Symmetry Problems in Jet Flow Simulation

[tianninglyu]

Hello everyone,

I hope this message finds you well. I have a specific concern regarding the GetReal function in Athena++ when working with negative numbers, particularly in the context of simulating jets.

Currently, I am conducting simulations of jets using Athena++ and have encountered an unexpected behavior. When I set the jet's magnetic field to both 0.1 and -0.1, I obtained identical simulation results. This outcome has raised doubts about the GetReal function's ability to correctly handle negative values.

Thank you very much for your attention and support. I am grateful for any assistance or suggestions you can provide regarding this issue.

Additionally, my timezone just entered June 1st. I would like to wish all athena++ developers and users a happy Children's Day, if you want your childlike innocence to last forever.

The focus of the question has changed. Please read below.

[felker]

GetReal can certainly handle negative values just fine--- which block and parameter are you setting in the input file?

[tianninglyu]

Another problem. Is this simulating a impulsive jet or a continuous jet? Because by reading jet.cpp, I seem to only understand the setting of the boundary conditions, but I still seem to know nothing about how the code makes the jet inflow.

[felker]

If the boundary condition function does not have some explicit time dependence, then it is enforced at all timesteps. So it is a continuous jet forced by constant density, pressure, and vx in the invisible ghost cells to the left of the domain (within the radius of the inflowing jet):

djet = 0.1
pjet = 6.0e-3
vxjet = 1.0

Is symmetry broken if you turn off magnetic fields?

I can't recall if x1 is special or not; you could try moving the boundary condition and domain size as you suggested.

[tianninglyu]

If turning off magnetic field, or turning on it but set all zero in the input file, the symmetry is very good, not broken. However I think there is maybe a mistake here. If I set the magnetic field as:

B_amb = (0,0,0)
B_jet = (0.1,0,0)

Are such magnetic fields illegal? because its normal component at the x1 boundary is discontinuous.

[c-white]

That boundary condition is valid, it just means there's a magnetic monopole in the boundary. This may mean the ghost cells are not physically realizable, but that's not uncommon with boundary conditions in MHD problems. The monopole can't advect onto the grid proper. You could always try to write a monopole-free boundary condition, but it's tricky (it will have to depend on the time-evolving state of the active cells), and in the end it wouldn't meaningfully change the results.

[tianninglyu]

Thank you for your reply!

[RenyueCen]

Hi Chris (and Kyle), thanks for helping. Tianning is learning Athena++ to do some experiments on how radio jets propagate through environments around SMBH to large scales. But he is running into the problem of jet becoming asymmetric with time for an apparent symmetric problem using an existing setup routine (in the hydro dir, though). Not sure what is going on. Any help is highly appreciated.

[tianninglyu]

Thank you for your kindness and consideration!