Predictive modeling of combustion processes
Author(s)
Sharma, Sandeep, Ph. D. Massachusetts Institute of Technology
DownloadFull printable version (12.13Mb)
Other Contributors
Massachusetts Institute of Technology. Dept. of Chemical Engineering.
Advisor
William H. Green.
Terms of use
Metadata
Show full item recordAbstract
Recently, there has been an increasing interest in improving the efficiency and lowering the emissions from operating combustors, e.g. internal combustion (IC) engines and gas turbines. Different fuels, additives etc. are used in these combustors to try to find the optimal operating conditions and fuel combination which gives the best results. This process is ad-hoc and costly, and the expertise gained on one system cannot easily be transfered to other situations. To improve this process a more fundamental understanding of chemistry and physical processes is required. The fundamental constants like rate coefficients of elementary reactions are readily transferable enabling us to use results from one set of experiments or calculations in a different situation. In our group we have taken this approach and developed the software Reaction Mechanism Generator (RMG), which generates chemical mechanism for oxidation and pyrolysis of a given fuel under a set of user-defined physical conditions. RMG uses group additivity values to generate thermochemistry of molecules and has a database of rate coefficients of elementary reactions. These two sets of data are used to generate chemical kinetic mechanism in a systematic manner. The reaction mechanisms generated by RMG are purely predictive and elementary rate coefficient from any reliable source can be added to RMG database to improve the quality of its predictions. The goal of my thesis was two fold, first to extend the capabilities and database of RMG and to release it as an open source software for the chemical kinetic community to use. (cont.) The second was to take a practical system of interest and use RMG to generate the chemical mechanism and thereby demonstrate the utility of RMG in generating predictive chemical mechanisms for practical situations. As a part of the second step our hope was to generate new chemical insights into soot formation processes which are of great interest. The three most important contributions of the thesis are listed below. 1. My work with RMG has resulted in order of magnitude improvements in the cpu and memory usage of RMG and it has added many useful features to RMG like ac- curate sensitivity analysis for better interpreting the final mechanism. I have also worked on extending the database of RMG, by adding thermochemistry of ringed species that cannot be treated adequately by group additivity. Also kinetic rate rules for intramolecular-H-migration reactions in OOQOOH molecules were added to RMG database, which are important in predicting the low temperature oxidation of alkanes. 2. Recently there have been considerable advances in the methodology for rate coefficient calculations for loose transition states, i.e transition states that are not saddle points. These type of transition states are encountered often in radical-radical reactions. In addition to these advances there has been significant progress in accurate calculation of the pressure dependent rate coefficients for complicated potential energy surfaces with multiple wells and multiple product channels. The method is based on the master equation formulation of the problem. These detailed equations are then appropriately coarse-grained to calculate the phenomenological rate coefficients. (cont.) I have used these state of the art techniques to calculate the rate coefficients for the formation of various aromatic species like benzene and styrene. The rate coefficients predicted by these methods were tested under certain conditions and are in good agreement with experimental data. 3. Finally to model a two-dimensional diffusion flame we have developed a solver that is able to solve a complicated set of highly coupled differential equations in an efficient manner to give accurate results. The solver in conjunction with chemistry that is developed using techniques mentioned in the last two points is used to solve the mole fraction profiles in the diffusion flame. The results of the simulations are compared to the experimental measurements and this process gives us insight into soot formation in diffusion flames.
Description
Thesis (Ph. D.)--Massachusetts Institute of Technology, Dept. of Chemical Engineering, 2009. Cataloged from PDF version of thesis. Includes bibliographical references (p. 161-169).
Date issued
2009Department
Massachusetts Institute of Technology. Department of Chemical EngineeringPublisher
Massachusetts Institute of Technology
Keywords
Chemical Engineering.