[Mesa-users] Change_E_flag

[Subedi, Shiv Kumar]

Thank you Frank. Replacing the "use_hydrodynamic_surface_bcs" with "use_compression_outer_BC" worked.


Also, thank you Bill for your suggestion. I'm currently doing the sensitivity study using the mesa-r7624. As you suggested, I'll move on to the new set of inlists provided in mesa-r10108 "test_suite" examples for ccsn. But before that I wish to run the inlists from previous mesa version (mesa-r7624) to first see how the results of the sensitivity study change while just moving onto the new mesa version (mesa-r10108). I'm planning to incorporate the new improvements in mesa-r10108 i.e.  treatment of shocks using the approximate Riemann solver (HLLC method) and treatment of the effects of Rayleigh-Taylor instabilities as well in my old inlists to see the change in the sensitivity study due to these two improvements.



Moving on implementing the "use_compression_outer_BC" flag in the inlist, I'd a couple of other error messages as well for the following flags.


(A) dt_div_dt_Cougrant_limit - It has no replacement neither in the controls.defaults of mesa-r10108 and nor in the inlists of the ccsn section of MESA INSTRUMENT PAPER 4. I hereby commented it out.


(B) use_energy_conservation_form- I replaced by the flag "use_time_centering"


With above changes the fortran runtime errors were solved, however the code then stopped working with the following error message.


stopping because of problems dt < min_timestep_limit
terminated evolution: cannot find acceptable model
termination code: min_timestep_limit


Initially the min_timestep_limit was set to 10E-12, which I decreased to 10E-15. However the problem still persists.


Could you please suggest what's the problem here and how it can be solved?


Best,

Shiv Subedi

Graduate Student- Ohio University


________________________________
From: Francis Timmes <fxt44 at mac.com>
Sent: Monday, June 25, 2018 7:20 AM
To: Subedi, Shiv Kumar
Cc: Frank Timmes; mesa-users at lists.mesastar.org
Subject: Re: [Mesa-users] Change_E_flag

> Could you please suggest what would be the solution for the problem: Fortran runtime error: Cannot match namelist object name use_hydrodynamic_surface_bcs ?


http://mesa.sourceforge.net/controls_defaults.html#use_compression_outer_BC
controls defaults - SourceForge<http://mesa.sourceforge.net/controls_defaults.html#use_compression_outer_BC>
mesa.sourceforge.net
specifications for starting model ¶. NOTE: if you are loading a saved model, then the following initial values are NOT USED to modify the model. in particular, you cannot use these to change Y or Z of an existing model. if you want to do that, see star_job.defaults controls such as change_Y. however, these are reported in output as the initial ...



