FGONG and gyre_nad

Bug/problem reports for any of the GYRE executables (gyre_ad, gyre_nad, etc)
pieterd
Posts: 9
Joined: Mon Apr 08, 2013 2:09 am

FGONG and gyre_nad

Postby pieterd » Wed Feb 05, 2014 11:35 am

Hi Rich,

I'm trying to run gyre_nad on an FGONG file, but I get

Code: Select all

Root Solving
       l    n_pg     n_p     n_g                 Re(omega)                 Im(omega)                       chi  n_iter        n
 ABORT at line 794 <gyre_coeffs_evol:tau_thm_1>:
 tau_thm is undefined
STOP Program aborted


after the adiabatic calculations went fine. I'm not familiar with the source code at all, but I see in the "gyre_fgong_file.f90" that when the base coefficients are initialized, the information on the opacity and opacity derivatives is not loaded:

Code: Select all


call ec%init(G, M_star, R_star, L_star, r, m, p, rho, T, &
                 N2, Gamma_1, nabla_ad, delta, SPREAD(0._WP, DIM=1, NCOPIES=n), &
                 deriv_type, add_center)



Could this be the reason? In the official FGONG file format description, all the columns you require should be there (nabla=6, kappa=8, kappa_rho=18, kappa_T=17, Omega_rot=16, epsilon=9, epsilon_rho=20, epsilon_T=19 (Fortran indexing)).

I can of course translate the FGONG file to an HDF5 file if this is really the issue, but I just thought you might know.

Cheers,
Pieter

User avatar
rhtownsend
Site Admin
Posts: 172
Joined: Sun Mar 31, 2013 4:22 pm

Re: FGONG and gyre_nad

Postby rhtownsend » Wed Feb 05, 2014 12:21 pm

Hi Pieter --

The routines in GYRE to read FGONG files are based off the description here:

http://www.astro.up.pt/corot/ntools/doc ... _Files.pdf

Do you have a more-recent description of the FGONG format, which I could use to update the code?

cheers,

Rich

pieterd
Posts: 9
Joined: Mon Apr 08, 2013 2:09 am

Re: FGONG and gyre_nad

Postby pieterd » Thu Feb 06, 2014 2:54 am

Hi Rich,

ah sorry, I was looking at the OSC format for my column description...

Only columns 37 to 40 are unused, but 37 and 38 are unofficially reserved for nabla and nablarad, that leaves only 39 and 40 for the other variables... that's a no go..

I will just stick to the GSM format then.

I don't know in how far you want to catch stupid user errors; but perhaps raising an error when running gyre_nad with impossible file formats can give some comprehensive output? Working with different codes, I sometimes forget that information is lost in file conversions.

Cheers & thanks,
Pieter

acsj
Posts: 1
Joined: Tue Sep 20, 2016 12:21 pm

Re: FGONG and gyre_nad

Postby acsj » Thu Sep 22, 2016 7:04 am

Hi GYRE team,

I am encountering the same problem that was raised in this report from 2013 when using gyre_nad in combination with garstec-gong files - while gyre_ad seems to be able to read these, gyre_nad cannot work with the supplied fgong files:

Root Solving
l n_pg n_p n_g Re(omega) Im(omega) chi n_iter n
0 1 1 0 0.2609750703206698E+01 0.0000000000000000E+00 0.6125537242514969E-14 6 571
0 2 2 0 0.4031881839925865E+01 0.0000000000000000E+00 0.1204521142524941E-13 6 571
0 3 3 0 0.5346291299614568E+01 0.0000000000000000E+00 0.5612047471814812E-14 6 571
0 4 4 0 0.6784089084278294E+01 0.0000000000000000E+00 0.2296951617124046E-14 6 571
0 5 5 0 0.8225799696929936E+01 0.0000000000000000E+00 0.1184834303157981E-13 6 571
0 6 6 0 0.9690708047750153E+01 0.0000000000000000E+00 0.1929800969225210E-13 5 571
0 7 7 0 0.1113221046886458E+02 0.0000000000000000E+00 0.1222363706211937E-13 6 571
0 8 8 0 0.1258054994512614E+02 0.0000000000000000E+00 0.1076265951011770E-13 6 571
0 9 9 0 0.1400711127178552E+02 0.0000000000000000E+00 0.1411590033455787E-13 6 571
0 10 10 0 0.1540179858447089E+02 0.0000000000000000E+00 0.2168522233300714E-14 6 571
0 11 11 0 0.1677316259619771E+02 0.0000000000000000E+00 0.3712457164431741E-13 6 571
0 12 12 0 0.1813071823486514E+02 0.0000000000000000E+00 0.9468405963672912E-14 6 571
0 13 13 0 0.1949602835464507E+02 0.0000000000000000E+00 0.1651407048933930E-13 6 571
Time elapsed : 0.329 s
Root Solving
l n_pg n_p n_g Re(omega) Im(omega) chi n_iter n
ABORT at line 785 <gyre_evol_model:nabla_1_>:
nabla is undefined
STOP Program aborted ABORT at line 785 <gyre_evol_model:nabla_1_>:

