[Mesa-users] Creating a Highly Irradiated Super-Earth

[Bill Paxton]

excellent.  i missed that.   perhaps as a workaround you might try making the core larger so that the innermost temperature and density would be lower.   that might get around the eos problems you are having.   

it would be nice to know if the Chen & Rogers scheme still works in the current mesa.   too bad we don’t have a test case that corresponds to their work - that would be wonderful as a way to make sure the necessary parts of the code are continuing to work as expected.

-b


> On Jan 25, 2020, at 10:15 AM, Brian Jackson <bjackson at boisestate.edu> wrote:
> 
> I was basing my calculations on their work, and I *think* I inserted
> an inert core into the center of my planet. At least, the .mod file
> (LF14_planet_relaxedM_20_ME_10-percent.mod) mentions a core:
> 
> R_center      1.3120716706688340D+09  ! radius of core (cm).  R/Rsun,
> avg core density (g/cm^3):      1.8852146191971522D-02
> 1.1357573371457304D+01
> 
> Did I misunderstand this line?
> 
> Thanks,
> Brian
> 
> On Fri, Jan 24, 2020 at 9:59 PM Bill Paxton <paxton at kitp.ucsb.edu> wrote:
>> 
>> 
>>> On Jan 24, 2020, at 4:09 PM, Brian Jackson <bjackson at boisestate.edu> wrote:
>>> 
>>> Is it possible to revert to a version of MESA that could successfully
>>> navigate these calculations? I see that Chen & Rogers (2016) used
>>> version 7623 to model highly irradiated super-Earths.
>> 
>> Note that they created a mesa model with an inert core and therefore were not trying to model the center locations where you are having problems.   You might try doing something similar to what they did.
>> 
>> b
>> 
>> 
>> 
>> 
>> 
>> 
>> 
>>> I was also able
>>> to use an earlier version to do similar calculations.
>>> 
>>> I know that I can download earlier MESA versions, but the earlier
>>> versions of the SDKs do not appear to be available. Any suggestions?
>>> 
>>> Thanks,
>>> Brian
>>> 
>>> On Fri, Jan 24, 2020 at 2:41 PM Brian Jackson <bjackson at boisestate.edu> wrote:
>>>> 
>>>> It's actually reassuring. At least I wasn't goofing something up. I'll
>>>> figure out some science workaround.
>>>> 
>>>> Thanks very much for the quick response,
>>>> Brian
>>>> 
>>>> On Fri, Jan 24, 2020 at 2:37 PM Bill Paxton <paxton at kitp.ucsb.edu> wrote:
>>>>> 
>>>>> this is definitely running into EOS problems.  (negative gamma1 -> imaginary sound speed)
>>>>> 
>>>>> it may well be related to a blending glitch combining OPAL eos and PC eos.
>>>>> 
>>>>> the solution is not likely to be trivial, so don’t hold your breath.
>>>>> 
>>>>> we’ll work on it, but it may take some significant time to fix.
>>>>> 
>>>>> sorry for the disappointing news.
>>>>> 
>>>>> b
>>>>> 
>>>>> 
>>>>> 
>>>>> 
>>>>> 
>>>>>> On Jan 24, 2020, at 11:53 AM, Brian Jackson <bjackson at boisestate.edu> wrote:
>>>>>> 
>>>>>> I've conducted additional experiments, and I think I'm starting to see
>>>>>> what the issues might be.
>>>>>> 
>>>>>> I've attached inlists and a model for a slightly different approach.
>>>>>> The model calculation behaves fine until about 4 Gyrs out, when the
>>>>>> timestep suddenly plunges and the code fails to complete the run. For
>>>>>> some reason, the central degeneracy parameter goes haywire shortly
>>>>>> before the run ends -- see attached History_Plot.png.
>>>>>> 
>>>>>> Looking at the 2011 MESA instrument
>>>>>> (https://ui.adsabs.harvard.edu/abs/2011ApJS..192....3P/abstract), it
>>>>>> seems the model core may be nearing a transition in the equation of
>>>>>> state (log(density) = -0.19, log(temp) = 3.79 -- see attached Figure
>>>>>> 1). Is there a reason why that transition would severely and abruptly
>>>>>> modify the degeneracy parameter? Is there a way to smooth this
>>>>>> evolutionary phase?
>>>>>> 
>>>>>> Or have I completely misunderstood what's going on?
>>>>>> 
>>>>>> Thanks,
>>>>>> Brian
>>>>>> 
>>>>>> On Wed, Jan 22, 2020 at 5:58 PM Brian Jackson <bjackson at boisestate.edu> wrote:
>>>>>>> 
>>>>>>> Dear MESA Users,
>>>>>>> I am still struggling to create highly-irradiated super-Earths using MESA.
>>>>>>> 
>>>>>>> I have attached the inlists and saved models I'm using to create a
>>>>>>> 20-Earth mass planet with an envelope mass fraction of 20% (i.e., a
>>>>>>> core mass of 16 Earth masses) and an irradiation 20 times Earth's
>>>>>>> (i.e., 2.72e7 ergs/s/cm^2).
>>>>>>> 
>>>>>>> I attempt to create the planet by irradiating a gas giant planet and
>>>>>>> then relaxing the mass down to my desired mass. (I've found doing it
>>>>>>> the other way around leads to badness.) I've tried tweaking lots of
>>>>>>> settings, but nothing seems to work. In the end, I wind up with a
>>>>>>> timestep below the suggested minimum of 9.9999999999999995D-07.
>>>>>>> 
>>>>>>> I am using Mac OS 10.13.1, MESA 11701, and the most recent version of
>>>>>>> the SDK (August 30 2019).
>>>>>>> 
>>>>>>> Please forgive me if I forgot any required information.
>>>>>>> 
>>>>>>> Thanks,
>>>>>>> Brian
>>>>>> <inlist><inlist_pgstar><history_columns.list><LF14_planet_relaxedM_20_ME_10-percent.mod><Paxton2011_Fig1.png><History_Plot.png>_______________________________________________
>>>>>> mesa-users at lists.mesastar.org
>>>>>> https://lists.mesastar.org/mailman/listinfo/mesa-users
>>>>>> 
>>>>> 
>>> _______________________________________________
>>> mesa-users at lists.mesastar.org
>>> https://lists.mesastar.org/mailman/listinfo/mesa-users
>>> 
>> 
> _______________________________________________
> mesa-users at lists.mesastar.org
> https://lists.mesastar.org/mailman/listinfo/mesa-users
>