Laminar flame speed of a novel liquid bio-fuel has been determined experimentally using a closed spherical combustion vessel of 29 L equipped with two pairs of fused silica windows for optical access at atmospheric pressure and elevated temperature conditions. Schlieren technique was used to visualize and record the temporal evolution of the outwardly spherical flame front, and an in-house developed Matlab code was employed to process the flame front images and calculate its area by applying several image processing techniques. The test conditions consisted of varying the fuel-air mixture equivalence ratio at atmospheric standard pressure and different initial temperatures. Validation of the present method was achieved by measuring and comparing the flame speed of methane/air and n-heptane/air mixture with their published counterparts. Experimental results revealed comparable laminar flame speed of the novel liquid biofuel (1, 3- dimethoxyoctane) to heavy liquid hydrocarbons such as n-heptane and isooctane, especially at stoichiometric and fuel rich conditions. Additionally, the flammability limits of this novel fuel showed similarities with those of gaseous hydrocarbons fuels (e.g. methane, ethane) but higher than those of liquid hydrocarbons (e.g. diesel, gasoline). / February 2016

Gasoline is a complex fuel containing hundreds of species, and, therefore, it is quite difficult to model all components present in gasoline. Alternatively, researchers tend to employ simpler surrogates that mimic targeted physical and chemical properties of gasoline. Two properties of gasoline, i.e., autoignition and laminar flame speed, play key role in the overall performance of spark-ignition and modern engines. For fuel-engine optimization, it is very important to have simple models which can accurately predict autoignition and laminar flame speed of gasoline. In this work, universal laminar flame speed correlation is proposed for typical gasolines. This correlation is based on Monte-Carlo simulations of randomly generated mixtures comprising of 21 gasoline-relevant molecules. Laminar flame speed of each molecule is numerically computed over a wide range of thermodynamic conditions using detailed chemical kinetic models, while flame speed of each mixture is estimated using a mixing rule. The proposed universal correlation is validated against experimentally-measured laminar flame speed of various gasoline fuels.

Natural gas is the primary fuel used in industrial gas turbines for power generation. Hydrocarbon blends of methane, ethane, and propane make up a large portion of natural gas and it has been shown that dimethyl ether can be used as a supplement or in its pure form for gas turbine combustion. Because of this, a fundamental understanding of the physical characteristics such as the laminar flame speed is necessary, especially at elevated pressures to have the most relevance to the gas turbine industry. This thesis discusses the equations governing premixed laminar flames, historical methods used to measure the laminar flame speed, the experimental device used in this study, the procedure for converting the measured data into the flame speed, the results of the measurements, and a discussion of the results. The results presented in this thesis include the flame speeds for binary blends of methane, ethane, propane, and dimethyl ether performed at elevated pressures, up to 10-atm initial pressure, using a spherically expanding flame in a constant-volume vessel. Also included in this thesis is a comparison between the experimental measurements and four chemical kinetic models. The C4 mechanism, developed in part through collaboration between the National University of Ireland Galway and Texas A&M, was improved using the data presented herein, showing good agreement for all cases. The effect of blending ethane, propane, and dimethyl ether with methane in binary form is emphasized in this study, with the resulting Markstein length, Lewis number (Le), and flame stability characterized and discussed. It was noticed in this study, as well as in other studies, that the critical radius of the flame typically decreased as the Le decreased, and that the critical radius of the flame increased as the Le increased. Also, a rigorous uncertainty analysis has been performed, showing a range of 0.3 cm/s to 3.5 cm/s depending on equivalence ratio and initial pressure.

This study employs a novel approach in order to satisfy the need in industry for a computationally inexpensive means to modelling soot formation in engines fuelled by natural gas. The complex geometries found in practical combustion devices along with the requirement to solve turbulent, chemically reacting, and multi-phase flows necessitates this goal. A two-equation model, which tracks soot mass and soot number density, is employed. The goal is to apply this model in engine simulations at Westport Innovations, an industry partner. Experimental data is used to validate the model in various operating conditions. Numerical data obtained from a detailed sectional soot model is also used to augment available validation data, especially with respect to soot formation/oxidation mechanisms. The developed model shows good agreement compared to experimental data and the detailed sectional soot model among all cases considered and will be further tested and applied in Westport’s natural gas engine simulations.

