[Mesa-users] Re3: Setup for overshooting
Ebraheem Farag
ekfarag at asu.edu
Wed Apr 26 18:21:38 UTC 2023
Hello Wenyu,
Thanks for sharing, I didn't realize you were running in r12778!
In both plots, the burning looks qualitatively identical. The Carbon shell
and Oxygen core burning look convective to me in both cases, with or
without hatching. Are you sure this isn't a plotting issue? A log-abundance
profile would show the difference, as convective burning leads to flat
composition gradients in the burning region.
These Kippenhahn diagrams only seem to show the evolution through O
burning, but you previously stated that your no overshoot models begin to
stall during Si burning? Again, there is not enough information presented
to say anything concrete here, especially since both kip plots look
identical short of hatching. Do both of these models stall during Si
burning, as before we were comparing a model with and without overshooting
in 'burn_Z', and now you are sharing two models with effectively no
overshooting on 'burn_Z'?
-EbF
On Wed, Apr 26, 2023 at 10:41 AM 辛文宇 <xinwenyu16 at mails.ucas.ac.cn> wrote:
> Hi EbF,
>
>
> Here are two plots for no overshooting. In 28M_none1, there are no
> convective zone during C, O burning for both shell and core.
>
> With a very small f and f0, there are convective zones. It seems in
> MESA-12778, the overshoot_scheme(1) = 'none' make the convection
>
> not work.
>
>
> In 28M_none, I use:
>
>
> overshoot_scheme(1) = 'exponential' ! one of exponential or step
> overshoot_zone_type(1) = 'burn_Z' ! one of burn_H, burn_He,
> burn_Z, nonburn, any
> overshoot_zone_loc(1) = 'shell' ! one of core, shell, any
> overshoot_bdy_loc(1) = 'any' ! one of bottom, top, any
> overshoot_f(1) = 1d-7
> overshoot_f0(1) = 1d-10
>
>
> overshoot_scheme(2) = 'exponential' ! one of exponential or step
> overshoot_zone_type(2) = 'any' ! one of burn_H, burn_He,
> burn_Z, nonburn, any
> overshoot_zone_loc(2) = 'any' ! one of core, shell, any
> overshoot_bdy_loc(2) = 'any' ! one of bottom, top, any
> overshoot_f(2) = 0.01
> overshoot_f0(2) = 0.004
>
>
> In 28M_none1, I use:
> overshoot_scheme(1) = 'none' ! one of exponential or step
> overshoot_zone_type(1) = 'burn_Z' ! one of burn_H, burn_He,
> burn_Z, nonburn, any
> overshoot_zone_loc(1) = 'shell' ! one of core, shell, any
> overshoot_bdy_loc(1) = 'any' ! one of bottom, top, any
> overshoot_f(1) = 0.01
> overshoot_f0(1) = 0.004
>
> overshoot_scheme(2) = 'exponential' ! one of exponential or step
> overshoot_zone_type(2) = 'any' ! one of burn_H, burn_He,
> burn_Z, nonburn, any
> overshoot_zone_loc(2) = 'any' ! one of core, shell, any
> overshoot_bdy_loc(2) = 'any' ! one of bottom, top, any
> overshoot_f(2) = 0.01
> overshoot_f0(2) = 0.004
>
> Thanks,
>
> Wenyu
>
> -----原始邮件-----
> *发件人:*"Ebraheem Farag" <ekfarag at asu.edu>
> *发送时间:*2023-04-25 17:12:48 (星期二)
> *收件人:* "辛文宇" <xinwenyu16 at mails.ucas.ac.cn>
> *抄送:* taylormorgan32 at gmail.com, mesa-users at lists.mesastar.org
> *主题:* Re: Re2: [Mesa-users] Setup for overshooting
>
> Hello Wenyu,
>
> >But when I turn off the overshooting, other with scheme:none or
> the overshoot_mass_full_off(1),
> When you turn off overshooting just for 'burn_Z' shells?
>
> >Are there any numerical or physical explanation for this change?
>
> It is difficult to speculate without more information (a plot or plots, an
> inlist or model directory, and/or a terminal output file). Overshoot is but
> one form of mixing and any number of other controls could contribute to the
> issue, including numerical resolution.
>
> A Kippenhahn diagram or log_abundance profile might help to identify the
> presence of a shell merger, if that is indeed the issue. Turning on pgstar
> could be quite useful!
