Brad Gibson Centre for Astrophysics & Supercomputing Swinburne University

Size: px
Start display at page:

Download "Brad Gibson Centre for Astrophysics & Supercomputing Swinburne University"

Transcription

1 N-body Simulations: Tree and Mesh Approaches Brad Gibson Centre for Astrophysics & Supercomputing Swinburne University

2 Demystifying the Jargon Trees Collisionless Boltzmann Equation Symplectic Time Integration Eulerian Reference Frame Hermite Time Integration Leap-Frog Timestepping Poisson s Equation Comoving Coordinates Meshe s Lagrangian Reference Frame Green s Functions Gravitational Softening Vlasov Equation Grids AMR

3 N-body Simulations: Overview Basic equations (by definition, anathema to a good talk!) Basic techniques PP, PM, P3M, Tree, FMM, TreePM Basic tools hardware, software Are all codes created equal? (strengths & weaknesses) Resources

4 N-body Simulations: Basic Structure Chris Power! All N-body codes consist of two basic modules: 1. Force field computation 2. Particle trajectory computation Both modules are called at each time step to ensure that the force field and particle trajectories evolve in a self-consistent manner. Chris Power

5 N-body Simulations: Basic Equations Dark matter and stars are modeled as self-gravitating collisionless fluids - i.e., they obey the Vlasov/Diffusion/Collisionless Boltzmann equation: where the self-consistent potential Φ is the solution of Poisson s eqn: F = -sφ and f(r,v,t) is the mass density in single-particle phase space. The N-body problem is the task of following Newton s equations of motion for a large number of particles under their own selfgravity.

6 N-body Simulations: Basic Equations Let s start simple and examine the N=2 scenario, and work in terms of the relative position: Newton s gravitational equation of motion then becomes: where and

7 N-body Simulations: Basic Equations Keep it further simple by setting G=1 and M 1 +M 2 =1: We now have the equations of motion for the relative position of one particle with respect to the other. We now need an algorithm to integrate these equations. Order of algorithm? Low? High? Structure of algorithm? Euler, Leapfrog, Hermite, Symplectic

8 N-body Simulations: Forward-Euler Algorithm What is the order of an algorithm? Signifies the rate of convergence Since no algorithm with a finite time step size is perfect, they all make numerical errors. In a 4th-order algorithm, this error scales as the 4th power of the time step. High-order integrators are complex to code and their region of convergence is limited There is an optimal order for each type of problem (well-behaved allows higher-orders)

9 N-body Simulations: Forward-Euler Algorithm What should the structure of the algorithm be? Simplest is 1st-order forward Eulerian integrator At each time step, you simply take a step tangential to the orbit you are on. At the next time step, the new value of the acceleration forces you to slightly change direction, and again you move for a step dt in a straight line in that direction. Orbit is thus constructed out of a number of straight line segments, where each one has the proper direction at the beginning of the segment but the wrong one at the end. Backwards Eulerian entails getting the right direction at the end of the time step (which is done via iteration guess a direction and then you correct for the mistake so that your second iteration is more accurate) Can also split the difference, which then becomes 2nd-order accurate (e.g. 2nd-order Runge-Kutta and Leapfrog)

10 N-body Simulations: Forward-Euler Algorithm What should the structure of the algorithm be? Simplest is 1st-order forward Eulerian integrator Above is notation for position and velocity of an individual particle, after time step dt. Acceleration is a. We now have enough in the way of equations to write an N-body code!

11 An N-body Code in <10 Lines (N=2) Initial Conditions (relative position & velocity)

12 N-body Simulations: Forward-Euler Algorithm Time step too large; artificial slingshot where orbital curvature is highest (just after pericenter)

13 N-body Simulations: Forward-Euler Algorithm Time step too large; artificial slingshot where orbital curvature is highest (just after pericenter)

14 N-body Simulations: Forward-Euler Algorithm Time step too large; artificial slingshot where orbital curvature is highest (just after pericenter)

15 N-body Simulations: Forward-Euler Algorithm Time step too large; artificial slingshot where orbital curvature is highest (just after pericenter)

16 N-body Simulations: Forward-Euler Algorithm Time step too large; artificial slingshot where orbital curvature is highest (just after pericenter)

17 N-body Simulations: Forward-Euler Algorithm In reality, many-body simulations do not have analytical solutions which would allow us to gauge convergence. Are there conserved quantities which could be used to better gauge convergence? Yes! Energy conservation is the simplest to employ, and must be rigorously conserved, so any deviations must be numerical. Our 1st-order forward Eulerian test does converge, but slowly!

