An investigation on methanol high pressure spray characteristics and their predictive models

This study leverages high-speed photography to explore how injection pressure, duration, ambient pressure, and temperature affect high-pressure methanol spray characteristics. Experiments across varied parameters reveal that increased injection pressure and duration augment spray tip penetration and cone angle. Ambient pressure rise diminishes penetration while enlarging the cone angle. Elevated temperatures reduce the liquid spray cone angle, stabilizing penetration length with minimal fluctuations. The Hiroyasu formula, adapted for diesel, was inaccurate for methanol spray tip penetration but was refined here to enhance accuracy. The Inagaki formula accurately predicted cone angles with a 1.31 % error. A CNN-LSTM-Attention model, with its 7 input and 2 convolutional layers followed by an LSTM layer, Attention layer, and output layer, predicts spray tip penetration with high precision, offering a determination coefficient of 0.99960 and an RMSE of 0.54 mm. These findings and models are instrumental for simulating methanol injection and optimizing engine combustion.