Degree centrality of combustion reaction networks for analysing and modelling combustion processes

Abstract

Listen

Translate article

Combustion research still needs more advanced fundamental understanding of combustion chemistry and dynamics from molecule scale to particle. The latter is also needed for soot and nanoparticles formation and combustion system control such as homogeneous charge compression ignition engine, and flame regimes and instability. The complex interactions between hundreds of species linked within thousands of reactions continue to be a challenge to analyse and model. The focus on this paper is to develop a method to facilitate the modelling and analysing of the detailed kinetics chemistry of fuels combustion. Through the use of combustion reaction networks (CRNs) analysis of degree centrality, principal species are identified during a combustion process by exploiting the introduced definition of principal species. A principal, central or the more active species of a combustion process, at a specific time step or cell of area/volume mesh, is the more tied up to other species in the CRN and so have the largest value of degree centrality. The accuracy of the dynamic identification of principal species, locally adapted to the thermochemical conditions at each time step/cell of the simulated combustion process, used by the employed directed relation graphs method of mechanisms reduction, is proved. The simulations were carried out using an adjusted dynamic adaptive chemistry approach of detailed chemistry implementing. It is demonstrated that an ‘active’ species in a combustion system would not necessary be considered as a part of important species set needed for its predictive simulations.