18 N-body Simulations: Leapfrog Algorithm The leapfrog algorithm is one of the most popular higher-order integrators in use today. The name derives from the fact that positions and velocities leap over each other. Positions are defined at times t i,t i+1,, spaced at constant dt, while velocities are defined at times halfway between, or t i-1/2,t i+1/2, The leapfrog algorithm then reads: Accelerations defined only on integer times, just like positions while velocities are defined only on half-integer times.

19 N-body Simulations: Leapfrog Algorithm At the beginning of the integration we set up the velocity at its first half-integer time step. Starting with initial conditions r 0 and v 0, we take 1st term in Taylor series expansion to compute 1st leap value for v : We then use to compute the new position r 1 using the 1st leap value for v 1/2. Next we compute the acceleration a 1, which enables us to compute the 2nd leap value v 3/2 using and so on.

20 N-body Simulations: Leapfrog Algorithm dt=0.01 dt=0.001 Forward Eulerian

21 N-body Simulations: Calculating the Force An N-body code solves the equation of motion and the Poisson equation self-consistently. Moving particles and computing the force for a given particle distribution are a code s two most important components. Currently there are two commonly used approaches for deriving the potential from Poisson s equation: Tree Codes (particle-particle/group summation) PM (particle-mesh: integrate Poisson s equation on a grid)

22 N-body Simulations: Tree Codes Calculating the force at every single particle position via direct summation of Newton s laws between all pairs of particles means the PP method scales like N 2 just not feasible for large N. Can we get more clever? Yes, by organising particles in a treelike structure. Particles located far away from the particle in question are grouped together as a single (but more massive) particle. This has the effect of dramatically reducing the number of calculations.

23 N-body Simulations: Tree Codes Decompose computational domain hierarchically into a sequence of cubes, where each cube contains 8 siblings (oct-tree), each with half the side-length of the parent. Each node will contain either 1 particle or is progenitor to further nodes, in which case the node carries monopole and quadrupole moments of all the particles that lie inside its cube.

24 N-body Simulations: Tree Codes Force computation then proceeds by walking the tree and summing the appropriate force contributions from tree nodes. A user-specified opening criterion is employed to decide whether or not a particular node needs to be opened into finer branches if the latter is deemed necessary, the walk will continue with all of its siblings. Smaller opening angle=higher accuracy=more expensive

25 Multi-resolution CDM with GCD+ High-z snapshot High Resolution Region m DM =2x10 5 M, ε DM =0.14 kpc, m gas =3x10 4 M, ε gas =0.08 kpc J-band 5kpc = 0.83 M vir = 6x10 9 M V max = 65 km/s

26 N-body Simulations: Particle-Mesh (PM) Codes Another way to obtain forces is to numerically integrate Poisson s equation. This demands the introduction of a grid in order to define density. The grid is usually of a regular (cubic) shape. Forces are then calculated as such: assign all particles to the grid to get ρ solve Poisson s equation s 2 Φ = 4πGρ This is where most differentiate to get forces F=-sΦ of the time is spent interpolate F back to particle positions The most common way to solve Poisson s equation on a grid is to make use of FFTs.

27 N-body Simulations: Particle-Mesh (PM) Codes The analytical solution to Poisson s equation is given by the integral: where which is readily solved in Fourier space: is the Green s Function of Poisson s eqn. FTs of the respective variables.

