Dynamic Adaptive Chemistry applied to homogeneous and partially stratified charge CI ethanol engines

The Dynamic Adaptive Chemistry (DAC) technique is extended in this work to multidimensional simulations of ethanol HCCI/PSCCI engines. Several DAC computations have been performed by using two kinetic reaction mechanisms of ethanol with different levels of detail, that include 57 species and 135 species, respectively. The specific choice of the DAC parameters, i.e. the set of search-initiating species and the tolerance value, has been carefully analyzed. The simulations show that very accurate results, in terms of pressure and heat release rate profiles and CO, CO2 and UHC emissions, are obtained with ethanol as the only species for the graph search both with the fuel uniformly distributed and by directly injecting liquid fuel in the combustion chamber. As regards NOx, specific attention has been addressed to the analysis of the NOx formation in order to correctly reproduce the paths that lead to NOx emissions for the different cases. The choice of ethanol-N2O as search-initiating set has given the best results with negligible errors with respect to the full mechanism. For the single-zone computations, the use of DAC provides a speed-up of the 135-species full mechanism more than 9, whereas, with respect to the 57-species mechanism, about 50% of the computational time is saved in the multidimensional simulations.