White Rose University Consortium logo
University of Leeds logo University of Sheffield logo York University logo

Experimental and theoretical studies of combustion rates at high pressure

AI-Shahrany, Ali Saied (2004) Experimental and theoretical studies of combustion rates at high pressure. PhD thesis, University of Leeds.

[img]
Preview
Text
uk_bl_ethos_410733.pdf

Download (15Mb)

Abstract

The thesis reports experimental and theoretical studies of premixed combustion rates at high pressure and temperature. It focuses on measurements of laminar and turbulent burning velocities at high pressures and temperatures approaching those in engines, with emphasis on flame instabilities. To encourage the development of such instabilities, mixtures with negative Markstein numbers were employed. Three different methods were used to measure burning velocities in a spherical bomb. The bomb was fitted with windows for observing flame propagation at the centre of the bomb and a transducer to measure pressure. Four fans at the wall of the bomb were employed for mixing and the generation of turbulence. The first two methods of measuring burning velocities were well established and involved central ignition. The third method was new and involved implosions of two flame kernels that originated at spark plugs mounted near the wall. It enabled the later stages of burning at the high pressures to be observed and burning velocities to be measured. The first method depended on highspeed schlieren photographic measurements of the flame speed, dr / dt , at different radii,r, supplemented by pressure measurements. The second method was employed when the flame front has propagated beyond the boundaries of the window and could no longer be observed. The expression for the burning velocity rested upon the assumption that the flame was spherical and the fractional pressure rise was equal to the fractional mass burned. Two different approaches were employed for the new third method, one was based on geometrical considerations, the other on the fractional pressure rise. A knowledge of the flame area and the appropriate geometrical analysis enabled two expressions to be obtained for the burning velocity. The agreement between the two different approaches for obtaining burning velocities, and the general consistency of the results for both initially laminar and turbulent flames, showed the technique to be accurate and suitable for obtaining burning velocities at high pressure. As a result, burning velocities, initially laminar, were measured for iso-octane - air at equivalence ratios ranging from 0.8 to 1.6 at initial pressures of 0.5 and 1.0 MPa. They were also measured for hydrogen - air mixtures at equivalence ratios of 0.3 to 0.5. Modification of the linear theory of flame instability of Bechtold and Matalon enabled the laminar burning velocity to be obtained from the values of unstable burning velocities. Enhancements of the laminar burning velocity of up to six fold were measured. Turbulent burning velocities were measured over a range of rms turbulent velocities ranging from 0.25 to 3 mls. It was found that these values of burning velocity were higher than those predicted from earlier expressions, derived predominantly from more stable flames close to atmospheric pressure. The possibility that turbulent burning velocities might be enhanced, not only by the effect of flame stretch at negative Markstein numbers, but also by flamelet instabilities was also investigated at high pressures and with mixtures with very low Markstein numbers. Stoichiometric and rich iso-octane-air flames were selected for this study and mixtures were ignited at initial pressures of 0.5 and 1.0 MPa. This enabled burning velocities to be measured up to 6 MPa.

Item Type: Thesis (PhD)
Academic Units: The University of Leeds > Faculty of Engineering (Leeds) > School of Mechanical Engineering (Leeds)
Depositing User: Ethos Import
Date Deposited: 06 Jan 2011 12:19
Last Modified: 07 Mar 2014 11:23
URI: http://etheses.whiterose.ac.uk/id/eprint/1180

Actions (repository staff only: login required)