28 N-body Simulations: Particle-Mesh (PM) Codes The PM approach is much faster than Tree codes. The most severe problem with PM codes is the lack of spatial resolution due to the underlying grid we cannot resolve structure on scales of the grid s cell spacing. Major problem! solution? Introduce finer grids in regions of high density

29 the simulations: n-bodies the n-body code grid hierarchy Multi Level Adaptive Particle Mesh only open source AMR code written in C extremely memory efficient extremely fast MPI parallelisation (planned ) hydrodynamics (planned ) on-the-fly analysis: MLAPM s grids naturally find substructure within halos!

30 the simulations: sample rd th refinement halo at level z=0 (out of 8) the satellite finder (Gill, Knebe, & Gibson 2003)

31 N-body Simulations: Tree vs PM Tree faster/more adaptive to highly clustered situations (Lagrangian = dynamic range) basically free of geometric restrictions two-body interactions more prominent PM faster for nearly homogeneous/non-clustered situations periodic boundary conditions more readily incorporated

32 Basic Tools: Supercomputers! 500 kcpu Swinburne, APAC, VPAC, NAOJ

33 Basic Tools: Software GCD+/GADGET Tree N-body + Smoothed Particle Hydrodynamics (parallel+vector) Star formation, energy feedback, chemical evolution, panchromatic MLAPM Adaptive Mesh Refinement (serial) Highly efficient, on-the-fly visualisation and analysis tools

34 Basic Tools: The Astonishing Approach Holmberg (1941) Truly astonishing and prescient N-body simulation Light Intensity=Gravity Used the fact that both fall off like 1/r 2 to simulate in-plane tidal interactions between disk galaxies! N=37*2 OK so N was only 74 it s still remarkable! Light Bulb Photocell + Galvanometer 80cm Initial Condition: Single Disk Galaxy; N=37

35 Basic Tools: The Astonishing Approach Simulations #hyperbolic & parabolic orbits #galactic diameter 2.5kpc #galactic mass10 11 M #interaction speed 450 km/ s #pericenter 2.5 kpc #realistic surface brightness #multi-mass via varying voltage #3mx4m graph paper #tangents via French Curves Results #prograde-prograde leads to trailing spiral arms #spiral/tidal arms appear prograde- retrograde-retrograde prograde

36 Resources N-body and Hydrodynamics Links (primarily cosmological and galactic) The Art of Computational Science (PP) GADGET (TreePM) Enzo (PM) MLAPM

37

High performance computing and numerical modeling

High performance computing and numerical modeling High performance computing and numerical modeling Volker Springel Plan for my lectures Lecture 1: Collisional and collisionless N-body dynamics Lecture 2: Gravitational force calculation Lecture 3: Basic

More information

N-body Cosmology: The billion-body problem. Brian O Shea Michigan State University

N-body Cosmology: The billion-body problem. Brian O Shea Michigan State University N-body Cosmology: The billion-body problem Brian O Shea (oshea@msu.edu) Michigan State University TIARA Winter School, Jan. 2011 This lecture Goals of cosmological N-body simulations Overview of an N-body

More information

Computational Astrophysics 1 Particle Mesh methods

Computational Astrophysics 1 Particle Mesh methods Computational Astrophysics 1 Particle Mesh methods Oscar Agertz Outline - The Vlassov-Poisson equation - The Particle-Mesh method in a nut shell - Symplectic Integrators - Mass assignment and force interpolation

More information

Treecodes for Cosmology Thomas Quinn University of Washington N-Body Shop

Treecodes for Cosmology Thomas Quinn University of Washington N-Body Shop Treecodes for Cosmology Thomas Quinn University of Washington N-Body Shop Outline Motivation Multipole Expansions Tree Algorithms Periodic Boundaries Time integration Gravitational Softening SPH Parallel

More information

Dynamics of Galaxies: Basics. Frontiers in Numerical Gravitational Astrophysics July 3, 2008

Dynamics of Galaxies: Basics. Frontiers in Numerical Gravitational Astrophysics July 3, 2008 Dynamics of Galaxies: Basics Frontiers in Numerical Gravitational Astrophysics July 3, 2008 c = G = 1 h = 0 Outline (a) Galaxies as physical systems. (b) The Collisionless Boltzmann Equation. (c) N-body

