ITR: A New Data Model and Extensible Software for Predictive Chemical Kinetics

Massachusetts Institute Of Technology, Cambridge MA

Investigators

Abstract

Project Summary This project develops a completely new data model for chemical kinetics and corresponding extensible software, including graphical user interfaces and appropriate examples, that make it feasible to predict the rates and products of complex chemical processes. Many of the difficult tasks now performed manually by kineticists are automated, including the determination of which intermediates and reactions should be included in a simulation (based on numerical significance). In the new data model, parameters are required only for each type of functional group, rather than for each chemical species, and the xml format encourages users to document information on the provenance and uncertainty of each parameter. Because there are many fewer types of functional groups than chemical species, the new data model is compact enough that a human can check all the parameters individually. Furthermore, the parameters are classified in a scheme that encourages comparisons between related functional groups and the identification of outliers. To jump-start adoption of the new data model and software, a large, well-documented set of rate and thermochemical parameters is provided, as well as software appropriate for modeling a wide range of thermal and oxidative chemistry of hydrocarbons; new software tools for manipulating, comparing, and analyzing large-scale chemical kinetic simulations; and many examples. Chemical kinetic simulations of real-world processes are both extremely valuable, and extremely data intensive. For example, a model for gasoline combustion chemistry widely used to help design low-emission engines involves over 20,000 numerical parameters as inputs, and even this large set is certainly incomplete. The existing data model for chemical kinetics, created more than 20 years ago, was never designed to handle this level of complexity. It is adequate for conventional small a posteriori kinetic models, but not appropriate for predicting the chemical kinetics of technologically important systems, especially those involving complex mixtures. Accurate, large-scale simulation of reacting mixtures requires a whole new paradigm for how engineers, chemists, and environmental scientists interact with complex chemistry modeling software, and how they document the many assumptions and uncertainties that underlie the simulations. By reducing the size of the required input, and making the connection between the transferable fundamental chemistry and simulation results more transparent, this new approach will make a priori chemical kinetic simulation accessible to a much broader range of scientists and engineers. This will revolutionize the field of kinetics, open up significant new possibilities in the design of chemical products and processes, and provide a firmer basis for business and regulatory decision-making. A variety of outreach and educational activities are included. To encourage current practitioners of kinetics to take advantage of the new IT technology, the new methodology is demonstrated at conferences and promoted using existing Web collaboratories. It is also necessary to change the way kinetics is taught to encourage new engineers and scientists to embrace the new paradigm; to this end, new teaching modules are developed to introduce chemical kinetics to undergraduate and graduate students in a variety of disciplines including chemistry, chemical engineering, mechanical engineering, and environmental science.