Reduced ethanol skeleton mechanism for multi-dimensional engine simulation

Abstract

Ethanol is considered one of the most popular renewable biofuels worldwide. Detailed chemical mechanisms are generated to numerically study the oxidation kinetics of this fuel. Although detailed mechanisms are commonly used in 0D and 1D modeling (e.g., ignition delay time and laminar burning speed), their utilization in multi-dimensional modeling (e.g., 2D and 3D) is rather limited due to a large number of species/reactions which increase the computational time consequently. Therefore, a smaller and accurate chemical mechanism is essential for the proper implementation of fuel-oxidizer kinetics in multi-dimensional modeling. In this work, a detailed ethanol mechanism generated by authors, i.e., PCRL-Mech1 is used to create three sets of reduced skeleton mechanisms at different temperature ranges that enable predicting the results of detailed mechanism within reasonable accuracy. PCRL-Mech1 is a detailed ethanol mechanism with 67 species and 1016 reactions which shows a reasonable accuracy in predicting combustion parameters in a wide range of engine-relevant conditions. The mechanism reduction is performed using the path flux analysis method along with sensitivity analysis. Three sets of skeleton mechanisms including reduced-low, -high, and -wide are generated targeting the temperature range of 700–900 K, 1000–1700 K, and 700–1700 K, respectively. The reduced mechanisms are applicable for lean, stoichiometric, and rich conditions with pressure ranging between 1-80 atm. These three reduced mechanisms are successfully validated by comparing the prediction of ignition delay time, laminar burning speed, and species concentration with the detailed mechanism as well as available experimental data in the literature. The final three mechanisms of low, high, and wide temperature range include 27, 25, and 39 species and 147, 185, and 370 reactions, respectively. These three skeleton mechanisms are used in 3D ethanol spray combustion simulations which predict the temperature profile with an acceptable accuracy compared to the available experimental measurements.