 |
There is a growing interest in processes of trans- and supercritical atomization in high pressure and temperature
applications, e.g. rocket engines, high pressure piston engines, chemical production, etc.. Therefore a detailed
understanding of trans- and supercritical processes is essential to improve such applications. Those applications
have an injection process in common. Thereby sub- or supercritical single as well as multi-component fluids are
injected, atomized and/or mixed. To support jet injection design tasks, numerical simulations are necessary. To
deal with usually involved complex configurations and geometries, it is well accepted to decrease the complexities
while keeping the essential features of the application under consideration. In this contribution a mixing layer which
includes a shear layer boundary to mimic the interface of separated fluids is studied. Thereby, complex species
diffusion and mixing processes take place.
Despite the increased available computational power, direct numerical simulations (DNS) of such high pressure
mixing processes are not efficiently feasible. In order to establish a numerical tool which is computationally efficient
and economically acceptable, the present contribution suggests a Large Eddy Simulation (LES) framework which
includes the Smagorisky subgrid-scale model and well adapted multi-component species diffusion and mixing mod-
els to study the Okong'o configuration. In this configuration two fluids, composed of pure oxygen and hydrogen,
respectively, are forced under supercritical conditions to build a predefined binary shear layer with temporal vortex
roll ups in order to explore the features of the interface mixing with initial density stratifications up to 24.4.
All LES simulations are conducted with the open-source solver OpenFOAM utilizing an in-house low Mach solver
which includes real gas properties by means of the Peng-Robinson equation of state to deal with high pressure
conditions.
Global shear layer mixing features are examined over time to describe the mixing. Developing vortices leading to
specific temporal enstrophy and vorticity are compared to reference DNS data, especially the mixing processes
within the temporal vortex roll ups, in order to validate the utilized diffusion and mixing models.
