In this study, natural gas and diesel dual-fuel combustion under a medium load was numerically optimized for a heavy-duty engine using a genetic algorithm optimization approach. This approach employed a micro-genetic algorithm optimization code coupled with an engine computational fluid dynamics code to perform the optimization. Initially, an optimization using premixed natural gas with double direct injection of diesel fueling strategy was conducted using both the stock piston bowl shape and a bathtub shaped piston. Low emissions and moderate combustion with thermal efficiency close to 50% were achieved for both piston configurations by optimizing the premixed natural gas to diesel fuel ratio, exhaust gas recirculation fraction, diesel injection timing, injection pressure, and injection split ratio. Based on this optimum point, a parametric study was performed to understand the effects of the optimization parameters. The results show that high efficiency and clean combustion (indicated thermal efficiency ≥ 45%, NOx ≤ 0.4 g/kW h, peak pressure rise rate ≤ 15 bar/°) can be achieved over a wide range of parameters. The optimized result of a single diesel injection strategy was also evaluated and compared with the double injection strategy. The results indicate that the single injection strategy is also able to yield near 50% thermal efficiency and clean combustion. Compared to the double injection strategy, the single injection strategy shows increased combustion losses due to reduced diesel fuel near the cylinder centerline. In addition, the bathtub shaped piston has lower heat transfer losses relative to the stock piston, leading to improved fuel efficiency. However, the piston shape shows less impact than other parameters on the overall combustion and emission characteristics for both injection strategies.
