[FLASH-USERS] Magnetic fields at refinement boundaries in USM

[John ZuHone]

Hi Sean and all,

> I’ve also encountered this same issue at refinement jumps with USM. I could never get HLLD to behave. What I found was ok was to use HLLC with high-order electric field construction. Then, even better, I turned on magnetic resistivity with a small value. That smoothed things out nicely. 

I suspect that this is not the best solution for all problems, especially if you’re trying to model some other non-ideal effect such as viscosity or resistivity. It probably depends on what you’re trying to do. 

> May “belief” at this point is that either: 1) staggered mesh CT simply does not behave well with AMR, or 2) the prolongation/restriction routines in PARAMESH for face variables is bugged somehow. 
> 
> John Z., Is this for the mesh refinement versions of Athena and Athena++? Or uniform grids? 

Mesh refinement version of both Athena and Athena++, since I have used both for the same problem as FLASH and not seen the issue. I have not used the GAMER code for this purpose (I will soon) but I have heard from the main developer Justin Schive that he hasn’t seen it either. For these reasons, I do suspect the issue is in PARAMESH or elsewhere in FLASH somewhere. 

John
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://flash.rochester.edu/pipermail/flash-users/attachments/20190329/a5342616/attachment-0001.htm>