The variation in the local radial heat transfer coefficient is reported for a disc rotating in still air up to 1650 r.p.m. and in a transverse or crossflow air stream of speeds up to 33m/s. These measurements have been made with the aid of a small sensor, using thermistors as heating elements. It is found that the heat transfer coefficient is governed by the main air flow, the rotation of the disc resulting in a small upward perturbation on this level. Tests with different disc aspect ratios, simulated by the disc protruding from a leading edge shroud, show that the radial distribution is greatly modified, but again the main air stream dominates the process. A thin film sensor has been developed to monitor the fluctuations in the heat transfer coefficient about the mean level as the disc rotates in the air stream. The local effects of rotation are examined closely. The velocity distributions around the stationary and rotating disc in still air and a transverse flow are presented. The experimentation is finally extended to the case of a simple train wheel shape, thus attempting to model the convective heat dissipation for the condition of train wheel braking.

No description available.

No description available.

No description available.

The present work covers three main areas. First, several suitable polymers (one-part silicone RTVs, including a certain silicone sealant and Dow Corning's DC 734) and swelling agents (aliphatic esters such as <I>iso</I>-Pentyl Ethanoate, <I>iso</I>-Butyl <I>iso</I>-Butanoate and <I>n</I>-Pentyl Propanoate; alkanes including <I>n</I>-Nonane and <I>n</I>-Decane) have been identified, along with effective methods of applying the rubber coatings to both flat and cylindrical substrates. The diffusion properties of these new solvent/polymer systems have also been measured. Also, various aspects of the holographic system have been examined. The fringe visibility problems mentioned above are found to stem from in the incorrect location of the optical diffusing screen and are exacerbated by an apparent design flaw in the thermoplastic holographic camera used by some workers causing the gradual re-localisation of the interference fringes after hologram development. Substantial improvements in the lifetime of the thermoplastic plates for the holocamera has resulted from surrounding the optical system with a simple plastic curtain to keep out airborne dirt and dust. The improved optical system has been used to look at natural convection mass transfer from a vertical plate, using a variety of swelling agents to cover a range of Rayleigh numbers (<I>Ra</I>). For the more volatile solvents the results are in fair agreement with the analytical prediction of Lorenz, <I>Sh</I>=0.411 <I>Ra</I><SUP>1/4</SUP>. For the less volatile swelling agents the measured values are significantly higher than predicted; this is believed to be due to background draughts increasing mass transfer rate. Natural convection from the cylindrical surface of a vertical rod has also been investigated. Other work performed in support of the main project includes the writing of a computer program to produce simulated fringe patterns in various geometries, the demonstration of a novel swollen polymer system using water and gelatin, and the investigation of temperature distributions in vertical rod arrays in free and mixed convection using both thermocouples and liquid crystal thermography.

A nonlinear two-dimensional free surface problem of an ideal jet impinging on an uneven wall is studied using complex variable and transform techniques. A relation between the flow angle on the free surface and the wall angle is first obtained. Then, by using a Hilbert transform and the generalised Schwartz-Christoffel transformation technique, a system of nonlinear integral-differential equations for the flow angle and the wall angle is formulated. For the case in which the wall geometry is symmetric, a compatibility condition for the system is automatically satisfied. Some numerical solutions are presented, showing the shape of the free surface corresponding to a number of different wall shapes. For the case of in which the wall geometry is asymmetric, a pair of conditions which determine the position of the stagnation point are revealed, using the integral form of the momentum equation. Thus the shapes of the free surface of the jet impinging on a few asymmetric uneven walls are shown. The stagnation point is located for each of the different cases. A flow passing through a porous film and then impinging on a flat solid boundary is analysed. Since the pressure field and fluid velocity are discontinuous along the film, the flow region is divided into two parts. At the film by analogy with Darcy'law the pressure difference is taken proportional to the flow rate. The free surface problem for this transpiration flow is formulated as a system of three coupled integral equations using a boundary integral method. As the system is solved numerically the normal speeds at the film and the shapes of the free surface for different permeability coefficients are presented.

No description available.

This thesis documents the application of a computer code developed by the author which makes possible the coupling of heat transfer between fluid and solid thermal models. The code was written using FORTRAN and couples the commercial computational fluid dynamics (CFD) software FLUENT with the Rolls-Royce finite element analysis program, SC03. The thermal modelling of a solid domain bounded by a fluid typically uses heat transfer correlations to define the heat flux at those boundaries. Considerable engineering judgement is required to appropriately select and apply these correlations, so that they accurately model the flow and geometry being considered. The objective of the coupling code is to replace the correlations with a CFD model of the fluid. The coupling is achieved by extracting metal temperatures determined from the finite element solver, using them to define CFD boundary conditions, and passing heat fluxes from the resulting CFD solution back to the finite element model. The finite element model then solves the newly defined problem and the process is repeated until a converged solution is obtained. The coupling code was evaluated through its application to two test cases. The first was an axisymmetric representation of a compressor stator well rig, the experimental apparatus or which comprised a two stage axial compressor, driven by a single stage axial turbine. The coupling code was used to model a temperature transient generated in the rig by injecting liquid nitrogen into the mainstream annulus, upstream of the compressor stages. For the second test case, an industrial application was chosen with real engine geometry. Using an axisymmetric finite element whole engine model of the Rolls-Royce Trent 500 aero-engine the code was employed to couple both axisymmetric and three dimensional representations of the fluid domain surrounding the pre-swirl system. Following the successful completion of these two test cases, the coupling code (now known as SC89) was production released by Rolls-Royce in July 2004 and is now available to their engineering community, as a design tool worldwide.

Tne effects of property variation on heat transfer to supercritical pressure fluids are reviewed, An experiment with supercritical pressure carbon dioxide is described for vertical flow in tubes of 5, 8 and 19mm diameter. Criteria are presented for the absence of buoyancy effects for upward and downward fLlow, and the published correlations for buoyancy free heat transfer to supercritical pressure fluids are critically examined, using the published data. Mixed convection correlations are presented for upward and downward flow in the form NU/Nu force = O (Gr/Reb 2.7) An experiment with atmospheric pressure water flowing in a vertical tube of 100mm diameter shows that the buoyancy correlations perform satisfactorily for mixed convection in fluids vrithout severe property variations.

Selective Laser Melting (SLM), a layer-based Solid Freeform Fabrication (SFF) process, was used to fabricate micro cross-flow heat exchangers from 316L stainless steel, bronze (Cu 90%, Sn 10%) and Inconel 718 powder. Their mechanical and thermal properties were determined using solid blocks of SLM material prior to the fabrication of the micro cross flow heat exchangers. Initially the process parameters for the fabrication of high density (>97%) parts for the different materials were defined. The mechanical and thermal properties of SLM parts were then measured. The tensile test results exhibited yield strength values superior to the parent metals, but also showed low tensile strength and ductility as a result of the inherent residual porosity (2-4%). Results obtained from the thermal conductivity of the stainless steel material system were in good agreement with the bulk material values. The heat transfer performance of the heat exchangers with either micro channels or lattice structures as heat exchange surfaces was investigated experimentally and the results were evaluated in terms of geometry and materials. The performance of the micro heat exchangers was found to be dependent not only on the choice of material but also on the heat exchanger media geometry.