More information

Swift: task-based hydrodynamics at Durham s IPCC. Bower

Swift: task-based hydrodynamics at Durham s IPCC. Bower Swift: task-based hydrodynamics at Durham s IPCC Gonnet Schaller Chalk movie: Richard Bower (Durham) For the cosmological simulations of the formation of galaxies Bower Institute for Computational Cosmology

More information

Galaxy interaction and transformation

Galaxy interaction and transformation Galaxy interaction and transformation Houjun Mo April 13, 2004 A lot of mergers expected in hierarchical models. The main issues: The phenomena of galaxy interaction: tidal tails, mergers, starbursts When

More information

NUMERICAL METHODS IN ASTROPHYSICS An Introduction

NUMERICAL METHODS IN ASTROPHYSICS An Introduction -1 Series in Astronomy and Astrophysics NUMERICAL METHODS IN ASTROPHYSICS An Introduction Peter Bodenheimer University of California Santa Cruz, USA Gregory P. Laughlin University of California Santa Cruz,

More information

A Comparison of AMR and SPH cosmology codes

A Comparison of AMR and SPH cosmology codes Chapter 3 A Comparison of AMR and SPH cosmology codes 3.1 Summary We compare two cosmological hydrodynamic simulation codes in the context of hierarchical galaxy formation: the Lagrangian smoothed particle

More information

Structure formation. Yvonne Y. Y. Wong Max-Planck-Institut für Physik, München

Structure formation. Yvonne Y. Y. Wong Max-Planck-Institut für Physik, München Structure formation Yvonne Y. Y. Wong Max-Planck-Institut für Physik, München Structure formation... Random density fluctuations, grow via gravitational instability galaxies, clusters, etc. Initial perturbations

More information

Simulations of structure-formation in theories beyond General Relativity

Simulations of structure-formation in theories beyond General Relativity Simulations of structure-formation in theories beyond General Relativity Hans A. Winther Diving into the Dark, Cairns 21 July 2016 Hans Winther University of Oxford 2016 Motivation Why modified gravity?

More information

CHAPTER 4. Basics of Fluid Dynamics

CHAPTER 4. Basics of Fluid Dynamics CHAPTER 4 Basics of Fluid Dynamics What is a fluid? A fluid is a substance that can flow, has no fixed shape, and offers little resistance to an external stress In a fluid the constituent particles (atoms,

More information

Outline. 1 Why PIC Simulations Make Sense. 2 The Algorithm. 3 Examples Surface High Harmonics Generation. 4 Extensions Of The PIC Algorithm

Outline. 1 Why PIC Simulations Make Sense. 2 The Algorithm. 3 Examples Surface High Harmonics Generation. 4 Extensions Of The PIC Algorithm PIC Simulations an Introduction GRK 1203 Meeting February 12-15, 2008, Oelde Outline 1 Simulations Make Sense 2 3 Surface High Harmonics Generation 4 Of PIC Plasma Physics Is Complex Experiment real thing

More information

Everything in baryons?

Everything in baryons? Everything in baryons? Benoit Famaey (ULB) Rencontres de Blois 2007 General Relativity -> Dark Matter R αβ - 1/2 R g αβ + Λg αβ = (8πG/c 4 ) T αβ very precisely tested on solar system scales (but Pioneer)

More information

The motions of stars in the Galaxy

The motions of stars in the Galaxy The motions of stars in the Galaxy The stars in the Galaxy define various components, that do not only differ in their spatial distribution but also in their kinematics. The dominant motion of stars (and

More information

A Universe in Motion: Testing the Cosmological Paradigm with Galaxy Dynamics. John Dubinski, Toronto

A Universe in Motion: Testing the Cosmological Paradigm with Galaxy Dynamics. John Dubinski, Toronto A Universe in Motion: Testing the Cosmological Paradigm with Galaxy Dynamics John Dubinski, Toronto Outline Cosmology and galaxy dynamics Tools of the trade and computational challenges Case Studies: Triaxial

