel methodology to evaluate potential economic trade-offs in utilising nanofluids within an immersed liquid cooled data centre. This methodology is used to calculate the minimum total costs and economically optimum volume fraction of a nanofluid with this type of data centre. The results show that Al2O3-water nanofluids showed the highest thermal performance with respect to other nanofluids because Al2O3 has the highest thermal conductivity, however Al2O3-water nanofluids are the most viscous giving it the highest pressure drop and consequently the highest pumping costs. Under the economic factors used, it is found that the most cost effective fluid for both the server with cooled plate having two parallel tubes and the server with cooled plate having serpentine tube configuration is the base fluid (water).

The objective of this work was to extend the range of turbulent burning velocities presented in the literature by performing new measurements at elevated pressures over a range of turbulent velocities, u' (r. m. s. deviation from the mean velocity). The influence of equivalence ratio was investigated for a fixed u'= 2 m/s at 0.5 MPa for methane, a 70 % methane/ 30 % hydrogen mixture, methanol, isooctane and a gasoline (Shell dutch-pura). The mixtures were varied from the lean ignition limit to rich limit up to a maximum equivalence ratio of 2. Further measurements were performed with iso-octane and the effect of r. m. s. turbulent velocity (u' = 0.5,1,2,4,6 m/s), pressure (P = 0.1,0.5,1 MPa) and equivalence ratio (0= 0.8,1.0,1.2 and 1.4) were investigated, to produce a database. Turbulent flames were centrally ignited in isotropic turbulence in the Leeds fan stirred bomb. Flame progress was monitored using high-speed schlieren photography and pressure measurements. The turbulent burning velocity based on the production of burned mass, ur was obtained from both techniques. The burning velocities obtained from schlieren imaging and pressure measurements were in good agreement. The turbulent flames continually accelerated throughout the time that they were monitored. This was the result of turbulent flame development, with the range of turbulent scales wrinkling the flame surface increasing as the flames grew. It was shown that flame development occurred primarily as a result of the kernel radius, rather than the time from ignition. For each turbulent condition, corresponding spherically expanding laminar flames were ignited and imaged with schlieren photography. The measurements were processed to give the stretch free laminar burning velocity and also the Markstein number, a measure of the influence of stretch on the burning velocity. The peak laminar burning velocity was found to be in the range 1>ø>1.2 depending on the fuel. At 0.5 MPa a number of the flames were observed to become cellular, in some cases this occurred from ignition, this has been linked to negative Markstein numbers which were observed with lean methane and rich iso-octane - air mixtures. The peak in the turbulent burning velocity with equivalence ratio varied considerably with the different fuels, and did not correspond to that for the laminar burning velocity. In the case of methane and the 70 % methane/ 30 % hydrogen mixture the peak turbulent burning velocity was lean of stoichiometric. In contrast, for iso-octane and the gasoline, the peak in %, was beyond 0=1.3. It was concluded that the shift in the peak could be explained by comparison with the measured Markstein numbers. In the iso-octane database, pressure was not observed to have a significant influence on the turbulent burning velocity for some conditions, however, when fuel rich higher pressure gave an increase in utr. Turndown of utr, (the burning velocity does not increase as with the turbulent velocity) was observed at high turbulence velocities, although this depended on the equivalence ratio. The measurements were then compared with existing turbulent burning velocity expressions and correlations. In general these expressions were found not to predict the effect of equivalence ratio well.