This study employs a novel approach in order to satisfy the need in industry for a computationally inexpensive means to modelling soot formation in engines fuelled by natural gas. The complex geometries found in practical combustion devices along with the requirement to solve turbulent, chemically reacting, and multi-phase flows necessitates this goal. A two-equation model, which tracks soot mass and soot number density, is employed. The goal is to apply this model in engine simulations at Westport Innovations, an industry partner. Experimental data is used to validate the model in various operating conditions. Numerical data obtained from a detailed sectional soot model is also used to augment available validation data, especially with respect to soot formation/oxidation mechanisms. The developed model shows good agreement compared to experimental data and the detailed sectional soot model among all cases considered and will be further tested and applied in Westport’s natural gas engine simulations.

Biomass is among the primary alternative energy sources that supplements the fossil fuels to meet today’s energy demand. Gasification is an efficient and environmental friendly technology for converting the energy content in the biomass into a combustible gas mixture, which can be used in various applications. The composition of this gas mixture varies greatly depending on the gasification agent, gasifier design and its operation parameters and can be classified as low and medium LHV gasified biomass. The wide range of possible gas composition between each of these classes and even within each class itself can be a challenge in the combustion for heat and/or power production. The difficulty is primarily associated with the range in the combustion properties that may affect the stability and the emission levels. Therefore, this thesis is intended to provide data of combustion properties for improving the operation or design of atmospheric combustion devices operated with such gas mixtures. The first part of this thesis presents a series of experimental work on combustion of low LHV gasified biomass (a simulated gas mixture of CO/H2/CH4/CO2/N2) with variation in the content of H2O and tar compound (simulated by C6H6). The laminar flame speed, lean blowoff limit and emission levels of low LHV gasified biomass based on the premixed combustion concept are reported in paper I and III. The results show that the presence of H2O and C6H6 in gasified biomass can give positive effects on these combustion parameters (laminar flame speed, lean blowoff limit and emission levels), but also that there are limits for these effects. Addition of a low percentage of H2O in the gasified biomass resulted in almost constant laminar flame speed and combustion temperature of the gas mixture, while its NOx emission and blowoff temperature were decreased. The opposite condition was found when H2O content was further increased. The blowoff limit was shifted to richer fuel equivalence ratio as H2O increased. A temperature limit was observed where CO emission could be maintained at low concentration. With C6H6 addition, the laminar flame speed first decreased, achieved a minimum value, and then increased with further addition of C6H6. The combustion temperature and NOx emission were increased, CO emission was reduced, and blowoff occurs at slightly higher equivalence ratio and temperature when C6H6 content is increased. The comparison with natural gas (simulated by CH4) is also made as can be found in paper I and II. Lower laminar flame speed, combustion temperature, slightly higher CO emission, lower NOx emission and leaner blowoff limit were obtained for low LHV gas mixture in comparison to natural gas. In the second part of the thesis, the focus is put on the combustion of a wide range of gasified biomass types, ranging from low to medium LHV gas mixture (paper IV). The correlation between laminar flame speed or lean blowoff limit and the composition of various gas mixtures was investigated (paper IV). It was found that H2 and content of diluents have higher influence on the laminar flame speed of the gas mixture compared to its CO and hydrocarbon contents. For lean blowoff limit, the diluents have the greatest impact followed by H2 and CO. The mathematical correlations derived from the study can be used to for models of these two combustion parameters for a wide range of gasified biomass fuel compositions. / <p>QC 20130411</p>