More information

N-body Simulations and Dark energy

N-body Simulations and Dark energy N-Body Simulations and models of Dark Energy Elise Jennings Supported by a Marie Curie Early Stage Training Fellowship N-body Simulations and Dark energy elise jennings Introduction N-Body simulations

More information

Numerical Simulations. Duncan Christie

Numerical Simulations. Duncan Christie Numerical Simulations Duncan Christie Motivation There isn t enough time to derive the necessary methods to do numerical simulations, but there is enough time to survey what methods and codes are available

More information

arxiv:astro-ph/ v1 10 Dec 1996

arxiv:astro-ph/ v1 10 Dec 1996 N-body Simulation of Galaxy Formation on GRAPE-4 Special-Purpose Computer arxiv:astro-ph/9612090v1 10 Dec 1996 Toshiyuki Fukushige and Junichiro Makino Department of General Systems Studies, College of

More information

Michela Mapelli. N-body techniques for astrophysics: Lecture 1 General Introduction

Michela Mapelli. N-body techniques for astrophysics: Lecture 1 General Introduction Michela Mapelli N-body techniques for astrophysics: Lecture 1 General Introduction PhD School in Astrophysics, University of Padova November 3-12, 2015 A few words about the course: This course is NOT:

More information

The Local Group Timing Argument

The Local Group Timing Argument The Local Group Timing Argument Real galaxies Milky Way (MW) Andromeda (M31) Indranil Banik (ib45@st-andrews.ac.uk) Supervisor: Hongsheng Zhao University of Saint Andrews MNRAS, 467 (2), 2180 2198 Basic

More information

Computational Astrophysics

Computational Astrophysics Computational Astrophysics Lecture 1: Introduction to numerical methods Lecture 2: The SPH formulation Lecture 3: Construction of SPH smoothing functions Lecture 4: SPH for general dynamic flow Lecture

More information

Numerical Cosmology & Galaxy Formation

Numerical Cosmology & Galaxy Formation Numerical Cosmology & Galaxy Formation Lecture 13: Example simulations Isolated galaxies, mergers & zooms Benjamin Moster 1 Outline of the lecture course Lecture 1: Motivation & Historical Overview Lecture

More information

Shedding Light on Dark Matter

Shedding Light on Dark Matter Shedding Light on Dark Matter can t actually be done while these particles make up 82% of the matter in the Universe, they don t interact with light. So how do we know anything about these invisible particles?

More information

The Magellanic Stream Tim Connors Dr. D. Kawata Prof. B. Gibson

The Magellanic Stream Tim Connors Dr. D. Kawata Prof. B. Gibson High-resolution N-body Simulations of Galactic Cannibalism The Magellanic Stream Tim Connors Dr. D. Kawata Prof. B. Gibson tconnors@astro.swin.edu.au Swinburne University of Technology Page 1 of 31 July

More information

AM 205 Final Project The N-Body Problem

AM 205 Final Project The N-Body Problem AM 205 Final Project The N-Body Problem Leah Birch Elizabeth Finn Karen Yu December 14, 2012 Abstract The N-Body Problem can be solved using a variety of numeric integrators. Newton s Law of Universal

More information

N-Body Methods. Chapter Numerical Stellar Dynamics Monte carlo methods Representing f (r;v;t)

N-Body Methods. Chapter Numerical Stellar Dynamics Monte carlo methods Representing f (r;v;t) Chapter 9 N-Body Methods The Collisionless Boltzmann and Poisson equations can be effectively solved with a Monte-Carlo technique which represents the distribution function f (r;v;t) as a set of N bodies.

More information

Figure Rotation of Dark Halos in Cold Dark Matter Simulations

Figure Rotation of Dark Halos in Cold Dark Matter Simulations Figure Rotation of Dark Halos in Cold Dark Matter Simulations by Sarah Bryan Submitted in fulfilment of the academic requirements for the degree of Master of Science in the School of Physics, University

More information

Dynamics of self-gravitating particles in an expanding universe!

