Releases: rhdtownsend/gyre
Releases · rhdtownsend/gyre
Release 7.2.1
New Features
- Added new build flags: FRONTENDS, IFACES, FORUM, SHARED
Changed Functionality
- The default values for build flags are now specified in the top-level Makefile
- Clarified the meaning of the table on the model capabilities page
- Relaxed the minimum required MESA SDK version to 21.4.1
- Updated the Frequently Asked Questions to remove sections on static builds and cluster execution
Bug Fixes
- The MESA interface now builds correctly
- Various errors in the installation instructions have been corrected
Release 7.2
New Features
- Added support for
'ANAPOLY'
(analytic polytrope) stellar models - Switched from the fpx3 pre-processor to the fypp pre-processor. This means
that a Python interpreter is now needed to build GYRE - Re-worked build system for faster (and more parallelized) compilation
Bug Fixes
- Fixed various typos in the oscillation equations
Deleted Features
- Building in single precision (32-bit) floating point is no longer supported
- Building without HDF5 support is no longer supported
Release 7.1
New Features
- Added a treatment of turbulent damping in convection zones (see the new Convection Effects section)
- Added the following output items for summary files:
n
dx_min
dx_max
dx_rms
- Added the following output items for detail files:
dx_min
dx_max
dx_rms
- Added support for stellar models in MESA format v1.2
- Added support for stellar models in GSM format v1.2
Bug Fixes
- Fixed an issue with the
'KAWALER'
option for thezeta_scheme
parameter, as described in Townsend & Kawaler (2023) - Fixed an issue when writing
zeta
output item to detail files (beta
was being written instead)
Changed Functionality
- The
'KAWALER'
option is now the default for thezeta_scheme
parameter
Release 7.0
New Features
- Added a new
gyre_tides
frontend for performing tidal calculations - Added the following parameters to the
&grid
namelist group:scaffold_src
file
file_format
- Added the following parameters to the
&mode
namelist group:static
- Added the following parameters to the
&osc
namelist group:alpha_kar
alpha_kat
alpha_trb
- Added the
&orbit
namelist group, for use withgyre_tides
- Added the
&tide
namelist group, for use withgyre_tides
- Added the
&tides_output
namelist group, for use withgyre_tides
- Added the following special sequences used to set the names of detail files:
%K
%k
- Added the following output items for summary files
Omega_rot_ref
Q
zeta
k
eul_Psi_ref
Phi_T_ref
Omega_orb
q
e
R_a
cbar
Gbar_1
Gbar_2
Gbar_3
Gbar_4
- Added the following output items for detail files:
Omega_rot_ref
Omega_rot
dQ_dx
Q
dzeta_dx
zeta
k
eul_Psi_ref
Phi_T_ref
eul_Psi
Phi_T
Omega_orb
q
e
R_a
cbar
Gbar_1
Gbar_2
Gbar_3
Gbar_4
Changed Functionality
- Renamed the following output items for summary files:
n_j
→n_row
j
→id
- Renamed the following output items for detail files:
n_k
→n
j
→id
eul_phi
→eul_Phi
deul_phi
→deul_Phi
eul_phi_ref
→eul_Phi_ref
deul_phi_ref
→deul_Phi_ref
- Renamed the following special sequences used to set the names of detail files:
%J
→%ID
%j
→%id
Removed Functionality
- Deleted the following parameters from the
&osc
namelist group:alpha_kap
(functionality replaced byalpha_kar
andalpha_kat
)
Bug Fixes
- Changed program termination behavior so that abnormal stopping
conditions produce an error exit code
Release 6.0.1
- Fixed typos in documentation
- Significantly improved performance in TAR routines
- Fixed undefined frequency bug when using
'ISOTHERMAL'
and'UNNO'
outer boundary conditions with radial modes
Release 6.0
New Features
- Created comprehensive documentation
- Added the following parameters to the
&mode
namelist group:static
- Added the following parameters to the
&osc
namelist group:outer_bound_cutoff
outer_bound_branch
alpha_grv
alpha_thm
alpha_hfl
alpha_gam
alpha_pi
alpha_kap
alpha_rht
zeta_scheme
x_atm
adiabatic
- Added a new
&rot
namelist group, for controlling rotation effects - Added the following parameters to the
&num
namelist group:ad_search
nad_search
- Added the following parameters to the
&scan
namelist group:freq_units
freq_min_frame
