Computational optimization of a reactivity controlled compression ignition (RCCI) combustion system considering performance at multiple modes simultaneously

摘要

Reactivity controlled compression ignition (RCCI) combustion has been shown to yield improved performance over conventional diesel combustion (CDC) in terms of efficiency and emissions at mid load conditions. However, operation under high load and low load conditions is a major challenge with RCCI combustion. Past research has shown that high load operation is achievable by using low compression ratio (CR) pistons (e.g., CR similar to 12:1). However, using a lower compression ratio piston might affect the operation at the low load conditions due to the long chemistry time scales corresponding to these loads. This shows that the optimum compression ratio could be different when both loads are taken into consideration. In addition, the optimal injector configuration, bowl geometry and air handling could all be very different considering the large difference in fuel mass associated with the low and high load operating conditions. Accordingly, the current analysis discusses the results from a computational optimization study that was conducted considering the performance at low load-high speed (2 bar, 1800 rev/min) and high load-low speed (20 bar, 1300 rev/min) operating conditions simultaneously. Detailed computational fluid dynamics (CFD) modeling in combination with a genetic algorithm (GA) was used to conduct a thorough optimization that considers 28 design inputs which includes parameters for bowl geometry, injector design, air handling and fueling strategy. The inputs were varied such that at any point during the optimization, the bowl geometry and injector design parameters would remain the same for the two operating modes while the air handling and fueling strategy inputs, which could be varied with load, were allowed to be different. The objective of the optimization study was to maximize the average of the efficiencies from the two operating conditions (i.e. an equal weighting was given to the two operating modes) considered for the study. The effect of giving a higher weighting to one operating condition over the other was also investigated by using a response surface model (RSM) that was built from the GA data of the two operating modes, using non-parametric regression techniques.