IMAS: VMEC Free Boundary Integration

[lazersos]

VMEC has been interfaced as a fixed boundary solver in IMAS via iWrap. The basic idea is that iWrap wrappers fortran code VMEC provides, allowing the passing of IDS's back and forth. All the iWrap stuff can be found in the IMAS_INTERFACE directory and the VMEC interface code can all be found in VMEC2000/Sources/TimeStep/vmec_imas.f90. This issue is to help track the implementation of free boundary VMEC into IMAS.

[lazersos]

A little background as to why I @lazersos didn't finish this part when I had a chance. TLDR: the IDS for coils are a mess. There are three IDS coil types in IMAS

• TF: Toridal field
• PF: Poloidal field
• 3D: General field coil descriptions

Now my thought was to the the interface working such that if the user could call the code with a mix of coil IDS. The general flow would be:

1. Read the IDS's to get something like a centerline of the coil.
2. Use the existing coil routines in LIBSTELL to generate a B-field on a grid in MGRID.
3. Since we'll have loaded info into the mgrid_mod, VMEC should just behave like it read an MGRID.

NOTE: some of my nomenclature here may be off.