Large Eddy Simulation of Diesel Spray Flames using the Unsteady Flamelet Progress Variable Approach with Soot Prediction

Abstract

Large scale simulations of realistic engine geometries with detailed chemistry and species transport are prohibited by the large size and the stiffness that accompanies detailed reaction mechanisms. Flamelet-based models are frequently used in spray combustion to provide a cost-effective approach to understand combustion and emissions characteristics of liquid sprays while accounting for turbulence chemistry interaction. In this study, a large eddy simulation (LES) approach utilizing the unsteady flamelet progress variable (UFPV) method is assessed for the Engine Combustion Network (ECN) Spray A configuration. The LES-UFPV framework is validated against measured ignition delay and flame liftoff at different ambient conditions. Good agreement between the numerical predictions and experimental measurements was achieved. The validated LES-UFPV model is then coupled with an empirical one-equation soot model, the Hiroyasu model, to investigate soot characteristics. While the Hiroyasu soot model qualitatively predicts the total soot mass, its performance in predicting the onset of soot inception is unsatisfactory. These shortcomings encourage further investigations using more advanced soot models for more accurate soot prediction.