Computer-Aided Construction of Chemical Kinetic Models

Green, William H.¹ 

[1] Massachusetts Inst. of Technology (MIT), Cambridge, MA (United States)

关键词: kinetic;    RMG;    reaction mechanism generation;    combustion simulations;   

DOI  :  10.2172/1167107
RP-ID  :  Final Report DE-FG02--98ER14914
PID  :  OSTI ID: 1167107

学科分类：数学（综合）

美国|英语

来源: SciTech Connect

【 摘 要 】

The combustion chemistry of even simple fuels can be extremely complex, involving hundreds or thousands of kinetically significant species. The most reasonable way to deal with this complexity is to use a computer not only to numerically solve the kinetic model, but also to construct the kinetic model in the first place. Because these large models contain so many numerical parameters (e.g. rate coefficients, thermochemistry) one never has sufficient data to uniquely determine them all experimentally. Instead one must work in ???predictive??? mode, using theoretical rather than experimental values for many of the numbers in the model, and as appropriate refining the most sensitive numbers through experiments. Predictive chemical kinetics is exactly what is needed for computer-aided design of combustion systems based on proposed alternative fuels, particularly for early assessment of the value and viability of proposed new fuels before those fuels are commercially available. This project was aimed at making accurate predictive chemical kinetics practical; this is a challenging goal which requires a range of science advances. The project spanned a wide range from quantum chemical calculations on individual molecules and elementary-step reactions, through the development of improved rate/thermo calculation procedures, the creation of algorithms and software for constructing and solving kinetic simulations, the invention of methods for model-reduction while maintaining error control, and finally comparisons with experiment. Many of the parameters in the models were derived from quantum chemistry calculations, and the models were compared with experimental data measured in our lab or in collaboration with others.