The structure of deflagration to detonation transition in gases is studied through high-resolution numerical simulations. Quasi-steady wave structures have been shown to play an important role in this transition process in previous theoretical studies assuming an Arrhenius single-step reaction model with asymptotically high activation energy. In chapter 2, these are considered in the context of numerical simulations and analysis tools are developed to automatically identify quasi-steady structures within one-dimensional simulation results. These tools are applied to simulations of shock-induced transition in order to reveal the evolution of the waves in time and demonstrate their persistence and scaling over a range of both activation energy and energy release rate. An extension of the system of equations to multiple ideal gases is considered in chapter 3. Particular attention is paid to the method of the numerical solution of this model to avoid errors produced at material interfaces, and an approximate Riemann-problem solution derived to allow for solution of the model by the Weighted Average Flux method. This model is validated for inert flows by the simulation of two-dimensional Richtmeyer-Meshkov experiments. The model is further extended to model reactivity flow by considering the reactant and product gases as two constituents of the system and modelling the conversion between these by a simple one-step reaction mechanism. The complete reactive multi-gas model is then applied to simulations of two-dimensional shock-flame interactions resulting in multidimensional DDT, and the DDT event in these simulations examined in detail. Finally, in chapter 4, the role of chemistry modelling for the DDT process is investigated.

No description available.

The aim of this work was to learn more about the chemical mechanisms by which nitric oxide (NO) is created an destroyed in combustion. Premixed, laminar flames of H<SUB>2</SUB> + O<SUB>2</SUB> + N<SUB>2</SUB> and CH<SUB>4</SUB> + O<SUB>2</SUB> + N<SUB>2</SUB>, burning at atmospheric pressure with final temperatures between 1840 and 2520 K were studied. Temperature was measured along the axes of these flames using coherent anti-Stokes Raman spectroscopy (CARS). A procedure was developed for measuring the concentration of OH along the axis of these flames from a two-dimensional image of laser induced fluorescence (LIF). This method compensated for variations in the width and intensity of the laser sheet used to excite the molecules, required only one calibration and was applicable to either fuel-lean or fuel-rich flames burning unshielded from the entrainment of air. Concentrations of NO<SUB>x</SUB> (here, NO<SUB>x</SUB> = NO + NO<SUB>2</SUB> + HNO, but [NO]/[NO<SUB>x</SUB>]>0.98) were measured along the axes of the flames. Whereas the Zel'dovich mechanism was sufficient to explain the formation of NO<SUB>x</SUB> in the burnt gases of fuel-lean flames, NO<SUB>x</SUB> was produced faster than predicted by the Zel'dovich mechanism alone in fuel-rich flames with final temperatures below 2100 K. Including the production of NO <I>via </I>NNH, in N<SUB>2</SUB> + H = N<SUB>2</SUB>H, followed by N<SUB>2</SUB>H + O → NO + NH, gave a reasonable fit to the measurements. Predictions of [NO<SUB>x</SUB>] obtained using a reaction scheme for combustion devised by Miller & Bowman (1989, <I>Prog. Energy Combust. Sci., </I>15, 287) were below the measured [NO<SUB>x</SUB>] in the cooler (T < 2050 K), fuel-rich flames of hydrogen, because this scheme does not include formation of NO <I>via</I> NNH. Miller & Bowman's (1989) mechanism was also found to underpredict [OH] in fuel-rich and [NO<SUB>x</SUB>] in fuel-lean and fuel-rich flames of methane. The concentration of NO<SUB>x</SUB> was also measured in flames of H<SUB>2</SUB> + O<SUB>2</SUB> + Ar, with traces of NO or NH<SUB>3</SUB> added to the burner supplies. For flames seeded with NO, some of the additive (up to 60%) was destroyed in the reaction zone. When either NO or NH<SUB>3</SUB> was added, [NO<SUB>x</SUB>] remained constant along the burnt gases of a fuel-lean flame but increased downstream of the reaction zone of a fuel-rich flame; this rapid loss and subsequent regeneration of NO<SUB>x</SUB> in a fuel-rich flame doped with NO was not predicted by several models of combustion chemistry.

No description available.

