Abstract
The design of future jet engines has to meet the requirements of lower pollutant emission and higher efficiency. One of the main pollutants is soot particulates, whose emission negatively affects the global climate and human health. As a result, minimizing soot formation in jet engines can reduce both the economic and social costs of transportation. In addition to minimizing soot formation, the overall efficiency of jet engines can be improved by replacing the current metal super-alloy materials with ceramics matrix composites (CMCs), particularly silicon carbide matrix/silicon carbide fiber (SiC) composites. SiC composites offer similar, if not better, maximum service temperature to the nickel-based super-metal alloys, but only at one-third of the weight. Employing the composites will increase 1) the combustion efficiency with higher allowable flame temperature, 2) the engine thrust with smaller air flow diverted for cooling, and 3) the fuel mileage with lighter engines.
However, there exists a number of challenges that hinder the achievement of minimal soot formation and cost-effective production of SiC composites. Numerous research efforts have been dedicated to tackling these challenges. Such efforts have resulted in a large number of detailed chemical kinetic models whose purpose is to predict the soot formation and SiC composite fabrication processes in CFD simulations. Unfortunately, the models are still facing two major limitations: 1) significant uncertainties in reaction rate coefficients and 2) huge numbers of species and reactions that render the models impractical for complex CFD simulations. In order to overcome these two limitations, there exists a need for fundamental experimental data to 1) minimize the reaction rate coefficient uncertainties, and 2) reduce the model size while preserving their performance for practical CFD simulations.
Recognizing this need for experimental data, the Reacting Flow Laboratory (RFL) at the University of Virginia (UVa) has constructed and verified an excellent analytical system for chemical kinetic studies. The main component of the analytical system is a microflow tube reactor (MFTR), which can be paired with different diagnostic tools to quantify stable and radical gaseous species, and solid products during thermal and oxidative decomposition of hydrocarbon fuels or other chemicals of interest.
This thesis work employed the established analytical system to study thermal decomposition (pyrolysis) of three different hydrocarbon fuels: ethylene, n-dodecane and Jet A. The objectives were to investigate the effects of temperature, residence time and fuel chemical composition on the formation of soot precursors in the gas phase. Besides traditional chemical kinetic models, the experimental data were also used to test a new concept in modeling real jet fuels: the Hybrid Chemistry (HyChem) approach. This approach considers combustion of jet fuels as a two step process. The fuel initially undergoes thermal or oxidative decomposition to produce about ten products, which then get oxidized to form the final combustion products. Given that the HyChem approach represents the fuel chemical composition by the pyrolysis step, it needs to be thoroughly examined in order to ensure that the HyChem approach can accurately model the jet fuel combustion.
In addition to pyrolysis of hydrocarbon fuels, this thesis work also employed the analytical system to investigate species transport and chemical kinetics during silicon carbide (SiC) deposition from pyrolysis of methyltrichlorosilane/hydrogen (MTS/H2) mixtures. First of all, the functionality of the analytical system to study the SiC deposition process was demonstrated with a set of SiC deposition experiments on quartz substrate over a wide range of experimental conditions. Then, the analytical system was used to investigate the formation of SiC precursors in the gas phase during pyrolysis of MTS/H2 mixtures. With the measured speciation data, optimization and simplification of a detailed chemical kinetic model was carried out. The result was an improved model with less than half the number of species and reactions.