fxt





> On Jun 25, 2018, at 12:05 PM, Subedi, Shiv Kumar <ss383615 at ohio.edu> wrote:
>
>
>
> Hello Bill,
>
>
> I'd some homework for the "Change_E_flag" and "use_hydrodynamic_surface_BCs" in two mesa versions (mesa-r7624 and mesa-r10108). Here's is what I found.
>
>  For mesa-r7624:
> grep -r "use_hydrodynamic_surface_BCs" test_suite/
> test_suite/ccsn/inlist_edep:      !use_hydrodynamic_surface_BCs = .true.
> i.e. "use_hydrodynamic_surface_BCs" is present in the inlist_edep of "ccsn" in test_suite.
>
> Also,
>
> grep -r "change_E_flag" test_suite/
> test_suite/envelope_shock/inlist_start_shock:      change_E_flag = .true.
> test_suite/ccsn/inlist_edep:      change_E_flag = .true.
> i.e. "change_E_flag" is present in the inlist_edep of "ccsn" and the initial_start_shock of "envelope_shock" of test_suite.
>
> For mesa-r10108:
> Both the directories "envelope_shock" and "ccsn" are absent in the test_suite and the above variables are not present in any of other current directories in the test_suite.
>
>
> MESA INSTRUMENT PAPER 3
> I've been using the inlists of Mesa instrument paper 3 for my work on ccsn with mesa-r7624. In the "inlist_edep" of the inlists provided, above variables come into play as follows.
>
> &star_job
> change_E_flag = .true.
>  new_E_flag = .true.
>
> &controls
> ! equation controls
>       use_ODE_var_eqn_pairing = .true.
>       use_hydrodynamic_surface_BCs = .true.
>       use_dvdt_form_of_momentum_eqn = .true.
>       use_dPrad_dm_form_of_T_gradient_eqn = .true.
>
>       use_dedt_form_of_energy_eqn = .true.
>
>
> MESA INSTRUMENT PAPER 4
> I then looked into the "inlist_edep" of the inlists provided for "ccsn" in Mesa instrument paper 4 (mesa-r10108) for all the cases of Core-collapse SN models.
> There is no "E_flag" in the "star_job" section and the "controls" section has following variables.
>
> &controls
>   ! equation controls
>       use_ODE_var_eqn_pairing = .true.
>       use_dPrad_dm_form_of_T_gradient_eqn = .true.
>       use_dedt_form_of_energy_eqn = .true.
>
>       !use_momentum_outer_BC = .true.
>       use_fixed_vsurf_outer_BC = .true.
>       fixed_vsurf = 0d0
>
> i.e. there is no "use_hydrodynamic_surface_BCs" in the equation controls here, and comparing above I cannot say for 100% what variable replaces the " use_hydrodynamic_surface_BCs".
>
>
> Also, there is no mention of the above two variables in the controls defaults section in MESA web page
> http://mesa.sourceforge.net/controls_defaults.html#use_compression_outer_BC.
>
>
>
> Could you please suggest what would be the solution for the problem: Fortran runtime error: Cannot match namelist object name use_hydrodynamic_surface_bcs ?
> Also what is best solution if any of such problem arises for a variable which has no definition in the defaults files while upgrading a new mesa version?
>
> Best,
> Shiv Subedi
> Graduate Student- Ohio University
>
>
>
>
>
>
>
>
>
> From: Bill Paxton <paxton at kitp.ucsb.edu>
> Sent: Sunday, June 24, 2018 11:18 AM
> To: Subedi, Shiv Kumar
> Cc: Frank Timmes; mesa-users at lists.mesastar.org
> Subject: Re: [Mesa-users] Change_E_flag
>
> time for some detective work.   look in star/test_suite from the old version to find cases that use the old control name (use_hydrodynamic_surface_BCs) in their inlists.
>
> then look at the corresponding inlists in the newer version to see what has taken its place.
>
> let us know what you discover.
>
> bill
>
>
>
>
>> On Jun 24, 2018, at 4:13 PM, Subedi, Shiv Kumar via Mesa-users <mesa-users at lists.mesastar.org> wrote:
>>
>> !### use_hydrodynamic_surface_BCs
>>
>>          ! 1) gradient of compression vanishes at surface
>>             ! see Grott, Chernigovski, Glatzel, 2005.
>>             ! d_dm(d_dm(r^2*v)) = 0 at surface
>>             ! by continuity, this is d_dm(d_dt(1/rho)) = 0 at surface
>>             ! finite volume form is
>>             ! (1/rho(1) - 1/rho_start(1)) = (1/rho(2) - 1/rho_start(2))
>>             ! this BC determines the density for surface cell.
>>
>>          ! 2) black body radiation at surface
>>             ! L = 4*pi*R^2*boltz_sig*T^4
>>             ! this BC determines the temperature for the surface cell given R and L.
>>
>>       use_hydrodynamic_surface_BCs = .false.
>>
>> However controls.defaults in mesa-r10108 doesnot have any description of the above variable.

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.mesastar.org/pipermail/mesa-users/attachments/20180626/15cb4d53/attachment.html>