Availability and sustainability of fuels for road and air transport is essential for economic development and growth of any nation. New alternative fuels provide an opportunity to limit the use of ever declining conventional petroleum oil reserves as well as offsetting CO2 generation from their use. Liquid fuels have the highest energy density for transportation applications and synthetic liquid fuels, which can be produced from renewable non-food bio feedstock offer an exciting opportunity for partial or even total substitution of remaining fossil fuel supplies. It is therefore of great interest to study the fundamental combustion characteristics of these fuels if they are to be used commercially. This work is aiming at characterising the auto-ignition properties of individual fuel components representative of the chemical families present in the synthetic fuels which in this case are toluene, iso-octane, n-heptane, and bio-alcohols; ethanol and n-butanol. The auto-ignition characterisation was made by measurements of ignition delay times, τ. The time τ for these fuels and their blends were measured after rapidly compressed to an elevated pressure and temperature using a Rapid Compression Machine (RCM). RCM provides good platform to study the fuel auto-ignition process without complicated physical effects in engines which are continually changing. However, they are not without problems, practical applications are usually not within the ideal conditions. Different machines have different extent of deviation from ideal conditions, making comparison of results between rigs difficult. In the present study, a dedicated work was conducted to study the difference between the measurements originated from these rigs and were characterised against their deviations from ideal conditions. These cover chemical reaction during the finite compression time, the effects of heat loss during the ignition delay period, the effects of piston displacement (piston bounce), and non-homogeneous auto-ignition. An interesting aspect of the study is that a plot of the measured different delay times at a given temperature, on the separate machines, against the corresponding degrees of reaction during compression, when extrapolated to zero reaction, yield a more accurate delay time for that condition. As the temperature is increased, so also are the oscillatory pressure amplitudes generated at the auto-igniting hot spots. This is in line with other studies of hot spot auto-ignition. Measurements of ignition delay times of different chemical groups separately and when blended with each other were made. They provided an understanding of how their interaction influences the overall ignition delay times. When blended the change of their τ values do not vary linearly especially when the blended components have large difference in reactivities. Toluene for example, which is commonly known for its long ignition delay times, was made extremely reactive when blended with n-butanol. Comparison of addition of bio-alcohols (ethanol and n-butanol) on gasoline surrogate fuel (TRF) showed that at lower temperatures, they both increased the ignition delay times of TRF, while at high temperatures they reduced TRF delay times to almost the same value. n-butanol started to reduce TRF delay times at lower temperatures compared to ethanol. Development of auto-ignition blending laws offers an opportunity to enable quick methods for choosing an appropriate blend for a particular application. In this work, a Linear by Mole (LbM) auto-ignition blending law was proposed, it uses the measured ignition delay times of individual components in the blend and varies them linearly with the fractional concentration of each component. This was found to be satisfactory only for blends of chemical families without NTC behaviour such as CH4/H2, for fuels with NTC behaviour an empirical based law was generated for the conditions studied. Overall, this study has broadened our understanding in auto-ignition behaviour of selected individual fuel components and their blends at varying conditions of pressure, temperature and concentration. It has also enabled substantial development of Leeds RCM to achieve fast compression with good piston damping.

There is a growing need for secure, flexible and inexpensive energy across the world, however there is also a need to simultaneously curb emissions of greenhouse gasses and toxic pollutants. Fossil fuel combustion is expected to meet a significant portion the world's growing energy demand, however CO2 emissions need to be mitigated to avoid potentially catastrophic effects caused by global warming. Carbon capture and storage (CCS) technologies have been developed to permit the use of fossil fuel combustion in a future with strict controls over greenhouse gas emissions. CCS technologies are still yet to be deployed at large industrial scales, and it is necessary to reduce the efficiency overheads associated with CCS before the technology is economically feasible. Computational modelling can play a significant role in designing and optimising CCS technologies for power generation due to its flexibility and comparatively low costs. The work in this thesis develops and validates models for predicting mercury oxidation and thermal radiation under oxyfuel combustion conditions, which is a promising CCS technology that is competitively placed for large-scale implementation. The oxidation of mercury is a key chemical process in mitigating emissions of the toxic metal, and predicting the principal oxidation pathways will improve the design of control technologies. Thermal radiation, which is the most significant mode of heat transfer at combustion temperatures, is a very important physical mechanism for predicting many properties of combustion, such as gas temperatures, chemical reaction rates and heat fluxes, and thermal radiation models must accurately account for changes in the combustion environment. The models developed and validated in this thesis provide new approaches to predict mercury oxidation and thermal radiation under oxyfuel conditions. The results and conclusions from this work offer clear guidance on methods to model thermal radiation in oxyfuel conditions, and provide new insight on the mercury oxidation mechanism.

