Integrating Laminar Burning Velocity Into an Empirical Kinetic Model for Predicting the Ignition Delay of Biodiesel Blends

Abstract

Abstract. Accurate prediction of ignition delay time is fundamental for optimizing combustion phasing in compression ignition engines. Classical empirical models often fail to capture the complex auto-ignition behavior of oxygenated fuels like biodiesel due to the omission of flame propagation characteristics. This study proposes an empirical kinetic model that integrates the maximum laminar burning velocity and cetane number to represent the synergy between physical transport and chemical kinetics. The model was developed using a robust experimental dataset derived from a Cooperative Fuel Research engine and a constant-volume bomb. Five primary parameters—cetane number, maximum laminar burning velocity, in-cylinder pressure at the start of injection, in-cylinder temperature at the start of injection, and equivalence ratio—were incorporated into a power-law formulation. To ensure predictive capability and avoid numerical overfitting, the 330 experimental data points were randomly divided into an 80% training set (264 points) for model calibration and a 20% independent testing set (66 points) for validation. The evaluation demonstrates stable accuracy across both datasets, with the independent testing phase achieving a mean absolute percentage error between 4.23% and 7.34% and a coefficient of determination reaching 0.922 for specific biodiesel blends. The integration of maximum laminar burning velocity enhances the model's sensitivity to the thermodynamic state, providing a reliable kinetic-based tool for advanced combustion simulations of sustainable fuels.