2024MNRAS.534..621D

We have numerically demonstrated that simulated cool star coronae naturally form condensations. If the star rotates slowly, with a co-rotation radius greater than the Alfvén radius (i.e. $R_{\mathrm{K}} \gt R_{\mathrm{A}}$), these condensations will form below the co-rotation radius $R_{\mathrm{K}}$ and simply fall back to the stellar surface as coronal rain. If, however, the star is more rapidly rotating, ($R_{\mathrm{K}} \lt R_{\mathrm{A}}$), not only rain will form but also 'slingshot prominences'. In this case, condensations collect into a large mass reservoir around the co-rotation radius, from which periodic centrifugal ejections occur. In this case, some 51 per cent of the coronal mass is cold gas, either in rain or prominences. We find that 21 per cent of the mass lost by our simulated fast rotating star is cold gas. Studies of stellar mass-loss from the hot wind do not consider this component of the wind and therefore systematically underestimate mass-loss rates of these stars. Centrifugal ejections happen periodically, between every 7.5-17.5 h with masses clustering around $10^{16}$ g, These results agree well with observational statistics. Contrasting the fast and slow rotating magnetospheres, we find that there are two distinct types of solutions, high-lying and low-lying loops. Low-lying loops only produce coronal rain whereas high-lying loops produce both rain and slingshots.