Title: Optimization of a Direct Injection Strategy using Fully Integrated CFD Design Exploration
This work presents an automated fully integrated process to achieve optimal turbulent kinetic energy and fuel uniformity at spark timing, in the combustion system of a GDI engine. The adopted hybrid search algorithm, SHERPA, is applied to high fidelity multi-dimensional and transient engine CFD simulations, encompassing the scavenging event of air intake into the cylinder combined with fuel spray injection. The design space is explored for feasible solutions, using fully parameterized injection events for a 6 hole nozzle injector, while meeting stringent constraints to reduce the total amount of non-vaporized fuel, fuel residing in the intake port and fuel charge residing outside the flammability limits, with minimal user intervention.
The results are indicative of a distinct trade-off between the turbulent kinetic energy and fuel homogeneity, explained by the spray and air interaction that boosts the turbulence levels while simultaneously augmenting the distribution of the fuel uniformity. The infeasible solutions are primarily bounded by the constraints, which are put upon the search algorithm in the design exploration. Despite the strict requirements on valid solutions, several feasible designs are found using the proposed method. In comparison with a baseline strategy, two particular strategies stand out; one where the intake port fuel and liquid fuel present at spark timing are reduced by 80% and 60% respectively, at a slight cost of reduced turbulent kinetic energy, and one improving vaporized fuel present in the engine, levels of turbulent kinetic energy and fuel homogeneity, all simultaneously, at acceptable penalty of fuel back flow to the intake port.