Multiple-Objective Optimization Of Direct Dual Fuel Stratification (Ddfs) Combustion at Different Loads

Abstract

Operating parameters of a heavy-duty engine run with the direct dual fuel stratification (DDFS) strategy are optimized over a full-load range using a combination of three-dimensional computational fluid dynamics simulation and genetic algorithm. Based on the optimized results, sensitivity analyses of operating parameters at different loads were conducted based on the Pearson method. The results indicate that DDFS strategy could achieve efficient and stable combustion in a full-load range after optimization. Initial operating parameters have dominated influences on engine performance at low-to-medium loads, while both of initial and injection parameters play important roles at high loads. Sensitivities of operating parameters increase with increasing of load, and the operating parameters with higher sensitivities have more concentrated distributions, while those with lower sensitivities have more dispersed distributions. At low-to-medium loads, the optimal cases essentially belong to the premixed-dominated combustion with somewhat reactivity stratification, which is sensitive to charge thermodynamics. By increasing the amount of diesel fuel with high reactivity, combustion efficiency and stability can be enhanced, especially at low loads. At high loads, a large amount of gasoline by direct-injection is employed in the optimal cases to create more notable stratified and diffusion combustion to prevent massive heat release rate with a penalty of certain fuel economy compared to the optimal cases at low-to-medium loads. Thus, the sensitivity of charge thermodynamics reduces while the effect of injection strategy increases at high loads.

Keywords: Direct dual fuel stratification (DDFS), Computational Fluid Dynamics, Genetic algorithm, Low-temperature combustion