[Mesa-users] Limiting the final radius of a simulation

[Ebraheem Farag]

Hello Ali,

It looks to me like you are currently only using "when_to_stop_atol =
0.0001d0", and the error is calculated using both "when_to_stop_atol" and
"when_to_stop_rtol":

"The system will automatically redo with a smaller timestep to hit a
stopping target. It calculates the following “error” term and retries if it
is > 1"

error = abs(value - target_value)/ &        (when_to_stop_atol +
when_to_stop_rtol*max(abs(value),abs(target_value)))


Since the default "when_to_stop_rtol = 1d99", your stopping condition is
not being satisfied.

Try setting  "when_to_stop_atol = 0.0001d0", with "when_to_stop_rtol = 0d0"
or vice versa, as per your goals. Just check the error equation to make
sure you're getting what you want!

In my inlists, I use:

! tolerances on stopping conditions

      when_to_stop_rtol = 1d-3

      when_to_stop_atol = 1d-3



Cheers,
EbF


On Wed, Feb 15, 2023 at 12:54 PM Ali Pourmand <pourmand at ualberta.ca> wrote:

> Dear Ebraheem and Francis,
>
> Thank you very much for your replies!
> I had one question though, how do we specify what variable we are using
> "when_to_stop_rtol" for? I tried it several times but it didn't have any
> effect.
> I have attached a sample inlist I am using.
>
>
> On Wed, Feb 15, 2023 at 12:02 PM Ebraheem Farag <ekfarag at asu.edu> wrote:
>
>> Hello Ali,
>>
>> If you're trying to stop at a specific radius, there are a few &controls
>> <https://urldefense.com/v3/__https://docs.mesastar.org/en/release-r22.11.1/reference/controls.html*photosphere-r-upper-limit__;Iw!!IKRxdwAv5BmarQ!d9I8umjyHg_Q5_hKiVe8qANnh3lSmfTP1-XEoKkhASa17Whom0yL-TS_x1TMC8Xtdhm94pM92aWt9s9nKOU$> that
>> can help:
>>
>> stop when photosphere_r is greater than this limit, in Rsun units
>>
>> photosphere_r_upper_limit = 1d99
>>
>>
>> stop when photosphere_r is less than this limit, in Rsun units
>>
>> photosphere_r_lower_limit = -1d99
>>
>>
>>
>> You can set a tolerance on your stopping condition using the &controls
>> <https://urldefense.com/v3/__https://docs.mesastar.org/en/release-r22.11.1/reference/controls.html*when-to-stop-rtol__;Iw!!IKRxdwAv5BmarQ!d9I8umjyHg_Q5_hKiVe8qANnh3lSmfTP1-XEoKkhASa17Whom0yL-TS_x1TMC8Xtdhm94pM92aWtLZtw6Vo$>
>>  'when_to_stop_rtol':
>>
>> Relative error criteria when hitting stop target time. The system will
>> automatically redo with a smaller timestep to hit a stopping target. It
>> calculates the following “error” term and retries if it is > 1.
>>
>> error = abs(value - target_value)/ &        (when_to_stop_atol + when_to_stop_rtol*max(abs(value),abs(target_value)))
>>
>> when_to_stop_rtol = 1d99
>>
>>
>> in your case you could set.
>>
>> when_to_stop_rtol = 1d-2 # for 0.01 precision, I use 1d-3 typically...
>>
>>
>> 'delta_lgR_limit' adjusts how large of a logarithmic change in the
>> stellar radius is allowed in each solver timestep, so it might help the
>> time resolution of your models, but won't exactly improve your stopping
>> condition.
>>
>> I hope this helps!
>>
>> -EbF
>>
>>
>> On Wed, Feb 15, 2023 at 11:24 AM Ali Pourmand via Mesa-users <
>> mesa-users at lists.mesastar.org> wrote:
>>
>>> Hello,
>>>
>>>  I want to stop several different MESA runs at the same final
>>> (photospheric) radius (with a certain precision), but since the physics is
>>> slightly different in each, I can't stop at the exact same radius. I
>>> realize that one could decrease the timesteps of evolution to achieve this,
>>> but was wondering if there is an easier way to achieve this.
>>>
>>> I am already using the stop to limit the photospheric radius command,
>>> but it always stops with a certain precision.
>>>
>>> What I want is a way to tell the code to stop when the photosphere
>>> reaches 7+-0.01 solar radii , is there any command that could achieve that?
>>> I tried delta_lgR_limit, but I don't think that does what I have in my
>>> mind.
>>>
>>> best
>>> Ali Pourmand
>>> M.Sc. Student
>>> University of Alberta
>>> _______________________________________________
>>> mesa-users at lists.mesastar.org
>>>
>>> https://urldefense.com/v3/__https://lists.mesastar.org/mailman/listinfo/mesa-users__;!!IKRxdwAv5BmarQ!YmagYdCG8_pS3vclu4IVf5A1KVI4YnXJrDXsGkGxNAo1nXX0t8RniyeS5XH7dQ03YuyYGOKa06ZVymwG9BU-jP-j$
>>>
>>>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.mesastar.org/pipermail/mesa-users/attachments/20230215/ddf4b29d/attachment.htm>