>
> -EbF
>
> On Tue, Apr 25, 2023 at 1:20 AM 辛文宇 <xinwenyu16 at mails.ucas.ac.cn> wrote:
>
>> Hi Morgan and EbF,
>>
>> Thanks for your suggestions.
>>
>> I have tried your ways for turn off the overshooting.
>>
>>
>> But here is another question.
>>
>> I am running a model with 28Msun from ZAMS to Fe core collapse.
>>
>> With the overshooting, the model can go through smoothly.
>>
>> But when I turn off the overshooting, other with scheme:none or
>> the overshoot_mass_full_off(1),
>>
>> there will be too much retries and as a results, the timestep of the
>> model will decrease to 1E-14 during the
>>
>> Si burning stage.
>>
>>
>> Are there any numerical or physical explanation for this change?
>>
>>
>> Thank you very much!
>>
>> Wenyu
>>
>>
>> -----原始邮件-----
>> *发件人:*"Ebraheem Farag" <ekfarag at asu.edu>
>> *发送时间:*2023-04-23 22:34:11 (星期日)
>> *收件人:* "Morgan Taylor" <taylormorgan32 at gmail.com>
>> *抄送:* "辛文宇" <xinwenyu16 at mails.ucas.ac.cn>, mesa-users at lists.mesastar.org
>> *主题:* Re: [Mesa-users] Setup for overshooting
>>
>> Hello Wenyu,
>>
>> >"I also checked the file in controls.defaults file, there is no choice
>> of 'none'."
>>
>> Though not listed in controls.defaults, 'none' should work. I confirmed
>> in r22.11.1 that the scheme I shared works, however I did making a
>> spelling error:
>>
>> it should be 'burn_Z', not 'burn_z'. Try this:
>>
>> ! no overshooting above Z burning shell (e.g. C, O, Si), overshoot e-fold
>> distance = 0.004*Hp
>>
>> overshoot_scheme(1) = 'none' ! one of exponential or step
>>
>> overshoot_zone_type(1) = 'burn_Z' ! one of burn_H, burn_He,
>> burn_Z, nonburn, any
>>
>> overshoot_zone_loc(1) = 'shell' ! one of core, shell, any
>>
>> overshoot_bdy_loc(1) = 'any' ! one of bottom, top, any
>>
>> overshoot_f(1) = 0.008
>>
>> overshoot_f0(1) = 0.004
>>
>>
>>
>> ! overshooting for all shells, superseded by overshoot_scheme(1) for
>> burning Z shells, overshoot e-fold distance = 0.01*Hp
>>
>> overshoot_scheme(2) = 'exponential' ! one of exponential or step
>>
>> overshoot_zone_type(2) = 'any' ! one of burn_H, burn_He, burn_Z,
>> nonburn, any
>>
>> overshoot_zone_loc(2) = 'any' ! one of core, shell, any
>>
>> overshoot_bdy_loc(2) = 'any' ! one of bottom, top, any
>>
>> overshoot_f(2) = 0.015
>>
>> overshoot_f0(2) = 0.005 ! 0.0075 for 0.5*f
>>
>>
>>
>> -EbF
>>
>> On Sun, Apr 23, 2023 at 7:30 AM Morgan Taylor via Mesa-users <
>> mesa-users at lists.mesastar.org> wrote:
>>
>>> Hi Wenyu,
>>>
>>> One suggestion is to specify the location/type of overshoot you want
>>> included.
>>> For example, here are some inlist controls I placed on a model to focus
>>> on core overshooting:
>>>
>>> overshoot_zone_type(1)='any'
>>>
>>>
>>> overshoot_zone_loc(1)='core'
>>>
>>>
>>> overshoot_f(1) = 0.016
>>>
>>>
>>> overshoot_f0(1) = 0.008
>>>
>>>
>>> overshoot_mass_full_on(1) = 0.01
>>>
>>>
>>> overshoot_mass_full_off(1) = 0.4
>>>
>>>
>>>
>>>
>>> overshoot_zone_type(1) is the 1st specified "type" which can be burn_H,
>>> burn_He, burn_Z, nonburn or any.
>>> overshoot_zone_loc(1) is the 1st area that wants the zone location to be
>>> in the core.
