Modeling combustion chemistry of China aviation kerosene (RP-3) through the HyChem approach

Abstract

To address the complexity of modeling combustion chemistry of real multi-component fuels, the Hybrid Chemistry (HyChem) approach has been developed and tested for some typical jet fuels such as Jet A, JP-8, JP-10, etc. Still, the development and evolution of HyChem remain ongoing, and its potential has yet to be fully explored. The primary objective of the present study is to develop a HyChem model for describing the combustion chemistry of RP-3 while demonstrating the evolutionary understanding of the HyChem approach. In addition to the comprehensive new datasets provided by the present study as well as the development, validation, and reduction of an RP-3 HyChem model, several innovations were made regarding the HyChem development. Firstly, pyrolysis and oxidation experiments were performed in a flow reactor and utilized, sequentially, to constrain the coefficient parameters of the lumped reactions of the fuel decomposition submodel of HyChem. Meanwhile, ignition delay time and laminar flame speed measurement experiments were conducted in a shock tube and a constant-volume combustion bomb respectively, to obtain new datasets. Secondly, the species 1,3-butadiene was characterized as an additional critical intermediate during the RP-3 decomposition, in addition to these identified during the Jet A decompositions, and the RP-3 HyChem model was thus proposed to be revised to contain 1,3-butadiene. Thirdly, the present study demonstrated that by taking advantage of a flow reactor system equipped with GC/microGC or GC-MS that was able to characterize a complete kinetic picture of intermediate species distribution at the millisecond reaction time scale, a reliable HyChem model could be effectively constructed. Lastly, a newly developed machine-learning-based approach DeePMR, through iterative sampling, perturbation, and deep neural network (DNN)-guided screening, was shown to effectively achieve compact reduced models with state-of-the-art accuracy. In summary, the present study revealed substantial evolutionary understanding of the HyChem approach, which would greatly improve accessibility for researchers through the selection and application of different experimental apparatus and diagnostics to explore the HyChem approach and to develop proper HyChem models, for evaluating next-generation fuels and engine applications.