radiation hydrodynamics of super star cluster...

1
Radiation Hydrodynamics of Super Star Cluster Nuclei Benny Tsz-Ho Tsang & Milos Milosavljevic The University of Texas at Austin From Stars to Massive Stars, Gainesville, Florida, 6 th – 9 th April 2016 Super Star Clusters Radiation Transport Matters What is the Role of Radiation? Simulating their Births Results Super star clusters (SSCs) are the most extreme star-forming sites in the Universe: massive stars at the cluster nuclei form from gas surface densities of 10 3 – 10 4 M pc 2 , with star formation rates above 10 – 100 M yr 1 . The intense radiation field from the dense population of young, massive stars is expected to be dynamically important in altering the star-forming gas. The resultant massive star densities are so high that stellar merging may be common, providing a potential pathway for massive black hole formation. Large distances to these extreme objects and dust obscuration have been limiting our observational access; theoretical models must realize the multidimensional complexity of dusty, radiation-dominated gas on scales < 0.1 pc inside the nucleus actively forming massive stars. Studies have shown that radiation pressure could dominate the disruption of parent molecular clouds in forming massive clusters (Krumholz & Matzner 2009, Murray et al. 2010, Kim et al. 2016). However, these predictions were based on one-dimensional models. Does stellar radiation pressure from young stars inhibit or terminate star formation in actual SSC? How long does star formation last in SSC? Does stellar radiation drive intense outflows? We implemented a closure-free, Monte Carlo radiation transfer scheme coupled with the hydrodynamic code FLASH. We confirmed that the accuracy of the radiation transport method could have a huge impact on the global behaviors of gas dynamics when radiation pressure is dominant and steep transitions in optical thickness is present. Initially static layer of dusty gas subject to a constant upward radiation flux and a constant downward gravity. Same exact setup has been tested with three other radiation transport methods, listed in increasing accuracy: flux-limited diffusion (FLD), M1 closure, and the variable Eddington tensor approximation (VET). We aim to identify the physical conditions (e.g., cloud mass and surface density) in which radiation pressure’s effects on turbulent molecular clouds is dominant. With a particle-based radiation transport scheme, our simulations will automatically deliver the full observable spectral energy distributions (SEDs) of SSCs. They will motivate and inform observations of SSCs in formation. Left: Gas density snapshots in the setup of radiation pressure-driven atmosphere. The flow morphology observed in our Monte Carlo scheme is consistent with other radiation transport methods. Right: Time evolution of the mass-weighted mean vertical velocity (upper) and velocity dispersion (lower). The late-time net acceleration and velocity dispersions are in agreement only with the VET method. It is therefore essential to adopt proper radiation transport in simulations with strong radiation pressure and steep transitions in optical thickness. Left: Time evolution of the fraction of gas mass with positive radial velocity (outflow) within 1 pc radius of the three most massive clusters. We simulate the formation of super star clusters from a 10 8 M giant molecular cloud, focusing on the feedback by stellar radiation pressure. Physics included are: Initial density/velocity distributions set by driven turbulence; Gas self-gravity and star-gas gravity; Star particles created as formed clusters and radiation sources; Frequency-dependent transfer of non-ionizing UV and dust- reprocessed IR radiation; Full coupling of radiation pressure with gas dynamics. The stellar radiation field can be super-Eddington locally around the massive clusters. It does not seem to effectively limit the infall of gas or terminate star formation due to the turbulent nature of the clouds and the strong infall of gas towards the growing clusters. Left: Mass-weighted gas density projection of the turbulent initial conditions. The initial surface densities before gravity is turned on are > 1000 M pc 2 , typical for starburst environments. Right: The same projection one free-fall time after gravity has been turned on, the black points mark the projected locations of the star particles representing the luminous clusters formed. Also shown is the zoomed-in view of the most massive cluster formed. Davis S. W., Jiang Y.-F., Stone J. M., Murray N., 2014, ApJ, 796, 107 (VET). Kim, J. G., Kim, W. T., Ostriker E. C., 2016, ApJ, 819, 137. Krumholz, M. R., & Matzner, C. D., 2009, ApJ, 703, 1352. Krumholz, M. R., & Thompson, T. A., 2012, ApJ, 760, 155 (FLD). Krumholz, M. R., & Thompson, T. A., 2013, MNRAS, 434, 2329 (FLD). Murray, N., Quataert, E., & Thompson, T. A. 2010, ApJ, 709, 191. Rosdahl J., Teyssier R., 2015, MNRAS, 449, 4380 (M1 closure). Tsang, B. T.-H., & Milosavljevic, M. 2015. MNRAS, 453, 1108. References: Right: Zoomed-in slice showing the radial velocity of the gas centered on the most massive cluster. It shows that the gas flows nears the growing cluster is highly anisotropic. Right: Radial velocity-radius phase diagram of the gas within 5 pc of the most massive star cluster. The pixel color reflects the total gas mass at a given state. A significant portion of the gas is subject to a strong infall even after the cluster reaches stellar mass of 10 6 M . M * = 10 4 M M * = 10 6 M

Upload: others

Post on 04-Feb-2020

6 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: Radiation Hydrodynamics of Super Star Cluster Nucleiconference.astro.ufl.edu/STARSTOMASSIVE/eproceedings/talks/tsang_b.pdfRadiation Hydrodynamics of Super Star Cluster Nuclei! Benny

Radiation Hydrodynamics of Super Star Cluster Nuclei!Benny Tsz-Ho Tsang & Milos Milosavljevic!

