[Mesa-users] How to change the reaction rates in the simplex_solar_calibration model

Sat May 10 14:00:12 UTC 2025

hi aman,

1) changing S(0) is equivalent, at first order, to directly scaling the reaction rate. section 3.2.1 of "nuclear physics of stars" by christian iladis can be useful here. a change in S(0) is easily accomplished in mesa by multiplying a rate by a constant factor, https://docs.mesastar.org/en/latest/reference/star_job.html#num-special-rate-factors, and was eb's first suggested option. if the end goal is to explore how solar models may change with reaction rates, then this is the easiest route.

2) the analytical expression is a fitting formula, a sum of a power function times an exponential times a polynomial. it is not derivable from first principles. for example, one could imagine a fitting formula that choses to use a rational function instead of a polynomial. it is also not possible to determine how the fitting coefficients may change without having the original data to fit, but again changing S(0) is tantamount to scaling a reaction rate up or down. alternatively, don't use a fitting function at all by interpolating a suitably dense table of reaction rates; https://docs.mesastar.org/en/latest/net/nets.html#change-rates-for-existing-reactions , which was eb's second suggested option.

3) please use a url to a published article (e.g., https://doi.org/10.1016/S0375-9474(99)00030-5) rather than stuffing ~1100 mesa-user's inboxes with a pdf file ;)

