Multiple-objective optimization of direct dual fuel stratification (DDFS) combustion at different loads

Abstract

The operating parameters of the direct dual fuel stratification (DDFS) strategy in a heavy-duty engine are optimized across a full load range by utilizing a combination of three-dimensional computational fluid dynamics simulation and genetic algorithm. After obtaining the optimized results, sensitivity analyses were conducted on the operating parameters at various loads using the Pearson method. The results show that the DDFS strategy can attain stable and efficient combustion across the entire full-load range after optimization. At low-to-medium loads, the engine’s performance is predominantly influenced by initial operating parameters, while both initial and injection parameters play critical roles at high loads. The sensitivities of operating parameters increase as load increases, with the operating parameters having higher sensitivities having more concentrated distributions, while those with lower sensitivities have more dispersed distributions. The optimal conditions for low-to-medium load combustion generally involve a premixed-dominated combustion regime with some degree of reactivity stratification, which is strongly influenced by charge thermodynamics. Increasing the proportion of high-reactivity diesel fuel can improve combustion efficiency and stability, particularly under low-load conditions. Under high-load conditions, the optimal combustion strategy involves using a significant amount of direct-injected gasoline to achieve a more distinct stratified and diffusion combustion regime, which helps mitigate excessive heat release rates. However, this approach may result in reduced fuel economy compared to the optimal strategy for low-to-medium loads. As a consequence, the role of charge thermodynamics becomes less significant while the injection strategy becomes more critical for achieving optimal combustion at high loads.