Reduced models represent a class of physics-based simplified kinetic models. The reduced number of species are systematically derived from a detailed mechanism, or more appropriately from a skeletal mechanism. There is no arbitrariness in the definition of the reaction rates. Reaction rates are still based on physically meaningful elementary reaction rates.
The key idea behind the development of a reduced reaction mechanism based on quasi-steady-state (QSS) approximation. Differential equations governing the evolution of the chemical species are reduced into much simpler algebraic relations by assuming that selected highly-reactive intermediate species are in QSS. The strength of reduced kinetic models is the ability to accurately capture complex combustion phenomena with a reduced set of chemical species since the details of the full starting mechanism (detailed or skeletal) are still indirectly retained.
In-Situ Adaptive Tabulation (ISAT) is a dynamic data storage/retrieval methodology for the evaluation of chemical kinetics based on table look-up and featuring adaptive error control. Within the context of a stiff chemistry solver, ISAT can significantly accelerate chemistry evaluations that are usually performed with computationally intensive Direct Integrations (DI).
A primary advantage of ISAT is that the operation does not require a foreknowledge of the evolution of the reacting flow field solution. Furthermore, the ISAT tables generated are of minimal size since the tables only contain the subset of thermo-chemical states accessed during a CFD simulation, as opposed to the entire realizable set of thermo-chemical states.
Dynamic Adaptive Chemistry
Adaptive chemistry with on-the-fly reduction is an innovative approach that bypasses the need to make assumptions by performing the reduction of the mechanism during the course of the CFD simulation based on instantaneous and local thermo-chemical conditions. The main advantage of dynamic reduction is that the active reaction pathways are identified automatically. Since the most computationally intense component of a CFD simulation with stiff chemistry is the time-integration of the reacting system, adaptive chemistry with on-the-fly reduction can determine the minimal yet sufficiently accurate reduced model for each chemistry valuation.
The key idea behind the concept of adaptive chemistry is that different regions of a flow field feature different levels of chemical activity (e.g., non-reacting far-field, after-burning base region, reacting shear layer, equilibrium plume region, etc.) and therefore it is not necessary to always use the most complex reaction mechanism to model the finite-rate chemical kinetics in all regions.