XSEDE Scholars Program

Scholar Glenna Dunn

Glenna Dunn (Vanderbilt University; graduate student; physics & astronomy)

Project Title: Forming direct collapse black holes with realistic Lyman-Werner radiation fields in cosmological hydrodynamic simulations

Research Advisor: Kelly Holley-Bockelmann, Department of Astronomy, Vanderbilt Univeristy


HPC Resources: Stampede


While we know that quasars must have formed as massive seeds, the mechanism by which this occurs is unresolved. One popular model that can explain how these objects form is known as direct collapse black hole formation. This model proposes that massive black holes form from the rapid collapse of a pristine gas cloud just as their host galaxies begin to take shape. A significant challenge that this model faces is how such a large quantity of gas, usually ten to one hundred thousand times the mass of our Sun, can collapse without breaking apart into smaller clouds that will form stars. The tipping point between forming a black hole or stars relies on a delicate interplay between gravity, the chemical properties of the gas, the local radiation field, and internal gas physics. These processes are best understood through detailed simulations.

We propose to implement an updated black hole physics module in the massively parallel cosmological hydrodynamic code Gasoline to simulate the formation of direct collapse black holes in the early Universe. We are particularly interested in the role of ambient ultraviolet radiation in a specific band, known as Lyman-Werner (LW) radiation, in creating an environment conducive to direct collapse black hole formation. Previous work, such as that of Shang et al (2009), has shown that if the LW radiation exposure of a primordial gas cloud exceeds a critical threshold, the cloud may meet the physical requirements to collapse into a massive black hole. We intend to use cosmological simulations to study the conditions under which primordial gas clouds experience high levels of LW radiation, and how this radiation affects massive black hole formation.

The first phase of this project will involve the development of a modified black hole formation module that will reflect the physical interplay between the amount of Lyman-Werner radiation felt by a gas cloud and the probability that the cloud will form a black hole. We plan to test this module by implementing it in a low-resolution simulation. The second phase of this project will involve the implementation of the black hole formation module in high-resolution Gasoline simulations. We expect to launch three simulations that explore the parameter space of the critical LW threshold to study how this value changes black hole seed formation, and plan to submit the production- level simulations to the XSEDE/TACC supercomputer Stampede. We intend to use this research to answer some of the open questions regarding the type of environment that allows massive black holes to form. The most highly anticipated results of this work will explore the proximity between LW sources and potential black hole formation sites required for direct collapse to occur. Additional results will offer constraints on massive black hole occupation fraction in the quasar epoch, and implications for reionization, high-redshift X-ray background radiation, and gravitational waves.