Presentation #116.11 in the session Stellar/Compact Objects.
Over the past few years, the NICER telescope has provided us with the best observations of Type I X-ray bursts to date. In particular, NICER’s soft X-ray response now allows us to track photospheric radius expansion (PRE) bursts throughout their whole evolution, even as the neutron star’s apparent area increases by a factor of up to 400 (Keek et al. 2018). This motivates the development of improved theoretical models describing the dynamics of this expansion, so that we may learn about the complex physics of nuclear burning and fluid motions happening on the neutron star surface, the ejection of heavy metals, and the constraints on neutron star masses and radii from burst observations. Motivated by the recent NICER observation of a Type I X-ray burst with a distinct “pause” feature during its rise (Bult et al. 2019), we show that bursts which ignite in a helium layer underneath a hydrogen-rich shell naturally give rise such pauses, as long as enough energy is produced to eject the outer layers of the envelope by super-Eddington winds. Moreover, we show that another distinct feature of such bursts is a slow rise in the luminosity following the pause. In a first set of simulations, we use the MESA stellar evolution code to model the accretion, nuclear burning and convective motions prior to and throughout the ignition of the burst, followed by the hydrodynamic wind driving the PRE. We show that both the duration of the pause and the subsequent rise of the lightcurve are determined by the gradient of hydrogen in the envelope left behind by convection. However, it is unclear how this convection should be treated in a one-dimensional simulation. In particular, we show that our results are sensitive to the definition of convective boundaries. In contrast, multidimensional simulations allow us to simulate the fluid motions involved in convection directly, rather than rely on approximate prescriptions. We show preliminary results of hydrodynamical simulations of convection in this interesting setting, using the low-mach number MAESTROeX code.