Context. The Gaia M-dwarf gap is a recently discovered feature in the colour-magnitude diagram that shows a deficiency of low-mass and low-metallicity stars at the lower end of the main sequence.Aims.
We aim at performing theoretical stellar modelling at low metallicities using a fine mass step and a fine time step, looking specifically for the transition of models from partially to fully convective since the convective kissing instability that occurs at this transition is believed to be the cause of the gap.Methods. Stellar evolution models with metallicities of Z=0.01, Z=0.001, and Z=0.0001 are performed using MESA, with a mass step of 0.00025 M-circle dot and a time step of 50 000 years.Results.
The small time step produced models that experience loops in their evolutionary tracks in the Hertzsprung-Russell (HR) diagram. The fluctuations in effective temperature and luminosity correspond to repeated events in which the bottom of the convective envelope merges with the top of the convective core, transporting He-3 from the core to the surface.
In addition to the episodes of switching from partially to fully convective, several near-merger events that produced low amplitude fluctuations were also found. Low-metallicity models undergo the convective kissing instability for longer portions of their lifetime and with higher fluctuation amplitudes than models with higher metallicities.
The small mass step used in the models revealed a discontinuity in the luminosity-mass relation at all three metallicities.Conclusions. The repeated merging of the convective core and envelope, along with several near-merger events, removes an abundance of He-3 from the core and temporarily reduces nuclear burning.
This results in fluctuations in the model's luminosity and effective temperature, causing loops in the evolutionary track in the HR diagram and leading to the deficiency of stars at the M-dwarf gap, as well as a discontinuity in the luminosity-mass relation.