A Sequential Chemical Kinetics-CFD-Chemical Kinetics Methodology to Predict HCCI Combustion and Main Emissions 2012-01-1119

This study presents the development of a new HCCI simulation methodology. The proposed method is based on the sequential coupling of CFD analysis prior to autoignition, followed by multi-zone chemical kinetics analysis of the combustion process during the closed valve period. The methodology is divided into three steps: 1) a 1-zone chemical kinetic model (Chemkin Pro) is used to determine either the intake conditions at IVC to achieve a desired ignition timing or the ignition timing corresponding with given IVC conditions, 2) the ignition timing and IVC conditions are used as input parameters in a CFD model (Fluent 6.3) to calculate the charge temperature profile and mass distribution prior to autoignition, and 3) the temperature profile and mass distribution are fed into a multi-zone chemical kinetic model (Chemkin Pro) to determine the main combustion characteristics.

Different numbers of zones have been tested in the multi-zone step to determine the effectiveness of the general methodology. 40 zones are needed to achieve acceptable thermal stratification resolution to accurately predict peak heat release rates, peak pressures rise rates and ringing intensity. However, a simplified 12-zone reduced model is developed and validated to study combustion variables. Simulation results for the main combustion variables and emissions are compared with experimental results from a multi-cylinder HCCI engine fueled with biogas (60% CH₄ + 40% CO₂), and operating at different intake conditions. Comparisons between the proposed numerical methodology and experimental results show good agreement for power output (measured as IMEPg), indicated efficiency, burn duration, peak pressure, individual ringing intensity, and HC and NO_x emissions. CO emissions are very sensitive to the input parameters of the 12-zone reduced model. However, when the peak temperature after ignition of boundary layer zones is properly handled; CO emissions are reasonably well predicted. According to the results, the methodology can successfully predict combustion parameters and emissions for HCCI engines in which the fuel and air are well mixed prior to ignition. Compared with previous sequential methodologies, the method proposed here allows for reduced number of zones, more uniform temperature profiles prior to ignition, more accurate estimation of mass located in each zone, reduced computing time and more accurate predictions of peak heat release rates, peak pressure rise rates, and ringing intensity.