Overcoming input file function topology limitations

[mikekryjak]

The fact that the core and PFR are inseparable in the input file means we can't use analytical expressions for two of the most useful applications in tokamak models:

• A profile of anomalous diffusion coefficients where core has one profile and SOL/PFR another
• A source of particles/energy at the first core ring

A solution to this would be to introduce a special input file variable named e.g. is_periodic which would work as a heaviside function that is 1 in the core and 0 elsewhere.

I seem to remember that there are also issues with the y variable which means it can't be used the same as theta. I can't remember exactly what was wrong with it, but when I was trying to add a neutral puff and pump (another two most common applications in tokamaks!), I found that a single y selection would result in more than one theta position.

[johnomotani]

The issue I know of with y is that in a double-null the upper (I think) PFR has the same y-values as parts of the core. It kinda has to, otherwise y would need 2 branch cuts in the core, not just the single branch cut at 0->2pi.