freq_max_frame
file
axis
- Added the following parameters to the
&grid
namelist group:dx_max
w_ctr
resolve_ctr
- Added the following output items for summary files
n_j
(number of modes found)domega_rot
(dimensionless rotational splitting)dfreq_rot
(dimensioned rotational splitting)
- Added the following output items for detail files:
domega_rot
(dimensionless rotational splitting)dfreq_rot
(dimensioned rotational splitting)dzeta_dx
(dimensionless frequency weight function)
- Added a new
'ISOTHERMAL'
option for theouter_bound
parameter - Added the ability to read WDEC stellar models
- Added support for mode isolation, via the
alpha_gam
andalpha_pi
parameters and the'GAMMA'
outer_bound
option - Added new contour and minmod methods for finding non-adiabatic modes
- Added the new
eval_lambda
executable, for calculating eigenvalues of Laplace tidal equations - Added support for statically linked compilation
- Added support for distributed execution using MPI
Changed Functionality
- Moved/renamed the following parameters in the
&model
namelist group:Omega_rot
→&rot
/Omega_rot
Omega_units
→&rot
/Omega_rot_units
- Moved/renamed the following parameters in the
&mode
namelist group:rossby
→&rot
/rossby
- Moved/renamed the following parameters in the
&osc
namelist group:rotation_method
→&rot
/coriolis_method
complex_lambda
→&rot
/complex_lambda
- Moved/renamed the following parameters in the
&grid
namelist group:alpha_osc
→w_osc
alpha_exp
→w_exp
alpha_thm
→w_thm
alpha_str
→w_str
- Moved/renamed the following parameters in the
&ad_output
/&nad_output
namelist groups:mode_template
→detail_template
mode_file_format
→detail_file_format
mode_item_list
→detail_item_list
mode_filter_list
→detail_filter_list
- Renamed the following output items for detail files:
dtau_dx_ss
→dtau_ss_dx
dtau_dx_tr
→dtau_tr_dx
- Changed the type of the
M_star
,R_star
&L_star
detail output items to array - Switched to IAU 215 / CODATA values for physical constants
- Switched to using the crlibm / crmath math libraries, enabling bit-for-bit consistency across platforms
Removed Functionality
- Deleted the following parameters from the
&osc
group:cowling_approx
(functionality replaced byalpha_grv
)nar_approx
(functionality replaced byalpha_thm
)narf_approx
(functionality replaced byalpha_hfl
)eddington_approx
(functionality replaced byalpha_rht
)
- Deleted the following parameters from the
&grid
namelist group:n_inner
(functionality replaced byw_ctr
)
- Deleted the following parameters from the
&ad_output
/&nad_output
namelist groups:prune_modes
Bug Fixes
- Fixed slow evaluation of
beta
anddbeta_dx
output items - Fixed bug in evaluating surface point for n=0 polytropes
- Fixed bug in reading OSC files
- Fixed bug in reading AMDL/FAMDL files
- Fixed bug in JCD variables_set implementation
- Fixed bugs in output routines when pruning modes (pruning is no longer necessary)
- Fixed bug in reading FGONG files that are order-reversed
- Fixed bug using TAR with radial modes
Release 5.2
New Features
- Added the following output items:
alpha_0
(excitation coefficient)alpha_1
(excitation coefficient)c_rad
(dimensionless radiative luminosity)E_ratio
(ratio of mode inertias inside/outside reference point)tau_ss
(steady-state torque)tau_tr
(transient torque)
- Updated the equations.pdf document to better reflect the set of equations that GYRE solves
- Added the
nar_approx
flag to the&osc
namelist; set to.TRUE.
to enable the non-adiabatic reversible (NAR) approximation.
Changed Functionality
- Removed the
'NULL'
option for rotation, since it is not strictly needed. The default is now'DOPPLER'
. To re-create the'NULL'
option, use a non-rotating model or setOmega_uni=.TRUE.
andOmega_rot=0
in the&model
namelist.
Bug Fixes
- Fixed various memory errors arising when the traditional approximation is used on a many-core machine (these errors originate from thread-unsafe I/O).
- Restored the
'psi_T'
and'psi_g'
output items, which for some reason had disappeared. - Fixed a problem where spaces in comma-separated list were not being ignored.
- Fixed a sign error in the evaluation of the
'I_1'
output item.
