Assessment of an E10 gasoline surrogate: Qualitative and quantitative comparisons of in-cylinder spray morphology

Abstract

A minimum-component gasoline fuel surrogate that captures both chemical and physical behaviors of a full-distillate fuel is needed for high-fidelity CFD simulations. This study evaluates gasoline spray characteristics in a direct-injection spark-ignition engine under motored operation. Two fuels are compared; PACE-20, which is a 9-component surrogate formulation of RD5-87, is compared with its target fuel RD5-87, which is a full-boiling range research grade E10 gasoline. The spray morphologies of both fuels are recorded for a centrally-located direct-injection 8-hole spray subject to intake air cross-flow during the early part of the intake stroke. High-speed imaging recorded scattered light of the side and axial projections of the liquid spray. Quantitative metrics were developed and employed to facilitate comparison of spray morphologies as well as to identify the transition in spray morphology due to flash boiling. This paper builds on a previous study of RD5-87 where coolant temperature (20°C–100°C), in-cylinder pressure (40–110 kPa), engine speed (650–1950 rpm), and injection pressure (60–180 bar) were systematically changed to span operating conditions with and without flash boiling. Images of the PACE-20 morphology are selected for a sub-set of operating conditions from the previous study where distinctive morphology changes occurred. Visual inspection of the images and quantitative metrics demonstrate that the PACE-20 spray morphology is equivalent to that of the RD5-87 in most cases. The exception was for changes in the ambient-gas pressure where the flash-boiling transition occurred at ∼5 kPa higher in-cylinder pressure for PACE-20. Three empirical metrics, Merging Index, Asymmetry, and Flash Index are proposed here and they were found to be useful both as quantitative comparisons of the fuel morphologies, and for identifying the transition in spray morphology due to flash boiling.

In-cylinder spray evolution in a motored central-injection gasoline engine: Imaging and simulating the effects of flash-boiling and intake crossflow

“…The fuel was PACE-20, a nine-component surrogate for a fully blended E10 research gasoline, RD5-87. 29 Engine experiments demonstrated that the PACE-20 surrogate reproduced the same spray morphology as the RD5-87 over a broader range of conditions 30,31 than those studied in this work. The chemical composition of PACE-20 is shown in Table 3.…”

“…At the 50 kPa ambient pressure in these tests, the transition to flash boiling is observed to occur at 60°C in this engine and in quiescent vessels (cf. Figure 3 and existing literature 31 ). However, anomalous behavior is observed at the temperature of 40°C.…”

“…This is consistent with the experimental observations comparing the spray in this engine with morphology in quiescent chambers for both flash and non-flashing boiling conditions. 31 The pink and red areas in Figure 10 show the fuel-vapor concentration at regions of low and high values, respectively, at this plane. Thus, these mark regions where droplet convection and subsequent vaporization have occurred, used in this figure to define the plume locations and radial growth at this plane.…”

“…In this section, the best practices for the flash-boiling model parameters were determined at 650 rpm, where the spray morphologies are the same as those in quiescent vessels, implying the engine crossflow effects on the morphology are negligible. 30,31 Following this validation of the simulation approach, the dynamics of the spray collapse are discussed. The second part evaluates the sensitivity of the simulation flash-boiling morphology to the KH-RT model parameters and the plume cone angle in the injection model, which were adjusted to provide the best simulated spray morphology.…”

“…For the bottom view, the spatial resolution was 292 × 292 pixel and the field of view was 35 × 35 mm. Detailed descriptions of the engine, operating conditions, imaging systems, can be found in the RD5-87 experiments of Reuss et al 30 and the PACE-20 experiments of Kim et al 31…”