Mesh advice for LEO plasma wake modeling

Hi Everyone,

New SPIS user here. Trying to model a plasma wake around a simple box in LEO conditions. The SPIS user manual says that the plasma volume should be meshed with grids smaller than 0.5 * Debye length. However, for a LEO plasma, the debye length is like ~1 cm, so this results in a very fine mesh. To have a sufficiently large computational volume, the resultant mesh is huge and SPIS errors because of insufficient memory. How can I set up a mesh for these conditions?

Thank you.

Depending on the obtained mesh size, one “brutal” solution could be to increase the memory available for SPIS, respecting the maximum memory available for your system.
If your run your simulation directly through the SPIS interface, you can change it in the preferences panel (and restart SPIS for the change to be taken into account).
If you run in 'batch mode", you can edit the launch script ( / Spis.bat) and change the HEAP_SIZE value in the file.

Thank you for the reply, @atrouche. I will try increasing the available memory.

Are there any more elegant ways to set the mesh size? As far as I can tell, the way to adjust the mesh size is changing the “element size factor” from within the mesh options window in Gmsh. Are there any more direct ways to set the meshsize? Thank you!

I figured out the solution to my prior question of setting the mesh size. The resolution parameter used when creating an object will determine the mesh size adjacent to that object. For example, if you create a box with a resolution of 0.01, then the meshsize near the box will be 1 cm.