Numerical and experimental studies of ethanol flames and autoignition theory for higher alkanes

In order to enhance the fuel efficiency of an engine and to control pollutant formation, an improved understanding of the combustion chemistry of the fuels at a fundamental level is paramount. This knowledge can be gained by developing detailed reaction mechanisms of the fuels for various combustion processes and by studying combustion analytically employing reduced-chemistry descriptions. There is a need for small detailed reaction mechanisms for alkane and alcohol fuels with reduced uncertainties in their combustion chemistry that are computationally cheaper in multidimensional CFD calculations. Detailed mechanisms are the starting points in identifying reduced- chemistry descriptions of combustion processes to study problems analytically. This research includes numerical, experimental and analytical studies. The first part of the dissertation consists of numerical and experimental studies of ethanol flames. Although ethanol has gained popularity as a possible low-pollution source of renewable energy, significant uncertainties remain in its combustion chemistry. To begin to address ethanol combustion, first a relatively small detailed reaction mechanism, commonly known as the San Diego Mech, is developed for the combustion of hydrogen, carbon monoxide, formaldehyde, methane, methanol, ethane, ethylene, and acetylene, in air or oxygen-inert mixtures. This mechanism is tested for autoignition, premixed-flame burning velocities, and structures and extinction of diffusion flames and of partially premixed flames of many of these fuels. The reduction in uncertainties in the combustion chemistry can best be achieved by consistently updating a reaction mechanism with reaction rate data for the elementary steps based on newer studies in literature and by testing it against as many experimental conditions as available. The results of such a testing for abovementioned fuels are reported here along with the modifications of reaction- rate parameters of the most important elementary steps and the addition and deletion of a few key steps relevant to these tests. A mechanism developed in such a hierarchical way starting with simpler fuels such as hydrogen and carbon monoxide to the fuels with one and two carbon atoms has reduced uncertainties in the combustion chemistry of a fuel. This reaction mechanism, consisting of 137 reactions among 30 species, provides a robust building block upon which an ethanol mechanism is developed. The San Diego Mech is extended for ethanol combustion by adding 55 new reactions and 6 new species. Specifically, 33 reactions are added that involve C₂H₅OH or one of the three isomers produced by abstraction of an H atom from it, CH₃CHOH, CH₂CH₂OH and CH₃CH₂O, and 22 reactions are added that involve acetaldehyde or one of the two isomers produced by abstraction of H from it, CH₂CHO and CH₃CO. Ethanol combustion is investigated on the basis of a new reaction mechanism, thus developed, consisting of 192 elementary steps among 36 species, augmented by 53 additional steps and 14 additional species to address the formation of the oxides of nitrogen and 43 steps and 7 species to address formation of compounds involving three carbon atoms. The mechanism is tested against shock-tube autoignition-delay data, laminar burning velocities, counterflow diffusion- flame extinction and measurements of structures of counterflow partially premixed and diffusion flames. Measurements on ethanol-air flames at a strain rate of 100 s⁻¹, employing prevaporized ethanol with a mole fraction of 0.3 in a nitrogen carrier stream, were made for the pure diffusion flame and for a partially premixed flame with a fuel-side equivalence ratio of 2.3 and involved thermocouple measurements of temperature profiles and determination of concentration profiles of C2H2OH, CO, CO2 , H2, H2O, O2, N2, CH4, C2H6 and C2H2+C2H4 by gas chromatographic analysis of samples withdrawn through fine quartz probes. Computational investigations also were made of profiles of oxides of nitrogen and other potential pollutants in similar partially premixed flames of ethanol and other fuels for comparison purposes. The computational results with the present mechanism are in reasonable agreement with experiment and perform as well as or better than predictions of other, generally much larger, mechanisms available in the literature. Further research is, however, warranted for providing additional and more stringent tests of the mechanism and its predictions, especially for condition at higher pressures. The second part of the dissertation consists of analytical study of autoignition of higher alkane fuels. It is shown that, above about 1000 K, ignition delay times for propane and all higher alkanes, as well as for a number of other fuels, can be calculated well by employing rate parameters of only three types of elementary steps, namely CmHn+HO2&;#8594;CmHn-1+H2O2, H2O2+M→2OH+M and 2HO2&;#8594;H2O2+O2, only the first of which is fuel- specific, the other two clearly being common to all fuels. The prediction of this remarkably simple result relies on a steady-state approximation for HO2, as well as steady states for more active radicals during induction. The resulting approximation to the chemistry exhibits a slow, finite-rate buildup of H2O2 and removal of fuel during the induction period. The criterion employed for termination of the induction period is the complete depletion of the original fuel subject to the approximations introduced. Numerical comparisons of the ignition-time formula with the experiments show that the predictions work well not only for higher alkanes but also for propene and JP-10. The analytical approximation thus produces reasonable results for a wide range of fuels. These results provide a new perspective on high-temperature autoignition chemistry and a general means of easily estimating ignition times of the large number of fuels of practical importanc