Chemical Kinetics and Modeling of Combustion

Saturday, February 16, 2013
Room 202 (Hynes Convention Center)
William Green , Massachusetts Institute of Technology, Cambridge, MA
Historically, very simple combustion models have been used in engine simulations, in part because no accurate fuel chemistry models were available, and also because it would have been impossible to solve the simulations. In addition, often the fuel composition was unknown (due to analytical chemistry difficulties), and the fuel composition was not changing much, making it an uninteresting variable.  However, now fuels are changing, and there have been significant advances on all fronts in recent years, making it possible to use much higher fidelity models. This talk presents recent advances in modeling the chemistry of fuel combustion. Many different alternatives to conventional petroleum fuels have been proposed. However, most of these fuels have not been tested in engines, many are not available at the scale needed for engine testing, and in any case it is not clear which engines to use for the tests, since by the time these new fuels are available at the huge scale required to significantly affect the fuels market, engine technology will have evolved significantly. We propose that the best approach to this challenge is to develop accurate computer models for the fuel behavior, based on a combination of quantum chemical calculations and experimental data, which can be used to predict how the new fuel would behave in any engine condition of interest. This predictive chemistry approach is applied to the combustion of butanol and several other fuels, and the predictions are tested against a large set of experiments. Successes, failures, challenges and opportunities for predicting fuel effects in engines are discussed.