>>> overshoot_mass_full_on(1) =0.01 says to begin overshooting when the star
>>> mass is above 0.01 Msun.
>>> overshoot_mass_full_off(1) is where you can tell overshooting to shut
>>> off. Since I focused on the core for a low mass star, I chose to turn off
>>> overshooting at 0.4Msun. You can change this to what you think is best.
>>> You could also just try setting the zone_loc to 'core' and see if that
>>> alone will avoid overshooting in the shell.
>>>
>>> Hope this helps,
>>>
>>> Morgan
>>>
>>> On Sun, Apr 23, 2023 at 7:10 AM 辛文宇 <xinwenyu16 at mails.ucas.ac.cn> wrote:
>>>
>>>> Hi Morgan,
>>>>
>>>>
>>>> Thank you very much for the suggestion.
>>>>
>>>>
>>>> If I just want to turn off the overshooting of C, O, Si burning shell,
>>>>
>>>> but keep the overshooting of other places as they are, how can I
>>>> realize it?
>>>>
>>>>
>>>> Thanks,
>>>>
>>>> Wenyu
>>>>
>>>>
>>>> -----原始邮件-----
>>>> *发件人:*"Morgan Taylor" <taylormorgan32 at gmail.com>
>>>> *发送时间:*2023-04-23 21:18:24 (星期日)
>>>> *收件人:* "辛文宇" <xinwenyu16 at mails.ucas.ac.cn>
>>>> *抄送:* "Ebraheem Farag" <ekfarag at asu.edu>, mesa-users at lists.mesastar.org
>>>> *主题:* Re: [Mesa-users] Setup for overshooting
>>>>
>>>> Hi Wenyu,
>>>>
>>>> >Is there any other way for no overshooting?
>>>>
>>>> Yes, in controls.defaults, there's this little segment:
>>>>
>>>> ! min_overshoot_q
>>>>
>>>> ! ~~~~~~~~~~~~~~~
>>>>
>>>> ! Overshooting is only allowed at locations with mass ``m >=
>>>> min_overshoot_q * mstar``.
>>>>
>>>> ! E.g., if ``min_overshoot_q = 0.1``, then only the outer 90% by
>>>> mass can have overshooting.
>>>>
>>>> ! This provides a simple way of suppressing bogus center
>>>> overshooting in which a small
>>>>
>>>> ! convective region at the core can produce excessively large
>>>> overshooting because of
>>>>
>>>> ! a large pressure scale height at the center.
>>>>
>>>>
>>>>
>>>>
>>>> So if you set min_overshoot_q = 1.0, you will turn off overshoot :)
>>>>
>>>> Morgan
>>>>
>>>> On Sun, Apr 23, 2023 at 5:52 AM 辛文宇 <xinwenyu16 at mails.ucas.ac.cn>
>>>> wrote:
>>>>
>>>>> Hi Ebraheem,
>>>>>
>>>>>
>>>>> Thanks very much for the suggestions.
>>>>>
>>>>> >>"Another question is that if I want to test with no overshooting for
>>>>> shell burning, can I set f = 0, or I need to give a very small value
>>>>> for f. Because f and f0 shoould >0."
>>>>>
>>>>> > You can do this by setting: overshoot_scheme(1) = 'none' , instead
>>>>> of 'exponential' in the &controls I shared above.
>>>>>
>>>>> I try the overshoot_scheme(1) = 'none', but there is warming of
>>>>> Invalid overshoot_scheme:none.
>>>>>
>>>>>
>>>>> I also checked the file in controls.defaults file, there is no choice
>>>>> of 'none'.
>>>>>
>>>>>
>>>>> Is there any other way for no overshooting?
>>>>>
>>>>>
>>>>> Thanks,
>>>>>
>>>>>
>>>>> Wenyu
>>>>>
>>>>>
>>>>> -----原始邮件-----
>>>>> *发件人:*"Ebraheem Farag" <ekfarag at asu.edu>
>>>>> *发送时间:*2023-04-20 18:11:27 (星期四)
>>>>> *收件人:* "辛文宇" <xinwenyu16 at mails.ucas.ac.cn>
>>>>> *抄送:* mesa-users at lists.mesastar.org
>>>>> *主题:* Re: [Mesa-users] Setup for overshooting
>>>>>
>>>>> Hello Wenyu,
>>>>>
>>>>> I believe the order of precedence for overshooting is such
>>>>> that overshoot_scheme(1) supersedes overshoot_scheme(2). In the
>>>>> current form you have written, overshoot_scheme(2) will not take effect
>>>>> because overshoot_scheme(1) will overwrite it. By switching the ranking, so
>>>>> that your 'burn_z' overshoot scheme is (1), and your 'any' routine for all
>>>>> other regions is set to (2), your scheme should work as intended.
