Response maps for chemical non-equilibrium reactions

In many applications (e.g. gas turbines, internal combustion engines), the gas-dynamic and chemical energy release mechanisms have comparable time scales, so that equilibrium chemistry is inadequate for predicting their interaction. The resulting requirement to couple non-equilibrium chemistry, fluid-dynamic, and initial- and boundary-condition equations results in large sets of numerically stiff equations; and their time integration demands enormous computational resources. The size and stiffness of these equation sets places practical limits on high level analyses such as computational fluid dynamics (CFD). Specifically, to make such analyses computationally affordable requires great simplification of either geometric resolution or the accuracy of the chemical reaction mechanism.; This dissertation presents the hierarchical development of response maps (in order of increasing information description, complexity and accuracy) that: (a) reduce the computational intensity of non-equilibrium chemistry by orders of magnitude; and (b) remove the stiffness of the equations while maintaining the non-equilibrium chemistry information. The main application examined is detonation of premixed stoichiometric propane-oxygen and propane-air mixtures. The response maps are developed for a range of impulsively compressed mixtures. Their behavior is examined and then tested for systems which are impulsively compressed and then react at either constant volume or constant pressure. Accuracies in predicting key parameters, generally within 3%, are achieved with computational speeds approximately two orders of magnitude faster than the detailed integration.