The Role of Computational Singular Perturbation (CSP)
in Chemical Kinetics

Harvey S. H. Lam
Princeton University, Princeton, NJ 08540


    Why does anyone want to do simplified chemical kinetics? There are (at least) two good rationales:

(1) A simplified kinetics model can provides valuable understanding and insights  to complex chemical reaction systems,
(2) The computational resource required is smaller when the number of dependent variables is made smaller. In addition, reduced chemistry models are usually less stiff.
    Traditionally, competent theoretician exploit their experience to "analytically derive" reduced chemistry models. In problems involving a flow field, it is obvious that in different regions of the flow field, the local reduced chemistry models are, in general, different. If rationale #1 is the primary issue, CSP can "computationally derive" local reduced chemistry models on the fly. However, the computational cost involved is inherently expensive.

    The central theme of this workshop appears to be: is there an inexpensive way to achieve rationale #2?

    I intend to talk about possible tradeoffs between rationales #1 and #2.

Selected Bibliography