A numerical–experimental assessment of wall impingement models for spark-ignition direct-injection engines

Abstract

Liquid wall impingement in direct-injection engines can cause soot and hydrocarbon emissions as well as reduced combustion efficiency. This study focuses on detailed evaluations of numerical droplet impingement criteria that govern the onset of splash. The five selected splash criteria, which all extrapolate from single-droplet impacts to full sprays, are representative of those currently in use for spark-ignition direct-injection engines. The computations examined the sensitivity of impinging spray simulations to the splash criteria for a high-pressure, direct-injection swirl spray under atmospheric conditions impinging at a 45° angle onto a flat plate. The numerical results were compared to an unusually extensive set of experimental data: Mie scattering and light transmission imaging, plus quantitative refractive index matching measurements of the fuel film area and thickness, and phase Doppler interferometry measurements of droplet size and velocity near the plate. Good qualitative and at least fair quantitative agreement was obtained for the global spray impingement and wall film formation, especially for single-drop criteria that include the effect of viscosity. The film area and shape were insensitive to the splash criteria, illustrating the importance of film thickness measurements for validating simulations. The results also revealed the sensitivity of impingement calculations to droplet arrival frequency when that is taken into account. In general, the comparisons indicated the need to capture the effect of multiple droplets impinging on the wall at irregular frequencies in the criterion, as well as other important physics of the droplet–wall interaction that may mask the true effect of the impingement criterion.

Spray-wall interactions in direct-injection engines: An introductory overview

“…Encouragingly, the models show good agreement with the broad range of measurements, with the only—but important—exception being the amount of liquid mass deposited, as shown by the RIM data. Even for this liquid mass metric, the splash criteria of Grover et al 3 and Cossali et al 42 come within 14% of the experimental value, which is reasonably close. The overall good performance of the splash criteria presents a strong case for future work being targeted at improving these models, particularly with regards to the issue of fractional mass deposited.…”

“…To visualize and measure the formation and evaporation of the fuel films on the piston crown, they developed a high-speed imaging technique based on refractive-index-matching (RIM). 30 This technique, which exploits the change in light scattering (reflection 4,9,12 or transmission 3 ) from a rough surface when it is wetted by a liquid, has also been used in three papers in this special issue. 3,4,9 Results under warmed-up, part-load stratified conditions showed that—in contrast to some earlier expectations—the fuel-film mass correlated quantitatively with engine-out soot mass, whereas the amount of fuel on the piston was small (typically ∼0.1−1%) and did not contribute significantly to HC emissions.…”

“…Three papers in this special issue focus on modeling and high-fidelity simulations of SWI. 1–3 The most fundamental is the work of Markt et al, 1 whose contribution characterizes the secondary droplet population created by the impingement of a train of ethanol droplets. With the use of detailed VoF simulations and employing their own post-processing algorithm, they provide a quantitative and qualitative assessment of the evolution of secondary droplet statistics, including visually striking illustrations of instability-driven breakup of lamellae.…”

“…The paper by Grover et al 3 evaluates the performance of five commonly used droplet-splash criteria (including the SR1996 model) that are implemented into the RANS code GMTEC. 44 The results are validated with Mie -scattering and light -transmission images, RIM measurements of liquid wall-film thickness and mass, and phase-Doppler–interferometry (PDI) measurements of the size and velocity distributions of the impinging and secondary droplet population.…”

“…This special issue of International Journal of Engine Research (IJER) presents 11 papers that investigate spray-wall interactions (SWIs) and their effects at both fundamental and practical levels. These studies include numerical simulations 1–3 and measurements 4–8 of SWIs and fuel-film formation and dynamics, as well as measurements of fuel-film combustion and emissions processes. 9–11…”