ABSTRACT
Observations of intermediate species in simple flames that can be readily modeled have found wide use as a test of the accuracy of our knowledge of combustion chemistry. This validation effort, when agreement is achieved, should then lend confidence in the predictability of advanced computational tools applied to model more complex real systems. The chemical complexity of even these simpler flame experiments indeed requires such comparisons, because the accumulated effects of uncertainties in the elementary rate con stants of the mechanism plus the occasional larger error or omitted pathway can lead to significant model disagreement. In a complex way there is data of kinetic mechanistic value in many of these observations, which for exam ple has been used in the systematic optimization procedure that produced the GRI-Mech kinetic mechanism [1,2]. The error limits on individual reac tion kinetics measurements allow a set of permissible mechanisms, many of which will provide poor predictions. The optimization is designed to select the most favorable.