fxt

> On May 8, 2025, at 10:11 PM, aman awasthi via Mesa-users <mesa-users at lists.mesastar.org> wrote:
> 
> HI Ebraheem, 
> 
> I took so much time to respond as I was  trying to implement the option 3 suggested by you. I further have questions based on my efforts till now.  
> 
> I am trying to understand and derive the analytical formula for the reaction rates as used in the nacresubdirectory in MESA. I have outlined the formula in the attached PDF file.
> Could you kindly suggest some references or resources where I can find the exact analytical form used in MESA (as in the NACRE compilation)? Additionally, I would like to understand how the coefficients in the analytical formula would change if I modify the S(0) value for a reaction.
> I have tried solving the integration involved, but I was not able to reproduce the analytical expression. I would be grateful for any suggestions or insights you might have on how best to derive or access this formula, and how the coefficients would be affected by changes in S(0).
> I am also attaching a reference paper which is cited by MESA while calculating the reaction rates using the  analytical expression .
> Thank you very much for your time and guidance. Any suggestions you can offer would be greatly appreciated.
> 
> Best Regards
> Aman Awasthi 
> 
> On Thu, Apr 10, 2025 at 11:07 AM Farag, Ebraheem <ebraheem.farag at yale.edu> wrote:
> Hello Aman,
> 
> I see you made two separate emails/threads for this issue on the mailing list. Perhaps that was in error, but let's try to keep it to one thread per issue 🙂.
> 
> >1. How to change the reaction rates of the above reactions in the simplex_solar_calibration model?
> >2.Are there other methods available within MESA to change the reaction rates aside from modifying the rate_tables directory?
> 
> There is more than one option, depending on how deep you are willing to go.
> 
> The first option is to just apply a factor to the current rates to achieve the desired effect. Seehttps://docs.mesastar.org/en/latest/reference/star_job.html#filename-of-special-rate. I don't know if this will work for the pp reaction since it is a weak reaction though.
> 
> The second option is to calculate the rates yourself, tabulate them, and read them into MESA.
> 
> The third, and likely most robust option is below.
> --- beware I'm about to suggest touching MESA source code below---
> The more involved although not terribly difficult option is to directly change the s_factor when MESA calculates the rate. Navigate to $MESA_DIR/rates/private/ratelib.f90, and check out this example branch I made where I directly modified the astrophysical s-factor for n14pg (albeit through a factor instead of replacing the default). You can ignore most of the files and focus on ratelib.f90.https://github.com/MESAHub/mesa/compare/main...EbF/add_rate_uncertainty_option
> 
> If you change the s_factor in the ratelib, you will need to recompile rates and recompute the rate files. 
> 
> This involves cleaning your rates_cache directory (The global directory is contained in $MESA_DIR/data/rates_data).
> 
> The easiest way forward is to add a file in $MESA_DIR called "skip_test", change what needs changed in $MESA_DIR/ratelib, then do a fresh install "./clean" "./install" of MESA. This will allow you to make your desired changes without worrying about satisfying the unit tests MESA conducts during installation.
> 
> If you want MESA to perform testing during the installation, you'll have to do the following instead: 
> 
> I would recommend first cleaning the rates_cache, then: 
> Navigate into $MESA_DIR/rates and change what you see fit. then:
> "cd $MESA_DIR/rates"
> "./clean"
> "./mk"
> "./build_and_test"
> 
> You'll probably receive an error stating that the "tmp.txt" file does not match "test_output", so do:
> "cp $MESA_DIR/rates/test/tmp.txt $MESA_DIR/rates/test/test_output", to replace the old test data with your new test data. 
> 
> Now, in $MESA_DIR/rates, you can again do:
> "./clean"
> "./mk"
> "./build_and_test"
> 
> if this is successful, follow it with:
> "./export"
> 
> .or. In my opinion, do a fresh install to make sure everything checks out:
> In $MESA_DIR, do:
> "./clean"
> "./install"
> 
> if you run into testing issues in "$MESA_DIR/net", you will likely have to repeat the procedure above for updating the testing files.
> 
> 
> >3. Is it possible to update the reaction data directly on the REACLIB website? I am unsure about how to proceed with the verification process and editing the libraries on the site.
> 
> I'm not sure.
> 
> ---
> 
> On solar models in MESA:
> If your plans are to use the simplex module for solar calibration, keep in mind it is only a starting point for computing standard solar models. The standard is to only change the initial Y, Z, and alpha_mlt, and then to run the calibration until L ~ Lsun, R~ Rsun, and Z_surf/X_surf ~ the observed Z_surf/X_surf at age ~ solar age (say ~4.568d9 yr). I have  an example of using the simplex calibration for solar models in my own work (seehttps://ui.adsabs.harvard.edu/abs/2024ApJ...968...56F/abstract, where the MESA files can be found in the zenodo here https://zenodo.org/records/10798600). Keep in mind, i used a very large resolution and strict timestepping in my models, which could be considered overkill. I also make other inlist choices that you should consider carefully before adopting. Another thing to note is that evolving from the pre-ms can make the calibration noisy (a mystery yet to be solved). If you instead evolve from the ZAMS with smart initial conditions, you should be able to create a converged solar model in under 5-10 iterations (as opposed to 100s in my work). I would recommend checking out the methods in (https://doi.org/10.1093/mnras/199.3.735) for good insight into how to structure a solar model for faster convergence.  
> 
> While the simplex_solar_calibration works, it is not yet a plug and play test_suite for computing standard solar models. 
> 
> 
> -EbFFrom: Mesa-users <mesa-users-bounces at lists.mesastar.org> on behalf of aman awasthi via Mesa-users <mesa-users at lists.mesastar.org>
> Sent: Wednesday, April 9, 2025 5:40 PM
> To: mesa-users <mesa-users at lists.mesastar.org>
> Subject: [Mesa-users] How to change the reaction rates in the simplex_solar_calibration model
>  Hi everyone, 
> 
> I am currently working on modifying the reaction rates of the following reactions in MESA's simplex_solar_calibration module
> 1. For p(p, e+𝜈e)d, taking S factor as S11 (0) = (4.03 ± 0.006). 10^{−25} MeVb
> 2. For 7Be(p, 𝛾)8B, taking S factor as S17 (0) = (2.13 ± 0.01). 10^{−5} MeVb 
> 3. For 14N(p, 𝛾)15O, taking S factor as S114 (0) = 1.59. 10^{−5} MeVb 
> 4.For 3He(4He, 𝛾)7Be, taking S factor as  S34 (0) = (5.72 ± 0.12). 10^{−4} MeVb
> 
> So I tried to look for it, I found a good conversation for it on mesa-user email arxiv (https://lists.mesastar.org/pipermail/mesa-users/2016-December/007313.html) that discusses how to modify reaction rates. 
> However, I’m unclear on how to generate the new data files for these updated reaction rates, which I would need to place in my rate_tables local directory. 
> For this I explored the REACLIB database (website link - https://reaclib.jinaweb.org/index.php). It mentions that we can change and  edit   libraries there. I created an account in hopes of gaining access to modify the reaction rate data, but I am still awaiting the verification email to complete the process.
> 
> I have a few questions regarding this process:
> 1. How to change the reaction rates of the above reactions in the simplex_solar_calibration model?
> 2.Are there other methods available within MESA to change the reaction rates aside from modifying the rate_tables directory?
> 3. Is it possible to update the reaction data directly on the REACLIB website? I am unsure about how to proceed with the verification process and editing the libraries on the site.
> 
> I would greatly appreciate your guidance on these points. Thank you for taking the time to consider my questions.
> 
> Regards
> Aman Awasthi 
> Phd student
> IIT Bombay , India<NACRE_reaction_rate_calculation_formula-1.pdf><A_compilation_of_charged_particle_induced_thermonuclear_ reaction_rates_Angulo_et_al.pdf>_______________________________________________
> mesa-users at lists.mesastar.org
> https://lists.mesastar.org/mailman/listinfo/mesa-users