Computing time at IT4I

Our postdocs were successful in obtaining supercomputer time on the Czech supercomputer infrastructure IT4Innovations. Diego Calderón won 26 000 node hours (approx. 900 000 CPU hours) on Barbora cluster to perform radiation hydrodynamics simulations of transients. Damien Gagnier won 36 000 node hours (approx. 4 600 000 CPU hours) on the brand new Karolina cluster to develop 3D hydrodynamic simulations of common envelope.

Supernovae in colliding-wind binaries: observational signatures in the first year

When a core-collapse supernova explodes in a binary star system, the ejecta might encounter an overdense shell, where the stellar winds of the two stars previously collided. It is worth looking more closely at this colliding-wind shell, because it’s structure can be relatively well predicted based on the properties of the binary. Together with Diego and Petr, we investigated what observable effect for light curves and flash ionization arises from supernova hitting the colliding wind shell. In the process, we developed a semi-analytic model of CSM distribution in colliding-wind binaries calibrated by Diego’s 3D AMR simulations. We combined this model with a thin shell model for supernova shock interaction that is angularly dependent. The conclusion is that observable signatures still require relatively high wind mass-loss rates, which might not be completely realistic, and nearly equal-mass companions at large separations, which might be rare. Preprint available here.

Observing trip to Valašské Meziříčí

We are a primarily theoretical group, but close connection to observations is essential to keep us grounded in the reality. We spent three nights in Valašské Meziříčí with the aim of using 20cm telescope with CCD camera at the local Observatory. Although only one night was clear, we learned how to use the telescope, got familiar with basic deep sky objects, and tried processing time-series photometry on archival data. Media coverage of our visit.

Two Master theses defended

Dominika Hubová graduated with thesis “Angular momentum loss from binary systems due to stellar winds”, where she combined radiative acceleration with Roche geometry in the context of winds from massive stellar binaries.

Jakub Cehula presented thesis “Stability of stars undergoing rapid mass loss”, where he worked out ideas on how to model mass transfer in binary stars.

Both theses were awarded “excellent” grades. Congratulations!

Moving-mesh radiation-hydrodynamic simulations of wind-reprocessed transients

New paper spearheaded by Diego Calderón, who developed a new radiation module for moving-mesh code JET by Paul Duffell. The moving mesh is constrained to move only radially, essentially working as a set of 1D Lagrangian wedges coupled by lateral fluxes. Radiation is implemented in flux-limited diffusion approximation and with implicit timestep update. All of these features make this code a perfect tool for studying various explosions and transients in astrophysics, especially the ones lacking spherical symmetry and powered by radiative shocks.

As a first application, we looked at wind-reprocessed transients (think tidal disruption events or AT2018cow-like supernovae), where we verified the analytics of Piro & Lu (2020) and observed how the result changes for anisotropic central radiation sources.

Preprint available at https://arxiv.org/abs/2105.08735.

The luminous red nova AT 2018bwo in NGC 45 and its binary yellow supergiant progenitor

New paper spearheaded by Nadia Blagorodnova and Jakub Klencki on Luminous red nova in NGC 45. For the first time, the progenitor of LRN was analyzed using binary star evolution tracks instead of single star tracks finding that the location exactly matches thermal-timescale (case B) mass transfer in a massive binary. The models suggest that quite a bit of mass (few Msun) was lost from the system in 100s of years before the merger. We take the structure of the final model of the primary and based on standard common envelope formalism predict how much mass should be ejected. We compare this prediction with inference from the transient, modeled as either scaled-down Type II-P or shock-interaction powered. We need transient models calibrated specifically for LRNe! We’re also somewhat puzzled by not seeing a lot of dust in the progenitor – something for future work.
https://arxiv.org/abs/2102.05662

ERC-funded position in computational astrophysics

We invite applications for a postdoctoral position in the area of computational astrophysics in our research group. The successful candidate will lead development of new computational abilities to understand multi-dimensional evolution of catastrophic interactions of binary stars (mergers, common envelope, compact objects, supernovae). The work will be performed within the project “Cat-In-hAT” funded by the ERC Starting Grant.

Applicants must have a PhD in astronomy, computational physics, or a related field by the start date of the appointment. Candidates with experience with simulations of stellar convection, turbulence, and magnetohydrodynamics are particularly encouraged to apply. The position includes competitive salary and research budget on the level of international fellowships.

Continue reading “ERC-funded position in computational astrophysics”