Realistic Simulations of Collisionless Black Hole Accretion Flows

这个项目将是对天体物理学特别感兴趣的一种重要类型的黑洞流动的第一个现实研究。这些特别的流动是非常热的，但不是非常密集的，大多数情况下都是使用假设对这种特定条件组合是错误的方法来处理的。这项工作将从基本原理开始，将小尺度和大尺度的重要物理都包括在新的数值计算中。初级研究人员将学习尖端技术，这也有助于推动计算机硬件和软件系统的发展。附近超大质量黑洞周围的吸积流包含热和扩散等离子体，其中碰撞平均自由程大于系统大小，耗散效应决定了它们的热力学和观测性质。这些吸积流的理论模型大多是基于理想磁流体动力学(MHD)，这是一个碰撞流体理论。然而，这个公式中的一些假设在黑洞周围热的、低光度的吸积流中被打破了。这个项目是一项第一原理的计算和理论研究，重点是能量耗散、输运和粒子加速。它首先使用了一种广义相对论测地线算法，旨在整合光子和粒子测地线，并针对图形处理单元(GPU)等特殊用途的架构进行了优化。大规模回转动力学模拟将研究热、湍流、无碰撞等离子体的粒子级和全球级特征。其次，将使用粒子在细胞中的代码来模拟等离子体的微观物理。这将研究MHD模拟中磁重联区域电流片中的粒子加速，以量化非理想效应，并包括低辐射效率吸积流的流体力学和光谱研究的结果。一名研究生将接受培训，一名博士后将接受计算、理论和等离子体天体物理方面的指导。这项研究使用NSF支持的GPU计算资源将有助于保持美国的尖端技术发展，粒子积分器算法将公开可用。研究人员还将开展广泛的外联活动，包括公开演讲、热门文章和纪录片。

This project will be the first realistic study of an important type of flow onto black holes, of special interest in astrophysics. These particular flows, which are very hot but not very dense, have mostly been treated using methods whose assumptions are wrong for this particular combination of conditions. This work will start from first principles to include the important physics at both small and large scales in new numerical calculations. Junior researchers will learn about cutting edge techniques, which also help to drive the development of computer hardware and software systems.Accretion flows around nearby supermassive black holes contain hot and diffuse plasmas where the collisional mean free path is larger than the system size and the dissipative effects determine their thermodynamic and observational properties. Most theoretical models of these accretion flows are based on ideal magnetohydrodynamics (MHD), which is a collisional fluid theory. However, a number of assumptions in this formulation break down in hot, low luminosity accretion flows around black holes. This project is a first-principles computational and theoretical study with a focus on energy dissipation, transport, and particle acceleration. It uses firstly a general relativistic geodesic algorithm designed to integrate both photon and particle geodesics, and optimized for special-purpose architectures such as graphical processing units (GPUs). Massive scale gyrokinetic simulations will study both particle-level and global-level characteristics of hot, turbulent, collisionless plasmas. Secondly, a particle-in-cell code will be used for simulations focused on the microphysics of the plasma. This will study particle acceleration in current sheets in magnetic reconnection regions of the MHD simulations to quantify non-ideal effects and include the results in hydrodynamic and spectral studies of low radiative efficiency accretion flows.One graduate student will be trained and a postdoctoral associate will be supervised in computational, theoretical, and plasma astrophysics. The study's use of NSF-supported GPU computational resources will help to maintain cutting edge technology development in the United States, and the particle integrator algorithm will be made publicly available. The researchers will also run an extensive outreach effort, with public talks, popular articles, and documentaries.