Characterization of pf_passive loops (to which part of the tokamak they belong)

[imbeauf]

This is a follow-up of an old ticket IMAS-3842, which discussed how to characterize to which part of a tokamak a passive loop belongs.
IMAS-3842 suggested that could implement this via an identifier at the loop level, the following initial list of possible options was suggested:
VV inner shell, VV outer shell, cryostat, triangular support

Is this still useful ?
Do you have suggestions for other options for this identifier ?

[MaxDubrov]

I think this is still useful, because the only identifier we have now is name field in the loop structure. Some identifier field with a convention for different parts of the machine (as we have for the coils), perhaps unique for each machine, would be appreciated.

[imbeauf]

The identifier we have in pf_active is about the coil primary function, I think for passive elements it's more a description of a machine part that you need.
We already have a machine part identifier used in the WALL IDS, do we want to extend this one and put it in utilities (as a common identifier for use in multiple IDS ?):
"port", "divertor", "baffle", "blanket", "limiter", "panel", "antenna"

[olivhoenen]

cf. https://imas-data-dictionary.readthedocs.io/en/latest/generated/identifier/wall_component_identifier.html

Question 1: would the new ID (e.g. VV inner shell, VV outer shell, cryostat, triangular support) make sense in the context of the wall IDS?
Question 2: what would be the backward compatibility implication of moving an identifier to be defined form within an IDS to the utilities section? (probably none unless we change the namespace for this group of identifiers, but would these additions make sense inside a wall_component_identifier)

@hrdoktorx @MaxDubrov

[hrdoktorx]

What do you mean by utilities section?

[imbeauf]

The utilities section is a folder in the DD repository in which we put all the material used by more than one IDS. I don't think there is any backward compatibility implication of moving the identifier file.
Olivier, you are right that all elements of the list won't make sense in both contexts (e.g. the cryostat isn't part of the wall) ... so using a single identifier file could lead to some confusion. So let's maintain two separate identifiers, while making sure that we use the same name and description for the common elements.
Unless there are further remarks, I will create an issue to implement a dedicated identifier with the following initial list: VV inner shell, VV outer shell, cryostat, triangular support

[hrdoktorx]

OK. Thanks for the clarification. No objection.

[olivhoenen]

As explained here #196 (comment) this is currently insufficiently motivated and the identifiers values are too machine-specific to make it through the DD as a generic identifier type.
Alternative approaches are to obtain some information by looking at the description, or to keep pf_passive IDSs that describe different parts separate and use the relevant ones in the simulation.