Experimental and numerical investigation of iC8H18 laminar combustion characteristics based on hydrogen – water complementary regulation

Abstract

The present study examines using water dilution and hydrogen as control measures to regulate the laminar burning velocity of iso-octane/air mixtures in a constant volume combustion bomb. Experimental results are obtained by processing schlieren photographs and simulation results are calculated by Chemkin software. The water dilution rates of the mixtures were between 0 and 5% and the hydrogen addition ratio was between 5 and 55%. The initial pressure was set at 101.3 kPa with the preheating temperature ranging from 400 K to 500 K. According to the results, laminar burning velocity varied almost linearly with the dilution of water and the addition of hydrogen, respectively. This study also distinguishes the thermal and kinetic effects of water on the combustion processes. The thermodynamic effect of water vapour on laminar burning velocity was higher than the kinetic effect. In kinetics, O, H, and OH were the most affected free radicals by the addition of water vapour. The main free radical change affected the rate of the elementary reaction. Besides, the kinetic effect of water vapour promoted combustion. However, this effect was offset by the thermodynamic effect. The introduced hydrogens of premixed mixtures did not lead to measurable improvements in laminar burning velocity. Additionally, the change in the reaction rate of the elementary reaction was not obvious, which was inconsistent with the change of laminar burning velocity. Lastly, a laminar burning velocity model using hydrogen and water vapor content as control parameters was proposed. Accordingly, the designated laminar burning velocity of the iso-octane/air mixture was achieved by altering hydrogen and water vapour ratios in the mixture. BP (Back-Propagation) neural network algorithm (a multi-layer feed forward neural network trained by error back propagation algorithm) and empirical data were used to verify the model, and the model’s maximum error was less than 5%.