A quasi-dimensional combustion model for spark ignition engines fueled with gasoline–methanol blends. (January 2018)

• Record Type: Journal Article
• Authors: Nguyen, Duc-Khanh; Sileghem, Louis; Verhelst, Sebastian
• Other Names: Pearson Richard guest-editor.; Gold Martin guest-editor.; Macey Brian guest-editor.; Turner James guest-editor.
• Is Part Of: Proceedings of the Institution of Mechanical Engineers. Volume 232:Number 1(2018:Jan.)
• Page Start: 57
• Page End: 74
• Publication Date: 2018-01
• Subjects: Methanol–gasoline blends -- alcohol -- laminar burning velocity correlation -- quasi-dimensional model -- spark ignition engines; Mechanical engineering -- Congresses; Transportation engineering -- Congresses; 629.2
• DOI: 10.1177/0954407017728161
• Languages: English
• ISSNs: 0954-4070

Abstract

The current work provides a quasi-dimensional model for the combustion of methanol–gasoline blends. New correlations for the laminar burning velocity of gasoline and methanol are developed and used together with a mixing rule to calculate the laminar burning velocity of the blends. Several factors (such as the laminar burning velocity, initial flame kernel, residual gas fraction, turbulence, etc.) have been investigated and the sensitivity of these factors and the used sub-models on the predictive performance was assessed. The simulation results were compared with measurement data from two engines on different gasoline–methanol blends. The results show the importance of the laminar burning velocity correlation, the method of initializing combustion and the turbulent burning velocity model. The newly developed laminar burning velocity correlation of gasoline performed equally or better than the existing correlations and the newly developed correlation of methanol outperformed the other correlations. The initial flame kernel size had a strong influence on the ignition delay. Changing the initial flame kernel to reproduce the same ignition delay was very effective to improve the simulations. Several turbulent combustion models were tested with the newly developed laminar burning velocity correlations and optimized ignition delay. In conclusion, the model of Bradley reproduced the trend going from gasoline to methanol much better than others due to the inclusion of the Lewis number.