No description available.

No description available.

No description available.

No description available.

No description available.

No description available.

Current work focuses on the development and performance evaluation of advanced flamelet models for turbulent non-premixed and partially premixed combustion in RANS and large eddy simulation (LES) based modelling. A RANS based combustion modelling strategy which has the ability to capture the detailed structure of turbulent non-premixed flames, including the pollutant NO, and account for the effects of radiation heat loss and transient evolution of NO, has been developed and incorporated into the in-house RANS code. The strategy employs an 'enthalpy-defect' based non-adiabatic flamelet model in conjunction with steady or unsteady nonadiabatic flamelets based NO submodels. The performanceo f the non-adiabaticm odel and its NO submodelsh asb eena ssessed against experimental measurements and steady flamelet model predictions for turbulent CH4/H2 bluff-body stabilized and CH4-air piloted jet flames. Appreciable improvements in the mean thermal structure predictions have been observed in the piloted jet flames by consideration of radiation heat loss through the non-adiabatic flamelet model. Since transient effects were weaker in the piloted jet flame, both unsteady and steady non-adiabatic NO submodels provided similar level of improvement in the pollutant NO predictions in comparison to their adiabatic counterpartsT. ransiente ffectsw ere, however,d ominanti n the bluff-body flame. The unsteady non-adiabatic NO submodel provided excellent agreement with measured NO distribution in comparison to the appreciably overpredicted distribution by its steadyc ounterpart.T he strategyo f non-adiabaticf lamelet model in conjunctionw ith unsteady non-adiabatic NO submodel seems to provide an accurate and robust alternative to the conventional strategy of steady flamelet model with steady NO submodel. While addressing the limitations of steady flamelet model in regard to radiation and slow chemistry of NO is one objective of this research, extending the applicability of the model to partially premixed combustion has been pursued as the second objective. Flamelet/progress variable (FPV) approach based combustion models, which have the potential to describe both non-premixed and partially premixed combustion, have been incorporated in the in-house RANS and LES codes. Based on the form of the PDF for reaction progress variable, two different formulations, FPV-8 function model and FPV-P function model, have been derived. (Continues...).

Laminar premixed flames have applications in many residential, commercial, and industrial devices and processes. Of primary importance in their description is the burning rate which directly affects the heat release and pressure development in energy conversion systems. In the present study, fundamental properties of laminar premixed flames have been studied experimentally and numerically with an objective to understand the factors which influence their burning rate measurements. Spherically expanding flames following central ignition were employed to quantify the flame structure and propagation rate. The experiments were carried out in fan- stirred Leeds combustion bombs, while the numerical solutions were obtained by solving the reactive Navier-Stokes equations. In the early stages of flame propagation, ignition energy was found to significantly affect the flame development. The effect of flame stretch rate, as quantified by the Markstein length (Lb) of the mixture, proved critical in the ignition energy dependency of flame propagation. For relatively high values, an underlying . common variation of self-sustaining flame speed with radius could be indentified by the rapid convergence of curves for different ignition energies. This allowed the quantification of an optimum ignition energy at which the flame became independent of ignition at a smallest radius. For relatively low values, low energy spark ignition was found to give rise to a distorted and wrinkled flame kernel. This problem was found to worsen as the Lb decreased, such that its apparent measured value became increasingly influenced by any distortions produced by the spark. In the later stages of flame propagation, the walls of the bomb started to influence the propagation such that assumption of flame propagation at constant pressure was invalidated. Due to this effect, the maximum flame radius, below which accurate measurements could be gathered in an experiment, was severely restricted. If this effect was ignored, the resulting choice of the flame radius range used for extracting the burning velocity would be too small. If this effect was not ignored, the resulting choice would be inaccurate. A correction factor was developed to correct the obtained measurements, to that of a freely propagating flame under constant pressure assumption. Linear and non-linear relationships between flame ( speed and stretch rate were examined for their usage in extracting the burning velocity. It was found that the size of the bomb significantly influenced the accuracy of burning rate measurements, which had not been recognized before.

A novel ignition system was studied experimentally, in which small volumes of hydrogen peroxide -of the order of pl/s- were injected at the immediate site of ignition, during the firing of a focus discharge igniter (FDI). Initially, the new ignition system was evaluated at an atmospheric expansion rig in the Mechanical Engineering of Leeds University. Afterwards, experiments were undertaken in an atmospheric testing facility with an industrial Rolls-Royce Olympus combustion chamber using kerosene Jet-A1 as the fuel and atmospheric air as the oxidizer. The study concentrated on the determination of the lean ignition limits o f the kerosene-air mixture at various air mass flow rates with and without the addition of H20 2. Notable improvements, from 6.5% to 44%, in the ignition limits of the fuel-air mixture were attainable by using only a maximum amount o f 10.8pl/s o f H20 2 during only the ignition process. The study suggests that these improvements are directly related to the increase in the ignition efficiency of the ignitor, by radical enhancement through the injection of the H20 2 plasma medium. Comparisons were made, between a fuel atomiser that was in normal service and of the same device but washed, in order to test the igniter’s ability to initiate combustion under poor and high fuel spray qualities (FSQ). The results indicate an enhanced improvement in the ignitability limits during poor air-fuel mixing quality when using the H20 2 as described above. A biodiesel fuel was also selected to test the effect that the new ignitor system has in a low-volatility fuel. The question of how to create the small amounts of hydrogen peroxide that will be used for ignition was also approached. An idea of producing the required H20 2 came by studying the bombardier’s beetle unique mechanism which produces H20 2 for defending itself from predators. Simulation work using Chemkin was conducted to investigate the production o f H20 2 by passing hydrocarbon fuels through a catalyst. When passing propane-air through a platinum/ rhodium catalyst the simulations show that H20 2 production is possible in rates enough to supply the proposed novel ignition system. A more specialised study in the chemistry of the production of H20 2, from gas-turbine fuels, is suggested. A cost effective method o f an onsite H20 2 production in small amounts would be an ideal topic for further study.

No description available.