A reduced-order model for turbulent reactive sprays in compression ignition engines

• Published in: Combustion and flame Vol. 236; p. 111751
• Main Authors: Deshmukh, Abhishek Y., Davidovic, Marco, Grenga, Temistocle, Lakshmanan, Raghavan, Cai, Liming, Pitsch, Heinz
• Format: Journal Article
• Language: English
• Published: New York Elsevier Inc 01-02-2022; Elsevier BV
• Subjects: Carts; Combustion; Computational efficiency; Computing costs; Delay time; Engines; Feedback control; Flame lift-off; Ignition; Reactive spray; Reduced order models; Reduced-order model; Soot; Turbulence; Turbulence models; Turbulent flow; Vapor phases; Turbulence; Reactive spray; Flame lift-off; Soot; Reduced-order model; Ignition
• ISSN: 0010-2180; 1556-2921
• DOI: 10.1016/j.combustflame.2021.111751

In this work, we present a reduced-order model for transient turbulent reactive sprays in compression ignition engines. The model is an extension of the previously developed cross-sectionally averaged spray model (CAS) (Deshmukh et al., 2021). A transient turbulence model is derived for the gas phase that allows to model the transient scalar dissipation rate, which is a key parameter in non-premixed combustion. Representative chemistry is solved in mixture fraction space interactively with the flow. The turbulent CAS model combined with the combustion model is termed the cross-sectionally averaged reactive turbulent spray (CARTS) model and can represent unsteady non-premixed combustion behavior. The CARTS model is validated against experimental data as well as three-dimensional (3D) numerical simulations. The model is able to reasonably predict the trends of ignition delay time, flame lift-off length, and soot emissions. The computational cost of the CARTS model is orders of magnitude lower than the 3D simulation methods. This low computational cost enables various applications, including but not limited to the rapid screening of novel fuel candidates as well as off-line training of models to be eventually used in closed-loop control.