boutproject / xhermes Goto Github PK

It would be useful to know this downstream, as to be consistent with normalised calculations, you need to normalise constants too, i.e. q_e and M_p become 1.

Currently extract_1d_tokamak_geometry guard replaces pos but no other variables. Guard replacement should be done for all variables at the same time and be user-settable with a flag.

The current calculation of cell volume and area in 1D assume dx = dz = 1. This is not always the case, in fact dz is 2pi by default even in 1! The formulas should use the proper expressions for consistency across dimensions:

Cell volume
dv = J * dx * dy * dz

Parallel cross-sectional area
da = dx * dz * J/sqrt(g_22)

As per:
https://github.com/mikekryjak/sdtools/blob/b336ec077fa9c9d4772965d2e59be9f00a60fcb2/hermes3/load.py#L1162-L1175

I put together some ideas for what functionality could go into xHermes. The concept is partially based on what I'm used to from my CFD life with ANSYS Fluent: being able to easily define 1D, 2D and 3D regions and obtain complex information from them quickly.

Part of this is already completed in my "private xHermes" (which needs tidying and has different names for the different bits but which can be found here: https://github.com/mikekryjak/sdtools/blob/next/h2D.py) which I am currently actively developing.

I haven't given design objectives much thought yet, but they would be along the lines of:

• Providing a framework for more complex post-processing based on sliced xBOUT datasets.
• The ability for a new user to perform a comprehensive particle, energy and momentum balance with trivial effort. This will help with keeping the code bug free but is also a great way to learn how the code works.
• The ability to easily track model convergence in a standardised way.

One major challenge will be to make sure this is all done in a way that makes sense and is crystal clear across all the different applications of Hermes-3 (1D, 2D, 3D, SN/DN tokamak, non-tokamak geometries, etc..). I think it makes sense to give this a lot of thought from the very outset.

Open to any and all discussions/comments! I would like to have a brainstorming session on this over Zoom if anyone is interested. Here are the notes - unfortunately I wrote them in OneNote and can't see an easy way to convert it to Markdown without a lot of editing so I'm putting it in as a picture and will amend later.