>>>>> ----
>>>>> Here is a pasted explanation from a previous email of how overshoot
>>>>> functions pasted from the overshooting section of
>>>>> $MESA_DIR/star/defaults/controls.defaults:
>>>>>
>>>>> Overshooting works by taking the diffusion mixing coefficient at the
>>>>> edge of the convection zone and extending it beyond the zone. But – and
>>>>> here’s the issue – at the exact edge of the zone the mixing coefficient
>>>>> goes to 0. So we don’t want that. Instead we want the value of the mixing
>>>>> coeff NEAR the edge, but not AT the edge. The “f0” parameter determines the
>>>>> exact meaning of “near” for this. It tells the code how far back into the
>>>>> zone to go in terms of scale height. The overshooting actually begins at
>>>>> the location determined by f0 back into the convection zone rather than at
>>>>> the edge where the diffusion coeff is ill-defined. The overshooting
>>>>> actually begins at the location determined by f0 back into the convection
>>>>> zone rather than at the edge where the diffusion coeff is ill-defined. So,
>>>>> for example, if you want overshooting of 0.2 scale heights beyond the
>>>>> normal edge, you might want to back up 0.05 scale heights to get the
>>>>> diffusion coeff from near the edge and then go out by 0.25 scale heights
>>>>> from there to reach 0.2 Hp beyond the old boundary. In the inlist this
>>>>> would mean setting the “f0” to 0.05 and the “f” to 0.25.
>>>>>
>>>>> ----
>>>>> A common practice, though not required, is to take a step back such
>>>>> that fo = 0.5*f, which helps establish a diffusion coefficient well inside
>>>>> the convection zone. In the current form, your diffusion coefficient for
>>>>> (1) is set to 0.006*Hp, as you are taking a 0.004*Hp step backward into the
>>>>> convection zone, and then overshooting 0.01*Hp from this position.
>>>>> In (2), you are taking a 0.004*Hp step backward and then overshooting
>>>>> 0.004*Hp forward, which is technically not overshooting above the
>>>>> convective boundary and is instead enhancing the diffusion coefficient near
>>>>> the edge of the convection zone.
>>>>> ----
>>>>> Here is a corrected overshoot scheme, which is what I believe you
>>>>> might be after? :
>>>>>
>>>>> &controls
>>>>> ! less overshooting above Z burning shell (e.g. C, O, Si), overshoot
>>>>> distance = 0.004*Hp
>>>>> overshoot_scheme(1) = 'exponential' ! one of exponential or
>>>>> step
>>>>> overshoot_zone_type(1) = 'burn_z' ! one of burn_H,
>>>>> burn_He, burn_Z, nonburn, any
>>>>> overshoot_zone_loc(1) = 'shell' ! one of core, shell,
>>>>> any
>>>>> overshoot_bdy_loc(1) = 'any' ! one of bottom, top, any
>>>>> overshoot_f(1) = 0.008 ! was 0.004
>>>>> overshoot_f0(1) = 0.004
>>>>>
>>>>> ! overshooting for all shells, superseded by overshoot_scheme(1) for
>>>>> burning Z shells, overshoot distance = 0.01*Hp
>>>>> overshoot_scheme(2) = 'exponential' ! one of exponential or
>>>>> step
>>>>> overshoot_zone_type(2) = 'any' ! one of burn_H, burn_He,
>>>>> burn_Z, nonburn, any
>>>>> overshoot_zone_loc(2) = 'any' ! one of core, shell, any
>>>>> overshoot_bdy_loc(2) = 'any' ! one of bottom, top, any
>>>>> overshoot_f(2) = 0.015 ! was 0.01
>>>>> overshoot_f0(2) = 0.005 ! was 0.004 and is fine, but changed by
>>>>> EbF for consistency with overshoot_scheme(1)
>>>>>
>>>>> ----
>>>>>
>>>>> >"Another question is that if I want to test with no overshooting for
>>>>> shell burning, can I set f = 0, or I need to give a very small value
>>>>> for f. Because f and f0 shoould >0."