Emissions of the gaseous pollutants NO and NO₂ (collectively known as NO_x) are subject to increasingly stringent controls. The following reaction scheme has been proposed as a method of reacting the toxic gases NO and CO: 3 NO + 2 Fe ® Fe₂O₃ + 3/2 N₂(VII) ; 3 CO + Fe₂O₃ ® 2 Fe + 3 CO₂. The net effect of these reactions is CO + NO ® CO₂ + ½ N₂. Reaction (VII), above, has been studied in detail in a very sensitive balance, a differential reactor and a laboratory scale fluidised bed. Kinetic parameters for reaction (VII), along with associated activation energies, were measured using the balance. The reaction was found to be marginally slower in the balance than in the differential reactor. The rate of reaction was found to be substantially faster (by a factor of up to ten) in a fluidised bed. The proposed explanation for the large difference in the rate of reaction measured in the balance or differential reactor and in the fluidised bed depends on the attrition of a weakly adherent oxide from the surfaces of iron particles in the fluidised bed. The latter stages of the reaction were also studied in the balance. It was found that after a thick layer of oxide had built up on the surface of a particle of Fe, the rate became almost independent of [NO] in the surrounding gas. This behaviour was ascribed to a solid-state diffusion-mechanism controlling the rate of reaction. Solid-state diffusion coefficients were measured, along with associated activation energies. Particles were also examined in an electron microscope both before and after reaction. Those particles which had been reacted in the balance were found to be very different in appearance from particles which had been reacted in the fluidised bed. In a separate study, the rates of oxidation of coal-chars in a fluidised bed which was contained in a quartz tube were measured. Three coal-chars of differing ranks were studied. The dependence of the rate of reaction on [O₂] and temperature were examined for each coal-char; rate constants were derived, along with activation energies. Chars from Cynheidre (high rank) and Reitspruit (middle rank) appeared to obey a similar rate law, the mechanistic implications of which are here studied in detail. Char from a lignite appeared to obey a very different rate law; a mechanism is also proposed for reaction of this char. A study of the method of addition of the char to the bed was also conducted. This concluded that the method for adding char to the bed was very important when considering the formation of NO_xduring combustion of the char; the method of addition had less effect on the rate of formation of CO and CO₂. It was also found that a stainless-steel heat-shield which prevented direct radiative heating of the upper surface of the fluidised bed and the freeboard above it had a marked effect on the rate of reaction of the coal-chars.

The first part of the thesis describes an experimental investigation of forced lean fully premixed turbulent flames with special emphasis placed on the amplitude dependence of their response. The heat release rate estimates from HO*/CH* chemiluminescence, flame surface density (<i>FSD</i>) from OH Planar Laser Induced Fluorescence (PLIF), and reaction rate (<i>RX</i>) imaging using simultaneous CH₂O and OH PLIF were used. The heat release response became non-linear after inlet velocity amplitudes of around 15% of the bulk velocity. This value depended on the forcing frequency and the equivalence ratio. The non-linearity was found to occur when the shear layers rolled-up into vortices. These vortices not only generated flame area when the flame wrapped around them, but also caused large-scale flame annihilation events. The results suggest that the flame kinematics play the major role in the saturation mechanism in these flames. The second part of the thesis describes the experimental investigation of the response of imperfectly premixed flames to acoustic oscillations. The results suggest that the flame response was controlled by two mechanisms, namely equivalence ratio perturbations and flame kinematics. The former mechanism dominated in low frequency conditions, under which the flame was compact, resulting in nonlinear amplitude dependence. The flame kinematics seemed to play an important role only when the flame length was comparable to the acoustic wave-lengths. The forced flame measurements were finally compared successfully with the response to self-excited flames which suggests that the mechanism of heat release variation investigated in the forced response studies is representative of the true limit-cycle behaviour of self-excited flames.

No description available.

No description available.

The aim of this thesis is the preparation and study of new materials to be used as electrolytes in Direct Methanol Fuel Cells (<i>DMFC)</i> in order to reduce crossover of methanol. Different materials have been prepared, including polymer dispersed liquid crystals, crosslinked polymers with high protonic conductivity and polymeric liquid crystals. Moreover, a methodology to characterise potential new materials for their application in <i>DMFC</i> has been developed. The absorption properties of water and methanol in the materials have been studied in different conditions and compared with the results obtained for commercial Nafion membranes used in DFMC. The absorption and diffusion properties of water and methanol through the materials have been studied by means of Fourier Transform Infrared Spectroscopy and thermal analysis techniques. The methodology provides important information to analyse the interactions between the different functional groups of the polymers and the solvents commonly used in <i>DMFC.</i> Finally, a device has been designed and set-up to analyse the effect on crossover of methanol of the electro-osmotic drag and the polarisation of the polymeric electrolytes.

No description available.