2006MNRAS.371.1381G

We study the angular momentum evolution of binaries containing two white dwarfs (WDs) which merge and become cool helium-rich supergiants. Our object is to compare predicted rotation velocities with observations of highly evolved stars believed to have formed from such a merger, which include the RCrB and extreme He stars.

The principal case study involves a short-period binary containing a 0.6-M_☉carbon-oxygen (CO) WD, and a 0.3-M_☉ He WD. The initial condition for the angular momentum distribution is defined by the orbital configuration where the secondary fills its Roche lobe.

Since mass transfer from the secondary is unstable, the WD breaks up on a dynamical time-scale. After accreting some mass, the primary is assumed to ignite helium and evolve to become a yellow supergiant with a He-rich surface. We assume conservation of angular momentum to compute the initial angular momentum distribution in a collisionless disc and subsequently in the giant envelope. At the end of shell-helium burning, the giant contracts to form a WD. We derive the surface rotation velocity during this contraction.

The calculation is repeated for a range of initial mass ratios, and also for the case of mergers between two helium (He) WDs; the latter will contract to the helium main sequence rather than the WD sequence.

Assuming complete conservation of angular momentum, we predict acceptable angular rotation rates for cool giants and during the initial subsequent contraction. However, such stars will only survive spin-up to reach the WD sequence (CO+He merger) if the initial mass ratio is close to unity. He+He merger products must lose angular momentum in order to reach the helium main sequence.

Minimum observed rotation velocities in extreme helium stars are lower than our predictions by at least one-half, indicating that CO+He mergers must lose at least one-half of their angular momentum, possibly through a wind during shell-helium burning, but more likely from the disc, following secondary disruption.