Confusing specification of q in input file

[jonmaddock]

In GitLab by @mkovari on Aug 3, 2015, 11:27

q /3.0/ : safety factor 'near' plasma edge (iteration variable 18): equal to q95 (unless icurr = 2 (ST current scaling), in which case q = mean edge safety factor qbar)

This is confusing. I suggest we use instead:

q95 /3.0/ : safety factor at 95% flux surface (iteration variable 18). Not used if icurr = 2 (ST current scaling).
qbar /3.0/ : mean edge safety factor qbar.  Only used if icurr = 2 (ST current scaling)

Note that the code should ideally be backwards compatible.

[jonmaddock]

In GitLab by @mkovari on Aug 3, 2015, 11:29

Title changed from Confusing specification of q to Confusing specification of q in input file

[jonmaddock]

In GitLab by @jmorris-uk on Nov 3, 2016, 16:15

Milestone removed

[jonmaddock]

In GitLab by @jmorris-uk on Jul 17, 2019, 18:04

removed milestone

[chris-ashe]

Looking at the definitions in input.f90 and the different function which use it, there is a greater case to keep q as q95 and qbar are only used in very specific circumstances.
The definition of q in input.f90 is sufficient and the definition in an input file can vary wildly depending on the input file chosen. Though this is a obsolete legacy problem as there is currently no definitions given in any of the regression test examples.

[mkovari]

Just to update the issue: this is the text in Variable Descriptions:

Name	Type	Datatype	Default Value	Description
q	Input	real	3.0	safety factor 'near' plasma edge (iteration variable 18) equal to q95 (unless icurr=2 (ST current scaling), in which case q = mean edge safety factor qbar)
q0	Input	real	1.0	safety factor on axis
q95	Output	real	-	safety factor at 95% surface

Here are the relevant lines from the large tokamak input file:

q = 6.4 
* DESCRIPTION:   Safety factor at 95% flux surface

In my opinion the variable name q is not well-chosen and could do with improvement. (Note my previous comment that it would be good if the code was backwards compatible, so it could still read the variable q from the input file if it is present.)

The output in the large tokamak file looks OK:

 Safety factor on axis                                            (q0)                          1.000     
 Safety factor at 95% flux surface                                (q95)                         3.656  ITV

[mkovari]

Name	Type	Datatype	Default Value	Description
q95	Input	real	3.0	safety factor at 95% flux surface (iteration variable 18) (unless icurr=2 (ST current scaling), in which case q95 = mean edge safety factor qbar)
q	Input	real	1.0	For backwards compatibility only, q95 can be entered using the symbol q.
q0	Input	real	1.0	safety factor on axis

[mkovari]

It would be good to update the documentation. Will this happen automatically at some point?