Cyclic thermal regenerators have an important industrial application, however most of the design techniques are old and use many simplifying assumptions. With the advent of digital computers many of these simplifying assumptions can be examined, however there is very little scope for practical verification. An apparatus has been designed, built and fully commissioned to investigate thermal regenerator characteristics. The analog operation of the apparatus is enhanced by four computer aspects. Commissioning of the regenerator highlighted how the response of apparatus can be influenced by the peripheral pipework around the test section, radiation effects from the heaters and jetting of air into the regenerator packed bed section. A series of experiments, comprising runs (a maximum of ten period changes) within a set-up (a known air flowrate and packed bed length) were formulated to examine the regenerator response, using three types of spherical packing, steel lead glass and alumina. A method of obtaining convective heat transfer coefficients was produced and represented in two forms. One in a graphical form which can be used for cyclic regenerator designs without the prior knowledge of a heat transfer coefficient, whilst the other form is data stored on a computer disc which produces convective heat transfer coefficients for regenerator raw data. Analysis of the practical data clearly shows the importance of the number of cycles required to reach equilibrium, test bed heat leak and packing intraconduction. Finally, prior to each cyclic set-up a single shot run was performed and the convective heat transfer coefficient obtained using Darabi's(1981) graphical technique. This offered a unique opportunity to compare cylic and single shot characteristics for the same physical system.

Described in this thesis are the results of an experimental and theoretical study of the effects of turbulence on combustion in single and dual chamber spark ignition engines. The techniques adopted in the experimental study included the use of high speed cine photography, and the collection of simultaneous cylinder pressure records using an on-line computer. The experimental results confirmed the potential of the dual chamber design for increasing burning rate, and for controlling the level of turbulence within an engine cylinder. High speed photographs were filmed through a perspex window in the engine cylinder head. These showed that flame propagation was much faster when the engine was fitted with a divided chamber cylinder head than when equipped with a disc shaped single chamber head. The acceleration of combustion rate has been shown to be a function of flow velocity through the interconnecting orifice during the compression stroke. At very high flow velocities the nozzle became choked, and engine performance was impaired. In the theoretical work, a computer model for the thermodynamic cycle of an engine was developed. The use in this model of empirical laws to describe combustion rate was shown to be inadequate; this was primarily because of uncertainty in the length of the combustion period, which one needs to specify when using this method. When burning velocity data (derived from work by colleagues using a turbulent combustion bomb) were incorporated into the model, good qualitative results were possible. The use of an empirical law to describe the effect of turbulence on the burning velocity of a developing flame was, however, shown to be inaccurate. The turbulent flame front in an engine is a thick reaction zone containing pockets of unbumt charge. Analysis o f data for flame projected area (derived from high speed photographs) and simultaneous cylinder pressure data, revealed that a considerable quantity of unburnt charge was present behind the visible flame front. There was some evidence that a greater proportion of unburnt charge was present behind the flame when the mixture was lean than when it was stoichiometric. Modelling of this effect by assuming that mass, once entrained, would burn at an exponential rate, was shown to produce reasonable results.

