MESAHub/mesa

missing atm options

Debraheem opened this issue · 18 comments

Debraheem commented

Hello, the following atm options don't appear to exist in the code any more, but they remain in the documentation. Does anyone know if they should be re-implemented or removed? I looked in $MESA_DIR/star/private/atm_support.f90, but no case exists for these options?

'paczynski_grey' Paczynski Grey Atmosphere
'fixed_Teff' Fixed Effective Temperature
'fixed_Tsurf' Fixed Surface Temperature
'fixed_Psurf' Fixed Surface Pressure
'fixed_Psurf_and_Tsurf' Fixed Surface Pressure and Temperature

evbauer commented

It looks like Paczynski_grey was present in releases up through r15140, but it's not clear that it ever actually worked after the atm overhaul circa r12115. E.g. I just tried to run a vanilla 1 Msun model with that option selected in r12778 and I just repeatedly get

 do1_create_Paczynski_atm failed in dopri853: idid          -2
 failed in do1_create_Paczynski_atm
 retry       1

Bill deleted the Paczynski_grey option in #143.

evbauer commented

It looks like fixed_Teff, fixed_Tsurf, fixed_Psurf, and fixed_Psurf_and_Tsurf were deleted in #200.

  • 👍1
Debraheem commented

Thanks Evan, If there is no response to your mailing-list, I'll make a pull to remove these options from the documentation.

  • 👍1
fxt44 commented

yep, circa jan 2021

remove Paczynski atm options
Paczynski_grey
trace_atm_Paczynski_grey
Paczynski_atm_R_surf_errtol
use_T_Paczynski_outer_BC

no real reason for it other than cleanup as it didn't seem to used much (or in the test suite).

on fixed_Teff, fixed_Tsurf, fixed_Psurf, and fixed_Psurf_and_Tsurf, i recall those were used in early versions of the conductive_flame test suite case. now it uses use_other_surface_PT = .true. to do its boundary condition magic.

warrickball commented

on fixed_Teff, fixed_Tsurf, fixed_Psurf, and fixed_Psurf_and_Tsurf, i recall those were used in early versions of the conductive_flame test suite case. now it uses use_other_surface_PT = .true. to do its boundary condition magic.

I have a (very long...) MESA to-do list on which I at some point made a note to try to figure out what happened to the fixed_* options for atm, so we must've discussed it before. I'm completely snowed under until mid-September and possibly until mid-October but can have a go at reviving those options then.

fxt44 commented

hmm. i'm not finding any email threads discussing the fixed* options.

warrickball commented

hmm. i'm not finding any email threads discussing the fixed* options.

I suspect it got little more than a passing mention in a telecon.

anwesha1111 commented

Hello,

regarding WD asteroseismology, the low optical depth part of the models needs to be well calculated. Now, the table atmosphere WD_tau_25 is a great option for analyzing interiors (esp the cooling timescales), but attaches the first model grid to optical depth of 25.14. In case of bursting DAs most of the interesting stuffs are out to a very low optical depth (ideally first model grid at 1e-1 depth and first atmosphere grid extending out to even lower depths say 1e-4)
Two earlier options that are now dropped--> atm_option = 'Paczynski_grey' and the test suite wd_sruf_tau1m4 allowed to put the surface at such low optical depts and compute the atmosphere structure

is there a way these can be brought back?

evbauer commented

The test suite case wd_surf_at_tau_1m4 isn't included in MESA anymore, but as far as I can tell all of the functionality it included is still supported. Is there something specific about the test case that needs to be brought back?

anwesha1111 commented

I don’t know what was there exactly in that test suite, but from the name it seems that it allows to put the first model grid at 10^-4 optical depth. Also working around this can be done by loading the desired model and relax it and to save a relaxed model (it should stop when surface reaches a defined optical depth —> this is the stopping condition) without doing anything else with it? What termination code should I be using in this case?

fxt44 commented

in 12778, the wd_surf_at_tau_1m4 test case used

  atm_option = 'T_tau'
  atm_T_tau_relation = 'Eddington'
  atm_T_tau_opacity = 'varying'

  set_tau_factor = .true.
  set_to_this_tau_factor = 0.00015

and terminated after ~1000 steps when
Teff_lower_limit = 1e5

here is one of the output plots:

Log_Tau_vs_logxq.pdf

anwesha1111 commented
fxt44 commented

fwiw, i'd be in favor of maybe reviving this test suite case if you cannot get it to go in r23.05.1 .

earlbellinger commented

Since this is marked as a release blocker: what is the status of this issue, and what remains to be done?

anwesha1111 commented
Debraheem commented

I would like to see the fixed options return. In trying to get the PPISN test suite to work with TDC, it's useful to remove material ejected from shocks forming near the surface of a model, and setting a new boundary condition below the surface during the removal. The fixed_Tsurf option makes the removal process possible.

I can look in previous versions and try to revive them.

evbauer commented

@Debraheem we can close this issue now that #612 has been merged, right?

evbauer commented

(got the go-ahead from @Debraheem on slack, so closing)