Evaluation of autoignition models for production control of a spark-assisted compression ignition engine

Abstract

The drive to improve performance and efficiency of internal combustion engines has greatly expanded the degrees of freedom of engine systems. As efficiency objectives exceed the capability of traditional combustion strategies, advanced combustion modes are more attractive for production. These advanced combustion strategies typically add sensors, actuators, and degrees of freedom to the combustion process itself. Spark-assisted compression ignition is an efficient production-viable advanced combustion mode characterized by a spark-ignited flame propagation that triggers autoignition in the remaining unburned gas. This research focuses on autoignition modeling for spark-assisted compression ignition combustion phasing control. This work comprehensively evaluates several autoignition model structures and identifies the real-time production control implications of each. The candidate models include four ignition delay correlations, an ignition delay lookup, three polynomial regressions, and an artificial neural network. All are computationally feasible using production controllers, but the artificial neural network model represents autoignition phasing significantly better than the other options evaluated. The polynomial regressions were similar in error and exceeded the accuracy of ignition delay models. The low performance of the induction time integral–based models stems primarily from the exclusion of low-temperature heat release. The regression models are also exercised on an experimental engine dataset to identify the impact of engine phenomenon such as charge stratification on the performance of each model structure. The trends in the model performance as well as the magnitude of the error were similar when evaluated on both spark-assisted compression ignition simulation data and homogeneous charge compression ignition experimental data.