Release 5.1
New Features
- Added the
eddinton_approx
flag to the&osc
namelist; set to.TRUE.
to enable the Eddington approximation (instead of the usual diffusion approximation). - Added the
quasiad_eigfuncs
flag to the&osc
namelist; set to.TRUE.
to calculate quasiadiabatic entropy and luminosity perturbations during adiabatic calculations. These are then used in the calculation of related quantities, such as work functions. - Added the
deps_scheme
parameter to the&osc
namelist; this (together with thedeps_file
anddeps_file_format
) can be used to incorporate frequency-dependent nuclear burning partials. - Added the
complex_lambda
flag to the&osc
namelist; set to.TRUE.
to use complex arithmetic when evaluating the angular eigenvalue lambda in non-adiabati calculationsc. If set to.FALSE.
, then lambda will be evaluated using real arithmetic, with the spin parameter q based on the adiabatic frequency of the mode being sought (this is preferable on numerical grounds). - Added the following output items:
c_lum
(luminosity thermal coefficient)c_thn
(optically-thin thermal coefficient)c_thk
(optically-thick thermal coefficient)c_eps
(energy generation thermal coefficient)lag_S_ref
(specific entropy perturbation at reference point)lag_L_ref
(radiative luminosity perturbation at reference point)eps_rho
(energy generation partial wrt density)eps_T
(energy generation partial wrt temperature)kap_rho
(opacity partial wrt density)kap_T
(opacity partial wrt temperature)
Changed Functionality
- Split the
freq_units
parameter into separatefreq_min_units
andfreq_max_units
parameters
Release 5.0
New Features
- Introduced support for models containing density discontinuities (represented by a double point)
- Updated the
build_poly
executable to allow creation of mixed polytropes with density discontinuities - Added the
poly_to_fgong
executable, which creates FGONG-format files from polytrope files - Added the following parameters to the
&model
namelist:Omega_units
dx_snap
repair_As
- Added the
rossby
parameter to the&mode
namelist: - Added the following parameters to the
&osc
namelist:nonadiabatic
time_factor
conv_scheme
- Added the
'MIRK'
(mono-implicit Runge-Kutta) option to thediff_solver
parameter - Added the following options to the
freq_min_units
andfreq_max_units
parameters:'LOWER_DELTA'
'UPPER_DELTA'
'ROSSBY_I'
'ROSSBY_O'
- Added the following output items:
Delta_p
(acoustic large frequency separation)Delta_g
(gravity inverse period separation)H
(total mode energy)W_eps
(work due to nuclear processes)nabla
(structure coefficient)dnabla_ad
(structure coefficient)c_dif
(structure coefficient)c_thm
(structure coefficient)y_
i (i=1,...,6; dimensionless eigenfunctions)dW_dx_eps
(differential work due to nuclear processes)dtau_dx_ss
(steady-state differential torque)dtau_dx_tr
(transient differential torque)
- Added support for versions 1.00 and 1.01 MESA-format files
- Added support for version >= 1000 FGONG-format files
- Added
%m
option to substitutions applied tomode_template
parameter
Changed Functionality
- Replaced the separate
gyre_ad
andgyre_nad
executables with a monolithic executable,gyre
, which handles both adiabatic
and non-adiabatic calculations - Replaced
&output
namelist by&ad_output
and&nad_output
namelists, to handle adiabatic and non-adiabatic output, respectively - Replaced multiple
&shoot_grid
and&recon_grid
namelist with a single&grid
namelist. Grid construction is now handled as a "one-and-done" operation, rather than a sequence of operations - Parameters for
'HOM'
models are now specified in the&model
inlist - Split the
freq_frame
parameter into separatefreq_min_frame
andfreq_max_frame
parameters - Renamed the
ivp_solver
parameter todiff_solver
(difference solver) - Renamed the
'PER_DAY'
option for thefreq_units
parameter to'CYC_PER_DAY'
- The default
variables_set
parameter is now'GYRE'
-- a canonical set of variables specified in the equations.pdf document - Renamed the
'ZERO'
outer boundary option to'VACUUM'
- Replaced the
C
output item withbeta
Bug Fixes
x_i
andx_o
parameters in&grid
inlist now actually work
Release 4.4
Bug Fixes
- Added regularization back into the
MAGNUS_GL*
schemes, to restore their convergence for giant models - Fixed bug in reading LOSC files