Add an initial homogeneous isotropic turbulence in PeleLMeX

[Rachel008996]

I noticed that in the tutorials ChallengeProblem, initial homogeneous isotropic turbulence has been introduced.

PeleLMeX/Exec/Production/ChallengeProblem/README.md

Line 28 in 2a03eb6

Velocity in the chamber is initialized from a precursor HIT simulation. Data are stored in a binary

However, the explanation provided is not detailed enough, leaving me a bit confused. Could you kindly guide me on how to set these parameters, or direct me to resources where I can find the underlying rationale for these settings?

[dmontgomeryNREL]

Hi @Rachel008996, the precursor HIT simulation is generated using the gen_hit_ic.py Python script in the utilities of PelePhysics. The documentation in PelePhysics will walk you through the process of generating the initial .dat file and the inflow files for turbulent inflow boundary conditions. The input file controls for TurbInflow, which are inherited by PeleLMeX, are discussed here. Once these files are generated you can copy them into your specific case.

I found it helpful to look at the GitHub workflow for the TurbInflow RegTest (line 175 of ./github/workflows/linux.yml) to see how the files are generated for that specific case. The files are not included in the PeleLMeX repository as they can be quite large depending on the parameters.

The challenge problem you are interested in injects turbulent inflow in 4 locations (see turbinflow.____ in input.3d). The ic.___ controls you are referring to above are related to a turbulent initial condition, which is a lot more involved than turbulent inflows, as the user has to:

• read in the hit_ic file and copy the data to GPU memory (starts at line 147 pelelmex_prob.cpp)
• interpolate the fluctuations to the multiple levels of the mesh (starts at line 102 of pelelmex_prob.H)

Adding a more straightforward functionality for initializing the velocity field with turbulent fluctuations would be a great addition to PeleLMeX. I hope this helps!