A GENERAL SEMICAUSAL STOCHASTIC-MODEL FOR TURBULENCE RADIATION INTERACTIONS IN FLAMES/

This paper presents a general two-dimensional non-stationary semicausa l model for the simulation of mixture fraction, which improves our pre vious causal model. The proposed model includes not only the pre-corre lation predictors (both in time space and geometric space) as well as the cross-correlation predictors, as in the causal model, but also pos t-correlation predictors. The latter makes possible the consideration of interactions of a scalar, such as mixture fraction, at a physical l ocation with that of all its adjacent locations. It has also been show n that the complicated second and higher-order correlation predictors can be neglected in the semicausal simulation of mixture fraction. To show the validity of the model, the stochastic mean and variance of th e spectral intensities at different wavelengths were predicted and com pared with derailed experimental data for turbulent carbon monoxide/hy drogen/air diffusion flames having different Reynolds numbers. These c omparisons showed excellent agreement with existing data and the impro vement over the prior causal model.