Dynamics of self-gravitating particles in an expanding universe! Dynamics of self-gravitating particles in an expanding universe! Michael Joyce LPNHE Université Pierre et Marie Curie - Paris VI Acknowledgements: my collaborators! T. Baertschiger (La Sapienza, Rome)

More information

COLA with scale dependent growth: applications to modified gravity and massive neutrinos

COLA with scale dependent growth: applications to modified gravity and massive neutrinos COLA with scale dependent growth: applications to modified gravity and massive neutrinos Kazuya Koyama Institute of Cosmology and Gravitation, University of Portsmouth Curvature Psaltis Living Rev. Relativity

More information

Cosmological N-Body Simulations and Galaxy Surveys

Cosmological N-Body Simulations and Galaxy Surveys Cosmological N-Body Simulations and Galaxy Surveys Adrian Pope, High Energy Physics, Argonne Na3onal Laboratory, apope@anl.gov CScADS: Scien3fic Data and Analy3cs for Extreme- scale Compu3ng, 30 July 2012

More information

Cosmological shock waves

Cosmological shock waves Highlights of Spanish Astrophysics VII, Proceedings of the X Scientific Meeting of the Spanish Astronomical Society held on July 9-13, 2012, in Valencia, Spain. J. C. Guirado, L. M. Lara, V. Quilis, and

More information

Practical in Numerical Astronomy, SS 2012 LECTURE 9

Practical in Numerical Astronomy, SS 2012 LECTURE 9 Practical in Numerical Astronomy, SS 01 Elliptic partial differential equations. Poisson solvers. LECTURE 9 1. Gravity force and the equations of hydrodynamics. Poisson equation versus Poisson integral.

More information

New Results from 3-D supernova models with spectral neutrino diffusion

New Results from 3-D supernova models with spectral neutrino diffusion New Results from 3-D supernova models with spectral neutrino diffusion Stuart C. Whitehouse and Matthias Liebendörfer Department of Physics, University of Basel, Switzerland Overview Introduction The Isotropic

More information

An optimum time-stepping scheme for N-body simulations

An optimum time-stepping scheme for N-body simulations Mon. Not. R. Astron. Soc. 376, 73 86 (007) doi:10.1111/j.1365-966.007.1147.x An optimum time-stepping scheme for N-body simulations Marcel Zemp, 1 Joachim Stadel, Ben Moore and C. Marcella Carollo 1 1

More information

OBSERVATIONAL EVIDENCE FOR DARK MATTER AND DARK ENERGY. Marco Roncadelli INFN Pavia (Italy)

OBSERVATIONAL EVIDENCE FOR DARK MATTER AND DARK ENERGY. Marco Roncadelli INFN Pavia (Italy) OBSERVATIONAL EVIDENCE FOR DARK MATTER AND DARK ENERGY Marco Roncadelli INFN Pavia (Italy) ABSTRACT Assuming KNOWN physical laws, I first discuss OBSERVATIONAL evidence for dark matter in galaxies and

More information

Simulating the Universe

Simulating the Universe Simulating the Universe Christine Corbett Moran, Irshad Mohammed, Manuel Rabold, Davide Martizzi, Doug Potter, Aurel Schneider Oliver Hahn, Ben Moore, Joachim Stadel Outline - N-body codes: where do we

More information

SPH simulations of galaxy formation

SPH simulations of galaxy formation SPH simulations of galaxy formation Looking under the hood of GADGET-2 Volker Springel Large N-body runs Algorithmic aspects of a new simulation code: GADGET- II Non-standard physics in simulations of

More information

AST541 Lecture Notes: Galaxy Formation Dec, 2016

AST541 Lecture Notes: Galaxy Formation Dec, 2016 AST541 Lecture Notes: Galaxy Formation Dec, 2016 GalaxyFormation 1 The final topic is galaxy evolution. This is where galaxy meets cosmology. I will argue that while galaxy formation have to be understood

More information

N-body Simulations. Initial conditions: What kind of Dark Matter? How much Dark Matter? Initial density fluctuations P(k) GRAVITY

N-body Simulations. Initial conditions: What kind of Dark Matter? How much Dark Matter? Initial density fluctuations P(k) GRAVITY N-body Simulations N-body Simulations N-body Simulations Initial conditions: What kind of Dark Matter? How much Dark Matter? Initial density fluctuations P(k) GRAVITY Final distribution of dark matter.

