화학공학소재연구정보센터
Combustion and Flame, Vol.122, No.3, 233-252, 2000
A turbulent reaction rate model for premixed turbulent combustion in spark-ignition engines
A new reaction rate model has been developed and validated for premixed turbulent combustion in spark-ignition (SI) engines. The formulation is within the context of the Bray-Moss-Libby (BML) formalism for turbulent transport in premixed flames. The laminar flamelet concept is employed to split the volumetric reaction rate into a reaction rate per unit surface area and a surface area per unit volume. The development has involved the implementation of an algebraic functional dependence of the laminar burning velocity on the reaction mixture strength, temperature, and pressure [1]. Also, the flame surface area per unit volume is modeled in a novel and physically reasonable manner that avoids any explicit dependence on local turbulent length and time scales, by modeling the laminar flamelet wrinkling length scale as a function of laminar flame properties as well as turbulent quantities. The resulting expression for the mean turbulent reaction rate is implemented in a computer code together with the BML second moment model for turbulent transport. Second-order accurate bounded spatial discretization is employed. The governing equations have been transformed into a moving coordinate system to take into account the piston motion. A feasibility study is carried out on the application of the new model to computation of flame propagation in SI engines. The empirical constants of the new modification of the model have been tuned and evaluated by capturing experimental engine cylinder pressure histories. A comparison between the present calculations and identical computations of engine cylinder pressure demonstrates the validity of applying the model in SI engine computations. Also, and unlike the standard Eddy Break-up (EBU) and basic flamelet models, the new model displays no tendency to produce excessive reaction rates in the presence of solid walls. A parametric study shows the model to behave in a satisfactory manner in response to changes in fuel type, equivalence ratio, ignition timing, compression ratio, and engine speed.