Optical diagnostics-based numerical modeling of a compression-ignition optical research engine implementing ducted fuel injection

Ducted fuel injection (DFI) is a promising strategy to dramatically reduce engine-out soot emissions from compression-ignition (CI) engines. To accelerate the integration of DFI into real-world engines, computational fluid dynamics (CFD) is a powerful tool for its understanding and optimization. Although CFD has been largely adopted for investigating DFI in CI engines, accuracy and reliability of the numerical outcomes remain often questionable. Therefore, this article proposes a step forward in the CFD model development and validation process of a CI optical engine implementing DFI, considering both global and local perspectives. Indeed, leveraging the optical diagnostics dataset, the correct reproduction of the combustion and emissions formation processes – across a range of loads and oxygen levels for both CDC and DFI configurations – was tested not only in terms of typical aggregated values (i.e., pressure, burn rate, engine-out emissions) but also locally in terms of lift-off length, flame evolution, and soot natural luminosity (NL). For the latter a proper methodology was used to virtually replicate soot NL distribution starting from the CFD variables. Despite the complexity of this luminosity-based comparison, the model was found to reasonably replicate the main trends, for both CDC and DFI, thus highlighting predictive capability, not only in terms of engine-out soot emission values, but also of the in-cylinder soot formation/oxidation processes. Moreover, the present results demonstrate the robustness of the methodology to compute a virtual NL from CFD outputs, which could be of interest to thoroughly validate combustion models every time soot prediction accuracy plays an important role.