[FLASH-USERS] value of Ye for single species runs

Slavin, Jonathan jslavin at cfa.harvard.edu
Tue Oct 6 09:54:07 EDT 2015


Hi Ernesto,

Thanks for your reply.  I'm still not clear though on whether there is a
problem with doing split hydro and unsplit diffusion.  I would think that
it's not a problem since the usual practice is use operator splitting such
that the hydro evolution is done separately from the diffusion, but I don't
know that for a fact.  Can anyone comment?

Regards,
Jon

On Mon, Oct 5, 2015 at 5:25 PM, Ernesto Zurbriggen <
ernesto at mail.oac.uncor.edu> wrote:

> Jon,
> integrate using Hydro/HydroMain/split or Hydro/HydroMain/unsplit is very
> different. That is if you put the shortcut "+uhd" in the setup line or not
> (in the hydro case). Things I told you are for the unsplit case and using
> Diffuse/DiffuseMain/Unsplit.
>
> While for the case Hydro/HydroMain/split and Diffuse/DiffuseMain/split
> (maybe also with Diffuse/DiffuseMain/unsplit) the parameter
> "addThermalFlux" is unkown for the code.
>
> On the other hand, every time you modify something in your flash.par or
> Config file the best would be run the setup line again.
>
>
>
>
> Quoting "Slavin, Jonathan" <jslavin at cfa.harvard.edu>:
>
> Hi Ernesto,
>>
>> I just tried the run again with the parameter addThermalFlux set to False
>> and got in the log file:
>> WARNING: Ignored Parameters :
>> These parameters were found in the flash.par file, but they were
>>  not declared in any Config file for the simulation!
>> ...
>> addThermalFlux
>>
>> so it seems that the code doesn't use addThermalFlux.  Could the parameter
>> have some other name?  I didn't rerun setup or build, but I wouldn't think
>> that would be necessary for a runtime parameter change.  Do you think it's
>> a problem to use split PPM and unsplit diffusion?  I think the diffusion
>> is
>> done separately from the hydro via operator splitting.
>>
>> Regards,
>> Jon
>>
>> On Mon, Oct 5, 2015 at 11:53 AM, Ernesto Zurbriggen <
>> ernesto at mail.oac.uncor.edu> wrote:
>>
>> Hi Jon,
>>> as I understand from what you told us, you should put
>>> addThermalFlux=.false.
>>> Because of you are already considering thermal conduction in the Diffuse
>>> unit, so you need the code doesn't enter the subroutine
>>> hy_uhd_addThermalFluxes.F90. In case of keeping addThermalFlux=.true.,
>>> you
>>> are considering thermal conduction two times. You need to turn off one of
>>> them.
>>>
>>>
>>> I think of the code has a bug putting addThermalFlux=.true. by default.
>>> That would be all.
>>>
>>> Best!
>>>
>>>
>>>
>>> Quoting "Slavin, Jonathan" <jslavin at cfa.harvard.edu>:
>>>
>>> ​Hi Ernesto,
>>>
>>>>
>>>> I don't specify the runtime parameter addThermalFlux.  Instead I set
>>>> diff_useEleCond, useConductivity, useDiffuse, and useDiffuseTherm to
>>>> .true.
>>>> and specify
>>>> ​REQUIRES
>>>> physics/materialProperties/Conductivity/ConductivityMain/PowerLaw
>>>> REQUESTS physics/Diffuse/DiffuseMain/Unsplit
>>>> ​Should I change one of those?​
>>>>
>>>> ​I note that in the log file that
>>>> hy_fluxrespresentation = fluxes
>>>> I'm not sure exactly what that means.  My goal here is to verfy that
>>>> ​thermal conduction is operating as I expect by comparing evaporation
>>>> of a
>>>> spherical cloud in 2D cylindrical symmetry with analytical results
>>>> derived
>>>> for steady flow including saturation of thermal conduction (Dalton &
>>>> Balbus
>>>> 1993).
>>>>
>>>> Regards,
>>>> Jon
>>>>
>>>> On Sat, Oct 3, 2015 at 1:00 PM, <flash-users-request at flash.uchicago.edu
>>>> >
>>>> wrote:
>>>>
>>>> From: Ernesto Zurbriggen <ernesto at mail.oac.uncor.edu>
>>>>
>>>>> To: flash-users at flash.uchicago.edu
>>>>> Cc:
>>>>> Date: Fri, 02 Oct 2015 17:45:40 -0300
>>>>> Subject: Re: [FLASH-USERS] value of Ye for single species runs
>>>>> Jon,
>>>>> looking at the end of the subroutine Eos/EosMain/Eos_getAbarZbar.F90
>>>>> you
>>>>> can see for a single species run Ye=Zbar/Abar too. And I think that is
>>>>> not
>>>>> a stub implementation for the single-species gamma-law EOS.
>>>>>
>>>>> On the other hand, if you turn on the thermal conduction managed by
>>>>> Diffuse/Split or Diffuse/Unplit units, I think the runtime parameter
>>>>> "addThermalFlux" must be false. On the contrary, you might be taking
>>>>> into
>>>>> account thermal conduction twice, within the Diffuse unit and Hydro
>>>>> unit
>>>>> via hy_uhd_addThermalFluxes.F90
>>>>>
>>>>> Best!
>>>>>
>>>>>
>>>>>
>>>>
>>>>
>>>> --
>>>> ________________________________________________________
>>>> Jonathan D. Slavin                 Harvard-Smithsonian CfA
>>>> jslavin at cfa.harvard.edu       60 Garden Street, MS 83
>>>> phone: (617) 496-7981       Cambridge, MA 02138-1516
>>>> cell: (781) 363-0035             USA
>>>> ________________________________________________________
>>>>
>>>>
>>>>
>>>
>>> ----------------------------------------------------------------
>>> This message was sent using IMP, the Internet Messaging Program.
>>>
>>>
>>>
>>
>> --
>> ________________________________________________________
>> Jonathan D. Slavin                 Harvard-Smithsonian CfA
>> jslavin at cfa.harvard.edu       60 Garden Street, MS 83
>> phone: (617) 496-7981       Cambridge, MA 02138-1516
>> cell: (781) 363-0035             USA
>> ________________________________________________________
>>
>>
>
>
> ----------------------------------------------------------------
> This message was sent using IMP, the Internet Messaging Program.
>
>


-- 
________________________________________________________
Jonathan D. Slavin                 Harvard-Smithsonian CfA
jslavin at cfa.harvard.edu       60 Garden Street, MS 83
phone: (617) 496-7981       Cambridge, MA 02138-1516
cell: (781) 363-0035             USA
________________________________________________________
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://flash.rochester.edu/pipermail/flash-users/attachments/20151006/d7a79bca/attachment-0001.htm>


More information about the flash-users mailing list