The University of Texas at Austin!

From Stars to Massive Stars, Gainesville, Florida, 6th – 9th April 2016!

Super Star Clusters!

Radiation Transport Matters!

What is the Role of Radiation?!

Simulating their Births!

Results!

Super star clusters (SSCs) are the most extreme star-forming sites in the Universe: massive stars at the cluster nuclei form from gas surface densities of 103 – 104 M¤ pc−2, with star formation rates above 10 – 100 M¤ yr−1. The intense radiation field from the dense population of young, massive stars is expected to be dynamically important in altering the star-forming gas. The resultant massive star densities are so high that stellar merging may be common, providing a potential pathway for massive black hole formation.!!Large distances to these extreme objects and dust obscuration have been limiting our observational access; theoretical models must realize the multidimensional complexity of dusty, radiation-dominated gas on scales < 0.1 pc inside the nucleus actively forming massive stars.!

Studies have shown that radiation pressure could dominate the disruption of parent molecular clouds in forming massive clusters (Krumholz & Matzner 2009, Murray et al. 2010, Kim et al. 2016). However, these predictions were based on one-dimensional models. !!•  Does stellar radiation pressure from young stars inhibit or terminate

star formation in actual SSC? !•  How long does star formation last in SSC? !•  Does stellar radiation drive intense outflows?!

We implemented a closure-free, Monte Carlo radiation transfer scheme coupled with the hydrodynamic code FLASH. We confirmed that the accuracy of the radiation transport method could have a huge impact on the global behaviors of gas dynamics when radiation pressure is dominant and steep transitions in optical thickness is present.!!•  Initially static layer of dusty gas subject to a constant upward radiation

flux and a constant downward gravity. !•  Same exact setup has been tested with three other radiation transport

methods, listed in increasing accuracy: flux-limited diffusion (FLD), M1 closure, and the variable Eddington tensor approximation (VET). !

We aim to identify the physical conditions (e.g., cloud mass and surface density) in which radiation pressure’s effects on turbulent molecular clouds is dominant. With a particle-based radiation transport scheme, our simulations will automatically deliver the full observable spectral energy distributions (SEDs) of SSCs. They will motivate and inform observations of SSCs in formation.!

Left: Gas density snapshots in the setup of radiation pressure-driven atmosphere. The flow morphology observed in our Monte Carlo scheme is consistent with other radiation transport methods. !!Right: Time evolution of the mass-weighted mean vertical velocity (upper) and velocity dispersion (lower). The late-time net acceleration and velocity dispersions are in agreement only with the VET method. It is therefore essential to adopt proper radiation transport in simulations with strong radiation pressure and steep transitions in optical thickness.!

Left: Time evolution of the fraction of gas mass with positive radial velocity (outflow) within 1 pc radius of the three most massive clusters. !

We simulate the formation of super star clusters from a 108 M¤ giant molecular cloud, focusing on the feedback by stellar radiation pressure. Physics included are:!•  Initial density/velocity distributions set by driven turbulence;!•  Gas self-gravity and star-gas gravity;!•  Star particles created as formed clusters and radiation sources;!•  Frequency-dependent transfer of non-ionizing UV and dust-

reprocessed IR radiation;!•  Full coupling of radiation pressure with gas dynamics.!

The stellar radiation field can be super-Eddington locally around the massive clusters. It does not seem to effectively limit the infall of gas or terminate star formation due to the turbulent nature of the clouds and the strong infall of gas towards the growing clusters.!!

Left: Mass-weighted gas density projection of the turbulent initial conditions. The initial surface densities before gravity is turned on are > 1000 M¤ pc−2, typical for starburst environments. !!Right: The same projection one free-fall time after gravity has been turned on, the black points mark the projected locations of the star particles representing the luminous clusters formed. Also shown is the zoomed-in view of the most massive cluster formed.!

Davis S. W., Jiang Y.-F., Stone J. M., Murray N., 2014, ApJ, 796, 107 (VET).! Kim, J. G., Kim, W. T., Ostriker E. C., 2016, ApJ, 819, 137.!Krumholz, M. R., & Matzner, C. D., 2009, ApJ, 703, 1352.! Krumholz, M. R., & Thompson, T. A., 2012, ApJ, 760, 155 (FLD).!Krumholz, M. R., & Thompson, T. A., 2013, MNRAS, 434, 2329 (FLD).! Murray, N., Quataert, E., & Thompson, T. A. 2010, ApJ, 709, 191.!Rosdahl J., Teyssier R., 2015, MNRAS, 449, 4380 (M1 closure).! Tsang, B. T.-H., & Milosavljevic, M. 2015. MNRAS, 453, 1108. !

References:!

Right: Zoomed-in slice showing the radial velocity of the gas centered on the most massive cluster. It shows that the gas flows nears the growing cluster is highly anisotropic.!

Right: Radial velocity-radius phase diagram of the gas within 5 pc of the most massive star cluster. The pixel color reflects the total gas mass at a given state. A significant portion of the gas is subject to a strong infall even after the cluster reaches stellar mass of 106 M¤. ! !

M* = 104 M¤! M* = 106 M¤!