[Mesa-users] Type I X-ray bursts using ns_he test case

[Wolf, William M]

Hi Triantafyllos,

Starting from the test case is a good first step. I imagine the problem is that the flash needs to expel matter. You might try turning on super eddington winds<https://docs.mesastar.org/en/24.08.1/reference/controls.html#super-eddington-scaling-factor> or some other mass loss mechanism. Without it, the envelope may become very extended and radiation dominated, which tends to be numerically unstable.

This paper<https://ui.adsabs.harvard.edu/abs/2023ApJ...954...54G/abstract> used MESA and super eddington winds to study type I X-ray bursts, so it might have some other helpful tips for evolving through a flash.

Finally, I’ve found that in radiation dominated envelopes like what you’re seeing here, the convergence_ignore_equL_residuals<https://docs.mesastar.org/en/24.08.1/reference/controls.html#convergence-ignore-equl-residuals> control to be essential. It basically tells the solver to stop caring about how large the residuals are in the luminosity equation. Doesn’t sound great, but it’s what you have to do to get through some phases of evolution.

Good luck!

Bill Wolf



From: Mesa-users <mesa-users-bounces at lists.mesastar.org> on behalf of Triantafyllos Kormpakis via Mesa-users <mesa-users at lists.mesastar.org>
Date: Friday, September 6, 2024 at 6:09 AM
To: mesa-users at lists.mesastar.org <mesa-users at lists.mesastar.org>
Subject: [Mesa-users] Type I X-ray bursts using ns_he test case
Hi everyone,

I'm quite a new user of MESA, and I'm trying to create a minimal working case of Helium type I X-ray bursts using MESA. I'm using mesa-r24.03.1.

I haven't created my own inlists, I'm just using the ns_he test case, and trying to modify it in order to follow the evolution of multiple bursts in time. The only modification I have done to the test case is to remove the lines:

required_termination_code_string = 'power_nuc_burn_upper_limit'
power_nuc_burn_upper_limit = 1d6

in the 'inlist_to_flash' inlist file. This in principle allows the model to evolve. The problem is that after some time, the hydro solver fails (using more iterations than the limit), and even if I increase the allowed iterations, the timestep starts getting smaller and smaller, until the hydro solver fails again. I have tried using the debugging options, saw that the 'equL' seems to be the limiting timestep factor, but when comparing the analytical and numerical derivatives following the MESA debugging principles, I find no apparent problem.

What I want to ask is, how should I modify the ns_he test case in order to have a minimal working simulation setup, on which I can afterwards build on? Before experimenting with different mass accretion, base luminosity etc. I want to achieve a stable setup.

Before anyone suggests, I have also looked into the inlists provided in the instrument paper of Paxton et al 2015, and other inlists from the MESA summer school for X-ray bursts, as well as Meisel 2018. I can make these inlists work, after looking into the changelog and fixing the backwards incompatibility issues, but even in these cases the simulations don't run as intended and I haven't figured out why.

Any help/suggestion would be appreciated. Thank you.

Triantafyllos Kormpakis
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.mesastar.org/pipermail/mesa-users/attachments/20240906/57d95d5f/attachment.htm>