>>>>>
>>>>> You can do this by setting: overshoot_scheme(1) = 'none' , instead of
>>>>> 'exponential' in the &controls I shared above.
>>>>>
>>>>> -EbF
>>>>>
>>>>> On Thu, Apr 20, 2023 at 2:35 AM 辛文宇 <xinwenyu16 at mails.ucas.ac.cn>
>>>>> wrote:
>>>>>
>>>>>> Hi Everyone,
>>>>>>
>>>>>>
>>>>>> I am going to change the overshooting of C-shell, O-shell and
>>>>>> Si-shell burning to a smaller value.
>>>>>>
>>>>>> Can I set the overshooting as follows:
>>>>>>
>>>>>>
>>>>>> overshoot_scheme(1) = 'exponential' ! one of exponential or
>>>>>> step
>>>>>> overshoot_zone_type(1) = 'any' ! one of burn_H,
>>>>>> burn_He, burn_Z, nonburn, any
>>>>>> overshoot_zone_loc(1) = 'any' ! one of core, shell, any
>>>>>> overshoot_bdy_loc(1) = 'any' ! one of bottom, top, any
>>>>>> overshoot_f(1) = 0.010
>>>>>> overshoot_f0(1) = 0.004
>>>>>>
>>>>>>
>>>>>> overshoot_scheme(2) = 'exponential' ! one of exponential or
>>>>>> step
>>>>>> overshoot_zone_type(2) = 'burn_z' ! one of burn_H,
>>>>>> burn_He, burn_Z, nonburn, any
>>>>>> overshoot_zone_loc(2) = 'shell' ! one of core, shell,
>>>>>> any
>>>>>> overshoot_bdy_loc(2) = 'any' ! one of bottom, top, any
>>>>>> overshoot_f(2) = 0.004
>>>>>> overshoot_f0(2) = 0.004
>>>>>>
>>>>>>
>>>>>> In first part, the overshooting of all of the cores and shells are
>>>>>> set to 0.01.
>>>>>>
>>>>>> Then second part change the 'f' of only the shells to 0.004, right?
>>>>>>
>>>>>>
>>>>>> Another question is that if I want to test with no overshooting for
>>>>>> shell burning,
>>>>>>
>>>>>> can I set f = 0, or I need to give a very small value for f. Because
>>>>>> f and f0 shoould >0.
>>>>>>
>>>>>>
>>>>>> Thanks,
>>>>>>
>>>>>> Wenyu
>>>>>>
>>>>>>
>>>>>> _______________________________________________
>>>>>> mesa-users at lists.mesastar.org
>>>>>>
>>>>>> https://urldefense.com/v3/__https://lists.mesastar.org/mailman/listinfo/mesa-users__;!!IKRxdwAv5BmarQ!bmHiP3-PihLTp-9DGHcp2mi-40kCuk2ZQdOueGJk74bQ3F8_AFGhOZIG-8uDaEdoJIo_zWhlZRoAIpQ8oJbcXb0RvQ$
>>>>>>
>>>>>> _______________________________________________
>>>>> mesa-users at lists.mesastar.org
>>>>> https://lists.mesastar.org/mailman/listinfo/mesa-users
>>>>> <https://urldefense.com/v3/__https://lists.mesastar.org/mailman/listinfo/mesa-users__;!!IKRxdwAv5BmarQ!ZfhBDF7qPDvk7dUmjWJJRNRD02WtriLQx_y3leencWpIpjIw9l1OQwA3eeNBiU49Ty42qWa8aQAdSCj0dSOD4mjs$>
>>>>>
>>>>> _______________________________________________
>>> mesa-users at lists.mesastar.org
>>>
>>> https://urldefense.com/v3/__https://lists.mesastar.org/mailman/listinfo/mesa-users__;!!IKRxdwAv5BmarQ!ZfhBDF7qPDvk7dUmjWJJRNRD02WtriLQx_y3leencWpIpjIw9l1OQwA3eeNBiU49Ty42qWa8aQAdSCj0dSOD4mjs$
>>>
>>>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.mesastar.org/pipermail/mesa-users/attachments/20230426/cab54a60/attachment.htm>
More information about the Mesa-users
mailing list