An ultrafast multi-zone HCCI model with Autoignition, Global reaction and Interpolation (AGI) for achieving comparable accuracy to detailed chemical kinetics models

Abstract

Homogeneous Charge Compression Ignition (HCCI) is a promising advanced combustion technology with high thermal efficiencies and low exhaust emissions, and thus, is a potential solution for future transportation applications. Quasi-dimensional multi-zone models with detailed chemical kinetics have reasonable computational cost and high accuracy for HCCI studies. However, they require significant tuning of the heat and mass transfers models against experiments, and solving large stiff equations for multi-zone chemical kinetics is still very time-consuming.In this study, a novel model is proposed. It is as accurate as conventional multi-zone chemical kinetic models, has much faster computational speed that is independent to the mechanism size, and require less tuning and calibration. Three main parts are included in this model: a control-mass Lagrangian (CML) framework, the Thermal Stratification Analysis (TSA) method, and the Autoignition, Global reaction, and Interpolation (AGI) model. The CML model framework avoids implementing interzonal mass and heat transfer correlations. The thermal stratification is realized by the TSA method. With it, case-by-case tuning is no longer needed. The combustion model is constructed by Autoignition, Global reaction, and Interpolation (AGI). A database of ignition delay times and burn rates based on constant-volume simulations is pre-generated. Then, the individual zonal ignition delay time and burn rate are interpolated from the database. A phenomenon called Burn Rate Equality ensures that the burn rate in the database equals the real engine simulation. This interpolation-based combustion model can run significantly faster than solving chemical kinetics at each time step, especially compared with a large-size chemical kinetic model and a large number of zones. The AGI model generally has errors less than 0.7 CAD of CA50 compared to the conventional chemical kinetics simulation. The performance boost is 100x to more than 10,000x depending on the size of the chemical kinetics model.