[mesa-users] dynamo effect on stellar str

Sat Jul 30 12:19:03 EDT 2016

Here's something to keep in mind concerning the extra dP_dm term set by other_momentum.  The routine is called before the start of the newton iterations and the value it sets for extra_dPdm is then held constant during the iterations.  This is the simplest approach and works fine if the term is changing on a timescale significantly longer than the timesteps.  But it can lead to instabilities if the timesteps are similar to the timescale or smaller -- those instabilities will probably show up as failures in the newton iterations.  The solution is to make the term "implicit" so that it is evaluated at each newton iteration and can adapt to changes in local values such as T and Rho for the cell and its immediate neighbors (see, e.g., other_energy and other_energy_implicit).  If you find we need to do that, we'll need to add a new hook that returns both the extra_dPdm and partials for use in the Jacobian that is the core of the newton solve.

But try it the simple "explicit" way first.  Just keep in mind the possible issues for the newton solver if the timescale for the term gets near the timesteps.

-B 

On Jul 30, 2016, at 4:39 AM, Valery Pipin wrote:

> Hi!
> 
> I think yes. I found that star/other contain modules for additional mixing and also additional terms in momentum equation.
> Thus, if nz correspond to the number of mesh points of the output files I can compute gradient of magnetic pressure myself and read it to the module other_momentum.
> The same can be done for additional diffusion.
> 
> Later I shall study if it is possible to transfer approach given in the paper by  Feiden & Chaboyer. Modifications goes to equation of state,
> equation of energy, mixing length and momentum.
> Best regards!
> Valery
> 
>> Hi Valery,
>> 
>> In light of Brian's comment, I'm wondering if just adding an extra pressure term to the hydrostatic equilibrium equation will be adequate to do anything useful.
>> 
>> Before I spend time adding something of no use, can you reply to this concern of mine?  I'm not going to take care of thermodynamic consistency for a new component of the pressure, so is that something you are prepared to deal with somehow?  Or do you plan to ignore the issue for now?
>> 
>> -Bill
>> 
>> 
>> On Jul 29, 2016, at 11:13 AM, Brian Chaboyer wrote:
>> 
>>> Hi,
>>> 
>>> If you are going to include the effects of a magnetic pressure, you will
>>> be have to be careful to ensure thermodynamic consistency.  One way to
>>> approach this for magnetic fields is discussed in Feiden & Chaboyer
>>> 2012, ApJ, 761, 30.
>>> 
>>> cheers,
>>> Brian
>>> 
>>> 
>>> On 7/29/16 12:55 PM, Bill Paxton wrote:
>>>> Hi Craig,
>>>> 
>>>> Good to hear you are also interested.  My goal, as I explained to Valery, is to provide the hooks in mesa/star so you crazy guys can try things -- it's the "E" in MESA at the level of things that require new software development.  You and Manos are already experienced with this idea.
>>>> 
>>>> My hope is we can do the current task by adding a new pressure component so that the total pressure in mesa/star is P = Pgas + Prad + Pmag.
>>>> I'll modify the guts of mesa star to include Pmag, but the actual code to calculate it at each cell will be done by a "hook" routine so it can be developed independently of the mainline mesa code.  Does that sound like it would give you enough rope to take care of the rest of the hanging?  ;)
>>>> 
>>>> Cheers,
>>>> Bill
>>>> 
>>>> 
>>>> 
>>>> 
>>>> 
>>>> 
>>>> On Jul 29, 2016, at 8:57 AM, Craig Wheeler wrote:
>>>> 
>>>>> We had meant to add our prescription for the MRI dynamo
>>>>> to MESA. Perhaps there can be some joining of forces here.
>>>>> 
>>>>> Craig Wheeler (for Manos, who did the work)
>>>>> 
>>>>> On 7/28/16 9:31 PM, Valery Pipin wrote:
>>>>>> Dear  List,
>>>>>> 
>>>>>> I wonder if MESA has a possibility to check effects of large-scale
>>>>>> magnetic field and flow on the stellar structure.
>>>>>> 
>>>>>> I have computed a set of nonlinear dynamo models. They include the 3D
>>>>>> mean-field dynamo coupled with angular momentum balance and heat transport.
>>>>>> 
>>>>>> The model use the reference thermodynamic  structure of a star 0.3Ms of
>>>>>> age 1Gyr  which was computed with help of MESA using the solar composition.
>>>>>> 
>>>>>> The equation for the mean-field heat transport was solved for
>>>>>> perturbation of the mean entropy caused by effects of the global
>>>>>> rotation and magnetic field.
>>>>>> 
>>>>>> In 1D stellar structure code the effect of mixing due to meridional
>>>>>> circulation can be treated as an addition to diffusion and effect of
>>>>>> magnetic field via the averaged addition to pressure.
>>>>>> 
>>>>>> Both factors vary along the radius. Can I add them to the code and in
>>>>>> what place. I would appreciate if you have some example.
>>>>>> 
>>>>>> The test suite contains example for 15M star with dynamo. I don't
>>>>>> understand where parameters of the magnetic field are defined?
>>>>>> 
>>>>>> Thanks
>>>>>> 
>>>>>> Valery
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> ------------------------------------------------------------------------------
>>>>>> _______________________________________________
>>>>>> mesa-users mailing list
>>>>>> mesa-users at lists.sourceforge.net
>>>>>> https://lists.sourceforge.net/lists/listinfo/mesa-users
>>>>> -- 
>>>>> J. Craig Wheeler
>>>>> Samuel T. and Fern Yanagisawa Regents Professor of Astronomy
>>>>> President, American Astronomical Society, 2006-2008
>>>>> Department of Astronomy
>>>>> 2515 Speedway, Stop C1400
>>>>> Austin, TX 78712-1205
>>>>> 512-471-6407
>>>>> http://www.as.utexas.edu/~wheel
>>>>> http://ebooks.cambridge.org/ebook.jsf?bid=CBO9780511536625
>>>>> http://www.amazon.com/Krone-Ascending-ebook/dp/B00926QEGW
>>>>> 
>>>>> 
>>>>> ------------------------------------------------------------------------------
>>>>> _______________________________________________
>>>>> mesa-users mailing list
>>>>> mesa-users at lists.sourceforge.net
>>>>> https://lists.sourceforge.net/lists/listinfo/mesa-users
>>>> ------------------------------------------------------------------------------
>>>> _______________________________________________
>>>> mesa-users mailing list
>>>> mesa-users at lists.sourceforge.net
>>>> https://lists.sourceforge.net/lists/listinfo/mesa-users
>>> 
>>> 
>>> ------------------------------------------------------------------------------
>>> _______________________________________________
>>> mesa-users mailing list
>>> mesa-users at lists.sourceforge.net
>>> https://lists.sourceforge.net/lists/listinfo/mesa-users
> 
> 