Predictive modeling of pre-chamber combustion engines relies primarily on the correct prediction of laminar and turbulent flame speeds. While the latter has been rigorously derived, the former correlations are mostly semi-empirical and valid for a limited range of operating conditions. The current work aims at highlighting the fundamental significance of correct laminar flame speed prediction on numerical modeling of ultralean prechamber engine combustion. Gulder's empirical correlation for laminar flame speed was chosen for the current work. It was modified for ranges beyond what it was originally derived for. It was initially observed that the numerical results that utilize Gulder's correlation for the laminar flame speed underperform compared to the one computed from the skeletal GRI3.0 by Lu and Law. In all cases, Peters' turbulent flame speed correlation was used, which evidences that any potential difference comes from the laminar flame speed. Using Lu and Law's chemical mechanism as a reference for laminar flame speed calculations, the values of the empirical constants α, η, and ξ in Gulder's correlation were optimized to yield more accurate flame speeds at ultralean engine conditions. The updated Gulder's correlation for methane was implemented in CONVERGE, a three-dimensional computational fluid dynamics (CFD) solver, and validated against the experimental engine results from KAUST. The flame topology was also explored to correlate the observed behaviors in the pressure predictions among all tested cases. Finally, the Borghi-Peters diagram provides insightful information on combustion regimes encountered in pre-chamber combustion engines.

Electric field can be a viable method in controlling various combustion properties. Comparing to traditional actuators, an application of electric field requires very small power consumption. Especially, alternating current (AC) has received attention recently, since it could modulate flames appreciably even for the cases when direct current (DC) has minimal effects. In this study, the effect of AC electric fields on small coflow diffusion flames is focused with applications of various laser diagnostic techniques. Flow characteristics of baseline diffusion flames, which corresponds to stationary small coflow diffusion flames when electric field is not applied, were firstly investigated with a particular focus on the flow field in near-nozzle region with the buoyancy force exerted on fuels due to density differences among fuel, ambient air, and burnt gas. The result showed that the buoyancy force exerted on the fuel as well as on burnt gas significantly distorted the near-nozzle flow-fields. In the fuels with densities heavier than air, recirculation zones were formed very close to the nozzle exit. Nozzle heating effect influenced this near-nozzle flow-field particularly among lighter fuels. Numerical simulations were also conducted and the results showed that a fuel inlet boundary condition with a fully developed velocity profile for cases with long fuel tubes should be specified inside the fuel tube to obtain satisfactory agreement in both the flow and temperature fields with those from experiment. With sub-critical AC applied to the baseline flames, particle image velocimetry (PIV), light scattering, laser-induced incandescence (LII), and laser-induced fluores- cence (LIF) techniques were adopted to identify the flow field and the structures of OH, polycyclic aromatic hydrocarbons (PAHs), soot zone. Under certain AC condi- tions of applied voltage and frequency, the distribution of PAHs and the flow field near the nozzle exit were drastically altered from the baseline case, leading to the formation of toroidal vortices. Increased residence time and heat recirculation inside the vortex resulted in appreciable formation of PAHs and soot near the nozzle exit. Decreased residence time along the jet axis through flow acceleration by the vortex led to a reduction in the soot volume fraction in the downstream sooting zone. Electromagnetic force generated by AC was proposed as a viable mechanism for the formation of the toroidal vortex. By varying applied AC in a wide range of frequency and voltage, several insta- bility modes were observed, including flicking flames, partial pinch-off of flames, and spinning flames. High speed imaging together with Mie scattering techniques were combined to reveal the flame dynamics as well as the flow structure inside the flames. Original steady toroidal vortices triggered by AC were noted to exhibit axisymmetric axial instability in the flicking and partial pinch-off modes and non-axisymmetric azimuthal instability in the spinning mode. Electrical measurements were also conducted simultaneously to identify the voltage, current, and electrical power responses. Integrated power was noted to be sensitive to indicate subtle variation of flames properties and to the occurrence of axial instability. Under low frequency AC forcing with electrical conditions not generating toroidal vortices, responses of flames were further investigated. Several nonlinear flame responses, including frequency doubling and tripling phenomena, were identified. Spectral analysis revealed that such nonlinear responses were attributed to the combined effects of triggering buoyancy-induced oscillation of the flame as well as the Lorenz force generated by applying AC. Phase delay behaviors between the applied voltage and the heat release rate (or flame size) were also studied to explore the potential of applying AC in controlling flame instability. It was found that the phase delay had large variations for AC frequency smaller than 80 Hz and became saturated at over 80 Hz, which has been explained based on the interaction between the buoyancy and ionic wind. Electrical measurement showed the power consumed by the AC was smaller than 0.01% of the heat release rate from the flame. To improve the understanding on the electric current resulting from applying electric field on flames, a simplified one-dimensional model was developed in that the reaction zone was modeled as a thin ionized layer. Model governing equations were derived from species equations by implementing mobility differences depending on the type of charged particles, especially between ions and electrons. The result showed that the sub-saturated current along with field intensity was significantly influenced by the polarity of DC due to the combined effect of non-equal mobility of charged particles as well as the position of the ionized layer in a gap relative to two electrodes. Experiments with quasi-one-dimensional flames under DC were conducted to substantiate the model and measured currents agreed qualitatively well with the model predictions.

