An experimental and modeling study on norbornane pyrolysis aided by chemical information from neural network-assisted molecular dynamics

H Xiao and ZH Chu and HD Chen and TC Zhang and JH Lu and CY Wang and L Zhao and B Yang, COMBUSTION AND FLAME, 274, 114039 (2025).

DOI: 10.1016/j.combustflame.2025.114039

Norbornane has been reported in recent years as a diesel additive that can improve soot characteristics or as the backbone of new jet fuels to increase fuel density and the net heat of combustion. However, there is still a lack of experimental and modeling studies on norbornane pyrolysis, which limits its further application. This work uses a flow tube reactor to conduct experiments at 30 torr, 923 K similar to 1373 K, with a photoionization molecular-beam mass spectrometer to identify and quantify the pyrolysis species. At the same time, an attempt is made to apply high-precision deep potential molecular dynamics (DPMD) to assist in kinetic model construction. Species analysis on MD simulations provides additional chemical information on key pyrolysis species in the norbornane pyrolysis system, which agrees with experimental results. Therefore, the reactions appearing in MD simulations are supposed to play a nonnegligible role in the experimental system, so that high- frequency reactions of selected are added to the pyrolysis kinetic model. Further comparison of experimental and modeling results shows that the modeling results can accurately predict experimental concentrations for most species. By analyzing the rate of production of all species in the system, we highlight their primary production and consumption pathways, especially the pathways from norbornane to benzene in the system. When conducting sensitivity analysis, it is found that the initial decomposition reactions of the fuel have large sensitivity coefficients on the experimental results, especially the concentration of norbornane; it is also noteworthy that the reactions extracted in the MD results have an essential impact on the concentration of 1,3-cyclohexadiene.

Return to Publications page