More information

A Fast N-Body Solver for the Poisson(-Boltzmann) Equation

A Fast N-Body Solver for the Poisson(-Boltzmann) Equation A Fast N-Body Solver for the Poisson(-Boltzmann) Equation Robert D. Skeel Departments of Computer Science (and Mathematics) Purdue University http://bionum.cs.purdue.edu/2008december.pdf 1 Thesis Poisson(-Boltzmann)

More information

Lecture V: The game-engine loop & Time Integration

Lecture V: The game-engine loop & Time Integration Lecture V: The game-engine loop & Time Integration The Basic Game-Engine Loop Previous state: " #, %(#) ( #, )(#) Forces -(#) Integrate velocities and positions Resolve Interpenetrations Per-body change

More information

Time scales. Notes. Mixed Implicit/Explicit. Newmark Methods [ ] [ ] "!t = O 1 %

Time scales. Notes. Mixed Implicit/Explicit. Newmark Methods [ ] [ ] !t = O 1 % Notes Time scales! For using Pixie (the renderer) make sure you type use pixie first! Assignment 1 questions?! [work out]! For position dependence, characteristic time interval is "!t = O 1 % $ ' # K &!

More information

Toward Binary Black Hole Simulations in Numerical Relativity

Toward Binary Black Hole Simulations in Numerical Relativity Toward Binary Black Hole Simulations in Numerical Relativity Frans Pretorius California Institute of Technology BIRS Workshop on Numerical Relativity Banff, April 19 2005 Outline generalized harmonic coordinates

More information

Geometric methods for orbit integration. PiTP 2009 Scott Tremaine

Geometric methods for orbit integration. PiTP 2009 Scott Tremaine Geometric methods for orbit integration PiTP 2009 Scott Tremaine spacecraft trajectories Cassini-Huygens trajectory around Saturn, 2004-2008 Planetary orbits lines = current orbits of the four inner planets

More information

Towards Understanding Simulations of Galaxy Formation. Nigel Mitchell. On the Origin of Cores in Simulated Galaxy Clusters

Towards Understanding Simulations of Galaxy Formation. Nigel Mitchell. On the Origin of Cores in Simulated Galaxy Clusters Towards Understanding Simulations of Galaxy Formation Nigel Mitchell On the Origin of Cores in Simulated Galaxy Clusters Work published in the Monthly Notices of the Royal Astronomy Society Journal, 2009,

More information

Circular Motion and Gravity Lecture 5

Circular Motion and Gravity Lecture 5 Circular Motion and Gravity Lecture 5 ˆ Today we talk about circular motion. There are two reasons to do this... ˆ Last week we talked about Newton s laws in problems dealing with straight-line motion.

More information

Benchmarks in Computational Plasma Physics

Benchmarks in Computational Plasma Physics Benchmarks in Computational Plasma Physics P. Londrillo INAF, Bologna, Italie S. Landi Università di Firenze, Italie What you compute when you do computations of the Vlasov equation? Overview A short review

More information

arxiv:astro-ph/ v1 22 Sep 2005

arxiv:astro-ph/ v1 22 Sep 2005 Mass Profiles and Shapes of Cosmological Structures G. Mamon, F. Combes, C. Deffayet, B. Fort (eds) EAS Publications Series, Vol.?, 2005 arxiv:astro-ph/0509665v1 22 Sep 2005 MONDIAN COSMOLOGICAL SIMULATIONS

More information

Stellar Dynamics and Structure of Galaxies

Stellar Dynamics and Structure of Galaxies Stellar Dynamics and Structure of Galaxies Gerry Gilmore H47 email: gil@ast.cam.ac.uk Lectures: Monday 12:10-13:00 Wednesday 11:15-12:05 Friday 12:10-13:00 Books: Binney & Tremaine Galactic Dynamics Princeton