Emissions of soot and polycyclic aromatic hydrocarbons (PAHs) from incomplete burning of hydrocarbon fuels pose a great threat to the environment and human health. To reduce such emissions, a comprehensive understanding of their evolution process is essential. In this work, a series of research studies were conducted to evaluate sooting tendencies and to experimentally and theoretically develop PAH mechanisms. The sooting tendencies of oxygenated fuels were quantitively investigated in counterflow diffusion flames. Sooting limits are described by critical fuel and oxygen mole fractions, measured with a laser scattering technique. The addition of dimethyl ether displays non-monotonic behavior on sooting tendencies at elevated pressures, which is attributed to the chemical effect from kinetic simulations. The tendency of incipient soot formation of other oxygenated fuels (e.g., alcohol, acid, ether, ketone, and carbonate ester) was also assessed, using a similar approach. As the precursor of soot, PAH measurement using laser induced fluoresecnce was implemented to track the evolution processes from PAHs to incipient soot. Developing a PAH mechanism is essential to the understanding of soot formation; however, PAH formation and its growth process are not well understood. Based on previous research, PAHs with 5-membered rings are abundant in flames. Therefore, the growth of PAHs with 5-membered rings was investigated, using acenaphthylene (A2R5) as the example. The density functional theory (DFT) and the transition state theory (TST) were adopted to calculate potential energy surfaces and reaction rate coefficients. The existence of 5-membered rings appreciably impacts PAH production by facilitating the formation of planar PAHs with C2H substitution, thereby improving existing PAH mechanisms. In PAH mechanisms, the thermochemistry properties are not all calculated, but are hypothesized to be equal to those of a similar structure. The simulation accuracy of the hypothesis is explored here by discussing the sensitivity of the thermochemistry parameters in flame simulations. The group additivity method utilizing THERM codes is used to calculate thermochemistry properties. PAH loading affects the sensitivity of thermochemistry properties to both flame temperature and product yields. These results show that either accurate thermochemistry properties, or reverse reaction rates should be provided in the mechanism to improve simulation accuracy.

The thesis comprises effects of electric fields on a fundamental study of spheri­cal premixed flame propagation.Outwardly-propagating spherical laminar premixed flames have been investigated in a constant volume combustion vessel by applying au uni-directional electric potential.Direct photography and schlieren techniques have been adopted and captured images were analyzed through image processing. Unstretched laminar burning velocities under the influence of electric fields and their associated Markstein length scales have been determined from outwardly prop­agating spherical flame at a constant pressure. Methane and propane fuels have been tested to assess the effect of electric fields on the differential diffusion of the two fuels.The effects of varying equivalence ratios and applied voltages have been in­vestigated, while the frequency of AC was fixed at 1 KHz. Directional propagating characteristics were analyzed to identify the electric filed effect. The flame morphology varied appreciably under the influence of electric fields which in turn affected the burning rate of mixtures.The flame front was found to propagate much faster toward to the electrode at which the electric fields were supplied while the flame speeds in the other direction were minimally influenced. When the voltage was above 7 KV the combustion is markedly enhanced in the downward direction since intense turbulence is generated and as a result the mixing process or rather the heat and mass transfer within the flame front will be enhanced.The com­bustion pressure for the cases with electric fields increased rapidly during the initial stage of combustion and was relatively higher since the flame front was lengthened in the downward direction.