Direct numerical simulation of an atomizing biodiesel jet: Impact of fuel properties on atomization characteristics

Abstract

The utilization of biodiesel is an effective approach to reduce pollution from internal combustion engines and thus has attracted steadily increasing interest in the recent years. As the viscosity of biodiesel is much higher than that of standard diesel, the atomization characteristics of a biodiesel jet can significantly deviate from those of a standard diesel jet under identical injection conditions. Since atomization of the injected fuel has a strong impact on fuel-air mixing and the following combustion processes, it is important to investigate the atomization of biodiesel and in particular to understand how the fuel properties affect the atomization process and the resulting spray characteristics. In the present study, three-dimensional direct numerical simulations are conducted to investigate atomizing biodiesel and diesel jets. The novel adaptive multiphase solver Basilisk is used for simulations. The statistics of droplets formed in the biodiesel jet is compared to the diesel jet under identical injection conditions.

Mesh Resolution Effects on Primary Atomization Simulations

“…This problem is solved using AMR with three grid refinement levels and the same refinement criteria. The physical properties of each phase are based on low-velocity diesel injection, following [5], using a pulsating boundary condition for the inlet, as done by [16,17]. This configuration has a narrow length-scale range and a simple deterministic injection velocity profile independent of grid refinement.…”

“…This periodic perturbation has been employed in the works of [16] to accelerate the atomization process near the nozzle. Moreover, this deterministic boundary condition produces simple hydrodynamic instabilities that are well resolved for all the grid size employed.…”

“…A better understanding of the relation between the jet evolution analyzed in the frames (a-d) in Figure 3 and the grid resolution effect on the spray characteristics requires a more detailed observation of the breakup mechanisms. These processes have been studied by [9] and [16], among others.…”

“…Besides these difficulties, DNS is still used to study early atomization stages on fuel injection problems. [16] and [17,18] applied the VOF technique with Adaptive Mesh Refinement (AMR) to analyze the primary atomization of diesel, biodiesel, and gasoline on injection problems at moderate speeds.…”

Spillback nozzle characterization using pulsating LED shadowgraphy

“…However, when applying this method to a full-scale system, the spatial discretization needed to observe the first stages of the atomization process is much smaller than the other characteristics lengths [5]. With the recent increase in computational power and developments of DNS, it is possible to conduct such analysis, but it is still computationally very demanding [2,[6][7][8][9]. As a simplification, it is possible to inject already atomized droplets in the computational domain, thus avoiding modelling directly the atomization process [10,11].…”