ABORT at line 785 <gyre_evol_model:nabla_1_>:
nabla is undefined
STOP Program aborted
nabla is undefined
STOP Program aborted/


Anything else I can do but convert to another format? If not, what procedure would you suggest? I have tried to convert the gongfile to an amdl file, using fgong-amdl.d supplied by ADIPLS, but gyre_nad can seemingly not work with this format either:

Reading from FAMDL file testgong.amdl
At line 126 of file gyre_famdl_file.f90
Fortran runtime error: Bad value during integer read


Cheers,

Andreas

User avatar
rhtownsend
Site Admin
Posts: 172
Joined: Sun Mar 31, 2013 4:22 pm

Re: FGONG and gyre_nad

Postby rhtownsend » Thu Sep 22, 2016 7:49 am

Hi Andreas --

acsj wrote:Hi GYRE team,

I am encountering the same problem that was raised in this report from 2013 when using gyre_nad in combination with garstec-gong files - while gyre_ad seems to be able to read these, gyre_nad cannot work with the supplied fgong files:

Root Solving
l n_pg n_p n_g Re(omega) Im(omega) chi n_iter n
0 1 1 0 0.2609750703206698E+01 0.0000000000000000E+00 0.6125537242514969E-14 6 571
0 2 2 0 0.4031881839925865E+01 0.0000000000000000E+00 0.1204521142524941E-13 6 571
0 3 3 0 0.5346291299614568E+01 0.0000000000000000E+00 0.5612047471814812E-14 6 571
0 4 4 0 0.6784089084278294E+01 0.0000000000000000E+00 0.2296951617124046E-14 6 571
0 5 5 0 0.8225799696929936E+01 0.0000000000000000E+00 0.1184834303157981E-13 6 571
0 6 6 0 0.9690708047750153E+01 0.0000000000000000E+00 0.1929800969225210E-13 5 571
0 7 7 0 0.1113221046886458E+02 0.0000000000000000E+00 0.1222363706211937E-13 6 571
0 8 8 0 0.1258054994512614E+02 0.0000000000000000E+00 0.1076265951011770E-13 6 571
0 9 9 0 0.1400711127178552E+02 0.0000000000000000E+00 0.1411590033455787E-13 6 571
0 10 10 0 0.1540179858447089E+02 0.0000000000000000E+00 0.2168522233300714E-14 6 571
0 11 11 0 0.1677316259619771E+02 0.0000000000000000E+00 0.3712457164431741E-13 6 571
0 12 12 0 0.1813071823486514E+02 0.0000000000000000E+00 0.9468405963672912E-14 6 571
0 13 13 0 0.1949602835464507E+02 0.0000000000000000E+00 0.1651407048933930E-13 6 571
Time elapsed : 0.329 s
Root Solving
l n_pg n_p n_g Re(omega) Im(omega) chi n_iter n
ABORT at line 785 <gyre_evol_model:nabla_1_>:
nabla is undefined
STOP Program aborted ABORT at line 785 <gyre_evol_model:nabla_1_>:

ABORT at line 785 <gyre_evol_model:nabla_1_>:
nabla is undefined
STOP Program aborted
nabla is undefined
STOP Program aborted/



Unfortunately, FGONG files lack the sufficient data to perform non-adiabatic calculations -- nabla = dlnT/dlnP is absent, and so are the partial derivatives of the opacity and the nuclear energy generation rate. That's ultimately why GYRE stops.

Anything else I can do but convert to another format? If not, what procedure would you suggest? I have tried to convert the gongfile to an amdl file, using fgong-amdl.d supplied by ADIPLS, but gyre_nad can seemingly not work with this format either:

Reading from FAMDL file testgong.amdl
At line 126 of file gyre_famdl_file.f90
Fortran runtime error: Bad value during integer read



AMDL won't work, because it suffers from the same problem. (But as an aside, the error message you get here is because you are trying to read an unformatted AMDL file as a formatted FAMDL file).

Ultimately, you're going to have to modify the code that writes the models, to output in a format which includes nabla and the partials. The OSC format (see https://www.astro.up.pt/corot/ntools/do ... _Files.pdf) is one suitable choice.

cheers,

Rich


Return to “Bug reports”

Who is online

Users browsing this forum: No registered users and 1 guest