A thesis submitted to the Faculty of Engineering, University of the Witwatersrand, Johannesburg, in fulfilment of the requirements for the degree of Doctor of Philosophy. Johannesburg, 1990 / This work was carried out in response to the need for a simple engineering method for the thermal analysis of direct-contact air-water heat exchangers. A simple method of performance analysis is developed which is directly analogous and consistent with the fundamental approach used in conventional heat exchanger analysis and one in which the algebraic form of the overall equation and the grouping of each of the parameters are apparent. The range of conditions considered are air and water temperatures of between 0 and 50 DC and barometric pressures ranging from 80 to 120 kPa. The air conditions considered range from completely dry to completely satucated with water vapour. Both air cooling and water cooling processes are considered. [Abbreviated abstract. Open document to view full version]

The determination of the heat transfer through a flat wall of which one surface is isothermal at a temperature of t_a and the other surface is isothermal at a temperature t_b is the simple problem of heat transfer. The equation: Q= (kA(t_a - t_b))/L Permits an easy solution of the problem where Q is the heat transfer, k is the thermal conductivity, A is the area through which the heat is transferred, and L is the distance between the two surfaces. The equation is only applicable where the area A is constant. This equation may be used without appreciable error for insulated enclosures such as furnaces where the insulation thickness is very small in comparison with the dimensions of the enclosure. Shape factors have been applied to this basic equation so the equation may be used in the determination of heat transfer where the area A is not constant and the effect of corners can not be neglected. The equation then becomes Q= (fkA((t_a - t_b))/L Where f is the shape factor. In 1947, T. S. Nickerson for a Master’s thesis at V.P.I. determined the values of the shape factor where the above equation is applied to cylindrical enclosures having flat ends and relatively thick walls of uniform thickness. Mr. Nickerson solved this problem analytically by the relaxation method. His solution depended upon the inside and outside surfaces of the insulation about the enclosure being isothermal surfaces. The values were calculated for combinations of ratios of insulation thickness to length of enclosure and length of enclosure to diameter of enclosure. This investigation is an experimental determination of these values using gypsum plaster cylinders of different combinations of ratios of length to diameter. However, before tests could be conducted on the cylindrical enclosures, the conductivity of gypsum plaster, the insulation about the cylindrical enclosure, had to be found. The method of determination of the conductivity and the values are given in Appendix A. / Master of Science

The primary objective of this Investigation was to determine the heat transfer rate and characteristics of three vibrated wires In a parallel flow. A simple experimental apparatus was constructed with which to pursue the investigation, and various results were obtained. Three wires of different diameters were individually tested in two configurations. One configuration consisted of electrically heated stationary wires which transferred heat to air flowing at various velocities and parallel to the wires. This was investigated to compare the results with previously reported investigations of small cylinders in parallel flow. The second configuration consisted of vibrating wires at various air velocities In parallel flow and at various frequencies and amplitudes. The results show that for the stationary wires the heat transfer characteristics are In accord In principle with the results of similar investigations. The results obtained when the wires were vibrated show that vibration in forced convection Increased heat transfer rate up to a certain value of flow Reynolds number after which the effect of vibration on heat transfer was negligible. The effect of amplitude and frequency on the Improvement of heat transfer rate was greater at a particular range of values of frequency and amplitude. As both frequency and amplitude were Increased above this range, the rate of Improvement In heat transfer was not as great. / Master of Science

A vibrated bed is a mobile layer of solid particles contained in a vessel that is vibrated vertically. This study investigates bed dynamics and heat transfer from a vertical surface in shallow vibrated beds in absence of aeration. In general, "shallow" means a depth-to-width ratio less than one. In this study, bed depth is 30 mm, and this ratio is about 0.2. All experiments are at 25 hertz and at vibrational amplitudes affording peak accelerations between 2 and 7 times gravity. The study uses spherical glass beads of two densities and "Master Beads," nearly spherical particles of a crude, dense alumina, in size fractions from 63 to 707 micrometers. A disc embedded in the vessel floor, vibrated at 4.5 kilohertz, gives data on bed-vessel separation, showing it to occur later than predicted by plastic, single-mass models. The delay is attributed to bed expansion, monitored by piezoelectric force gauges mounted on floor and wall of the vessel. In large-particle beds, bed-vessel collision occurs simultaneously everywhere. In small-particle beds, exhibiting an uneven top surface, collision occurs first at the side walls and moves toward the center. In small-particle beds, pressure gradients appearing during the bed's free flight drive a horizontal component of particle circulation from the vessel's side walls toward its center. An apparent viscosity of the bed, estimated crudely by pulling a rod through it, influences this component's velocity. In beds of large particles, circulation is almost entirely vertical, a layer of two or three particles moving downward at a wall, and a slow return flow moving upward elsewhere. The study confirms the downward wall motion to be driven by friction. Heat transfer closely follows trends in rate of circulation. Greater dependence upon vibrational intensity is seen in small-particle beds. Values as high as 578 W/m²-K are measured. Comparison of vertical-surface heater geometry with an earlier horizontal tube shows the former to be generally superior for surface-to-bed heat transfer. / Ph. D.