The turbocharger remains one of the best means available to the engine developer to satisfy the power density demands on a modern internal combustion engine. This simple device uses the otherwise waste exhaust gas energy to provide significant improvements in the volumetric efficiency or ‘breathing capacity’ of an engine. In order to maximize the energy of the exhaust driving the turbine, most applications utilize pulse turbocharging where a compact exhaust manifold feeds the highly pulsating exhaust flow directly into the turbine wheel. This thesis considers the influence that this pulse-charging has on a double-entry turbocharger turbine. The design of this turbine plays an important role in much of the research presented in this thesis. The turbine is equipped with a mixed-flow rotor with 12 blades that are fed by a 24 blade nozzle ring. The circumferentially divided volute is designed with two gas inlet passages that each feed a separate 180° section of the nozzle ring. Thus, there is no communication between the entries from the volute inlet to the exit of the nozzles. At the exit to the nozzle, the fluid from both inlets expands into an interspace that spans the circumference of the rotor inlet. This small volume that is formed between the nozzle and the mixed flow rotor is the first area where interaction between the flows can occur. The core of this report contains three main divisions: Steady flow experimental results, CFD modelling, and unsteady flow experimental results. These sections are preceded by an introduction explaining the background of the research study, and an essential outline of the equipment and the method of experimentation. The aim of this work is to use a combination of experiments and computational modelling to build up a picture of the performance of the turbine under a wide variety of flow conditions that will eventually lead to further insight into its unsteady performance. First, a comprehensive steady-state experimental data set was obtained to establish the base-line turbine performance. Steady, equal admission tests yielded excellent performance, peaking at 80% efficiency. Owing to the double-entry arrangement, steady flow could also be introduced in the two inlets unequally. During unequal, steady-state operation a notable decrease in performance was observed. The correlation between the ratios of entry pressures and the efficiency of operation was apparent but essentially independent of which flow was varied. In the extreme, when the turbine was only partially supplied with air, the consequence was a 28 point decrease in performance at the optimal velocity ratio. Despite the division between the two entries, the experiments showed that the flows through each inlet were interdependent. Compared to full flow,of the performance of the turbine under a wide variety of flow conditions that will eventually lead to further insight into its unsteady performance. First, a comprehensive steady-state experimental data set was obtained to establish the base-line turbine performance. Steady, equal admission tests yielded excellent performance, peaking at 80% efficiency. Owing to the double-entry arrangement, steady flow could also be introduced in the two inlets unequally. During unequal, steady-state operation a notable decrease in performance was observed. The correlation between the ratios of entry pressures and the efficiency of operation was apparent but essentially independent of which flow was varied. In the extreme, when the turbine was only partially supplied with air, the consequence was a 28 point decrease in performance at the optimal velocity ratio. Despite the division between the two entries, the experiments showed that the flows through each inlet were interdependent. Compared to full flow, 4 when the pressure in one entry was low, the second entry could swallow more mass, and when it was high, the second entry swallowed less. A three-dimensional CFD model was constructed in order to permit a detailed study of the flow in the double-entry design and answer specific questions regarding the observed steady-state performance. For both equal and unequal admission simulations, the model showed close agreement with the experimental mass flow behaviour and reproduced the measured efficiency trends quite well. The interdependence of the swallowing capacity of the two inlets was also predicted by the model, thereby allowing the analysis of the physical flow effects that drive this trend. It was found that the interspace region near the tongues was the site of much of the interaction between inlets. A major emphasis of this modelling work was also to discover areas of loss generation that could lead to the decrease in performance. By focussing on partial admission, this study found that the windage loss in the interspace region of the non-flowing entry proved to be one of the more significant areas of loss generation. Pulsating air flow was then introduced using the range of frequencies typically produced by an internal combustion engine. The operating point of the turbine, traced an orbit within a 3-D space defined by three non-dimensional parameters: velocity ratio, pressure ratio across inlet one, and pressure ratio across inlet two. Direct comparison between steady and unsteady values at the same pressure ratios and velocity ratio was possible due to the large amount of steady data measured. Thus, a quasi-steady versus unsteady comparison was made on the basis of efficiency, mass flow and output power. In general, under pulsating flow conditions, the turbine behaved quite differently than that predicted by the quasi-steady assumption. Lower frequency, higher amplitude pulsations produced the lowest unsteady cycle-averaged efficiency and also produced the most significant departure from quasi-steady behaviour.