Cesam2k20: A code for a new generation of stellar evolution models. I. Description of the code

TL;DR

Cesam2k20 advances the benchmark stellar evolution code CESAM by integrating state‑of‑the‑art transport for chemical elements and angular momentum within a collocation/B‑Spline numerical framework. It offers extensive options for opacities, equations of state, convection, overshoot, diffusion, and atmosphere treatments, together with a Python ecosystem for running, plotting, and grid generation. The paper demonstrates capabilities through solar models, showing improvements in acoustic structure and Li depletion relative to standard implementations, and discusses future directions toward MHD and atmosphere reconstruction. Overall, Cesam2k20 represents a comprehensive, extensible platform enabling precise, modular 1D stellar modelling with a close link to observational constraints and future PLATO needs.

Abstract

We present Cesam2k20, the latest version of the hydrostatic stellar evolution code CESAM originally developed by P. Morel and collaborators. Over the last three decades, it has undergone many improvements and has been extensively tested against other stellar evolution codes before being selected to compute the first-generation grid of stellar models for the PLATO mission. Among all the developments made thus far, Cesam2k20 now implements state-of-the-art models for the transport of chemical elements and angular momentum. It was recently made publicly available with an ecosystem of other codes interfaced with it: 1D and 2D oscillation codes ADIPLS and ACOR, optimisation program OSM, and Python utility package pycesam. This paper recalls the numerical peculiarities of Cesam2k20, namely, the use of a collocation method where the structure variables are decomposed as piecewise polynomials projected on a B-spline basis. Here, we review the options available for modelling the different physical processes. In particular, we illustrate the improvements made in the transport of chemical elements and angular momentum with a series of standard and non-standard solar models.

Figures (3)

• Figure 1: Schematic representation of steps followed by Cesam2k20 in computing a time step.
• Figure 2: Left panel: Kippenhahn diagram of two models of $1.3M_\odot$ computed with the method originally implemented in CESAM Morel1997, and with the method prescribed by Gabriel2014 and now implemented in Cesam2k20. The radiative zone corresponds to dark grey region for OrigM, and hatched region for G14M. Convective zone is the region in light grey for OrigM, and non-hatched region for G14M. Red regions are spurious convective zone. Vertical dashed line at age $2923$ Myrs marks the position in time of the model represented on right. Right panel: Profile of $\mathrm{d}\nabla \equiv \nabla_{\rm rad} - \nabla_{\rm ad}$ as a function of the mass coordinate, $m/M_\odot$, at each iteration of a given time step (at age $2923$ Myrs). Profiles for G14M are drawn in solid lines, and profiles for OrigM are in dashed lines. Up (G14M) or down (OrigM) triangles mark the location of each layer, at each iteration. Grey up (G14M) or down (OrigM) arrows mark the location of the convective boundary at the ten previous time steps. Lightest arrow if for the tenth previous time step and darkest arrow for the immediate previous one.
• Figure 3: Left panel: Relative differences between inverted and modelled sound speed as a function of the radial coordinate, for SSM1 to NSSM3. The inverted sound speed is taken from Basu2000. Right panel: Time evolution of the surface ${}^7{\rm Li}$ abundance for SSM1 to NSSM3. Horizontal black line is the present-day solar surface ${}^7{\rm Li}$ abundance as measured by Asplund2021, and the shaded area represents the associated uncertainty.