Key points are not available for this paper at this time.
Gas sampling methods have been crucial for the advancement of combustion science, enabling analysis of reaction kinetics and pollutant formation. However, the measured composition can deviate from the true ones because of the potential residual reactions in the sampling probes. This study formulates the composition estimation in stiff chemically reactive systems as a Bayesian inference problem, solved using the No-U-Turn Sampler (NUTS), and systematically investigates the causes of information loss for the initial value inference. Theoretical analysis shows that information loss arises from the restriction of system dynamics by low-dimensional attracting manifold, where constrained evolution causes initial perturbations to decay in fast eigen-directions in composition space. The methodological framework is demonstrated in the Robertson system and autoignition of hydrogen/air mixtures. Furthermore, a gas sample collected from a one-dimensional hydrogen diffusion flame is analyzed to investigate the effectiveness of the frozen temperature for alleviating information loss. The research highlights the importance of species covariance information from observations in improving estimation accuracy and identifies the intrinsic connection between inference failures and the rank reduction in the sensitivity matrix representing the system dynamics. Critical failure times for species inference are defined and analyzed in the Robertson and hydrogen autoignition systems are analyzed, providing insights into the limits of inference reliability and its physical significance.
Lu et al. (Fri,) studied this question.