Schlieren-based visualization and neural-network prediction of synthetic OME-1 fuel spray dynamics under elevated operating conditions

The growing demand for sustainable energy pathways has intensified research on low-carbon synthetic renewable (LCSR) fuels in internal combustion (IC) engines. In recent times, oxymethylene ether-1 (OME1) has emerged as a promising LCSR fuel and is primarily used as an additive due to its low octane and cetane numbers. As OME1 is better suited for modern IC engines with advanced combustion concepts, it is essential to ascertain its feasibility. Hence, the macroscopic spray characteristics of OME1–gasoline were investigated for the first time under elevated operating conditions. The Schlieren visualization technique was employed to examine spray evolution, and the captured images were subsequently processed in MATLAB to obtain quantitative data. In addition, a backpropagation neural network (BPNN) model was developed to predict the spray dynamics. The experimental results demonstrated that 100%-OME1 exhibited a narrower and faster spray penetration than gasoline, with a reduced liquid core region. With an increase in injection pressure from 100 to 400 bar, the spray penetration length, angle, and area were increased by 2.4 mm, 4.8°, and 286.8 mm2, respectively, at 1.5 ms ASOI. Further, the higher ambient temperature and lower boiling point of OME1 supported rapid fuel vaporization, increasing the spray area and reducing the average spray velocity. In BPNN modelling, the optimal network structure of 4 × 8 × 16 × 4 provided a reliable and robust prediction of spray characteristics, with an overall R2 and RMSE of 0.994 and 0.366, respectively. The obtained OME1 spray characteristics provide baseline data for the development of advanced engines.