Some aspects of the evolution in the knowledge about the PAH kinetics in combustion systems have been discussed. A simplified description of the improvements from early reaction models of PAH species formation and growth up to the newest models is given. More accurate determination of thermochemcial data for each species considered in the model, more accurate rate coefficients, and inclusion of relevant reactions that have not been considered yet are the ways for further improvements of kinetic models. The accurate determination of rate coefficients includes the elucidation of their pressure dependence, as PAH growth and soot formation are strongly dependent on pressure; skillful efforts have been recently devoted to this field, including RRKM and QRRK approaches on elementary reactions.

Numerical simulation of supersonic combustion of hydrogen in air has been done using point implicit finite volume method. This method treats all chemical species terms implicitly and all other terms explicitly. Solver is based on the solution of unsteady, compressible, turbulent Navier-Stokes equations, using Unstructured Finite Volume Method (UFVM) incorporating RNG based κ-ε two equation model and time integration using three stage Runge-Kutta method. Reaction of hydrogen with air is modeled using an eight-step reaction mechanism. The preconditioning has found to be effective in overcoming the stiffness in chemically reacting flows. The method is validated against standard experiments for CFD code validation. The predicted values of temperature and species production were in good agreement with experimental results. The code is used to simulate the combustion of hydrogen injected to the wake region formed by a wedge shaped strut.

The flame structure at small Damkohler number and large Karlovitz number is still unknown. In the present study, the reaction zone structure in a well-stirred reactor was investigated extensively. The OH-PLIF images suggested that there are no thin laminar flamelets and that the reacting eddies are distributed throughout the reactor. With the cross-correlation of ion signals, the scale of the reacting eddies was determined to be of the order of 4mm, and the convection velocity of these eddies or zones was found to coincide with the mean flow velocity of the order of 100m/s. In this combustion regime, NOx concentration was extremely low and this is very attractive for the practical use to meet the environmental requirements.

Combinations of optical interference band-pass filter and photomultiplier are often used for detecting OH^* and CH^* chemiluminescences in examining the correlation between chemiluminescent intensities and the combustion quantities, such as equivalence ratio and temperature. Since we consider that the influence of continuous background emission on chemiluminescent emission intensities should not be ignored, we propose a technique for removing those influences on OH^* and CH^* chemiluminescent emission intensities measured by the combination of optical band-pass filters and photomultipliers. Moreover, the correlation between the actual chemiluminescent emission intensities without the influence of continuous background emission and the physical quantities has been investigated in the present study. By separating chemiluminescent emission intensities of OH^* and CH^* from continuous background emission intensities, it becomes possible to measure equivalence ratio using chemiluminescent emission intensities at any positions of a flame. This technique holds wider applicability than the conventional methods.