More information

ASTR 610 Theory of Galaxy Formation Lecture 4: Newtonian Perturbation Theory I. Linearized Fluid Equations

ASTR 610 Theory of Galaxy Formation Lecture 4: Newtonian Perturbation Theory I. Linearized Fluid Equations ASTR 610 Theory of Galaxy Formation Lecture 4: Newtonian Perturbation Theory I. Linearized Fluid Equations Frank van den Bosch Yale University, spring 2017 Structure Formation: The Linear Regime Thus far

More information

8.1 Structure Formation: Introduction and the Growth of Density Perturbations

8.1 Structure Formation: Introduction and the Growth of Density Perturbations 8.1 Structure Formation: Introduction and the Growth of Density Perturbations 1 Structure Formation and Evolution From this (Δρ/ρ ~ 10-6 ) to this (Δρ/ρ ~ 10 +2 ) to this (Δρ/ρ ~ 10 +6 ) 2 Origin of Structure

More information

Things I don't understand about Vlasov-Poisson numerics

Things I don't understand about Vlasov-Poisson numerics Things I don't understand about Vlasov-Poisson numerics 1. Modeling stellar systems 2. Interpolation in simulations of cold dark matter 3. Orbit integration in N-body systems 4. How accurately should we

More information

Lecture IV: Time Discretization

Lecture IV: Time Discretization Lecture IV: Time Discretization Motivation Kinematics: continuous motion in continuous time Computer simulation: Discrete time steps t Discrete Space (mesh particles) Updating Position Force induces acceleration.

More information

Moving mesh cosmology: The hydrodynamics of galaxy formation

Moving mesh cosmology: The hydrodynamics of galaxy formation Moving mesh cosmology: The hydrodynamics of galaxy formation arxiv:1109.3468 Debora Sijacki, Hubble Fellow, ITC together with: Mark Vogelsberger, Dusan Keres, Paul Torrey Shy Genel, Dylan Nelson Volker

More information

Cluster Formation in Lagrangian Perturbation Theory. Lavinia Heisenberg

Cluster Formation in Lagrangian Perturbation Theory. Lavinia Heisenberg Cluster Formation in Lagrangian Perturbation Theory Zentrum fu r Astronomie (ITA/ZAH) Universita t Heidelberg 18.12.2009, University of Canterbury, Christchurch outline 1 Introduction 2 LPT 3 ELL 4 PINOCCHIO

More information

isocurvature modes Since there are two degrees of freedom in

isocurvature modes Since there are two degrees of freedom in isocurvature modes Since there are two degrees of freedom in the matter-radiation perturbation, there must be a second independent perturbation mode to complement the adiabatic solution. This clearly must

More information

Energy Balance in Clusters of Galaxies. Patrick M. Motl & Jack O. Burns Center for Astrophysics and Space Astronomy University of Colorado at Boulder

Energy Balance in Clusters of Galaxies. Patrick M. Motl & Jack O. Burns Center for Astrophysics and Space Astronomy University of Colorado at Boulder Energy Balance in Clusters of Galaxies Patrick M. Motl & Jack O. Burns Center for Astrophysics and Space Astronomy University of Colorado at Boulder X-ray and Radio Connections, February 6th, 2004 With

More information

Computer Simulation Using Particles

Computer Simulation Using Particles Computer Simulation Using Particles R W Hockney Emeritus Professor, University ofreading and J W Eastwood Culham Laboratory, Abingdon Institute of Physics Publishing Bristol and Philadelphia CONTENTS Foreword

More information

Ordinary Differential Equations

Ordinary Differential Equations Chapter 13 Ordinary Differential Equations We motivated the problem of interpolation in Chapter 11 by transitioning from analzying to finding functions. That is, in problems like interpolation and regression,

More information

Fluid Dynamics. Massimo Ricotti. University of Maryland. Fluid Dynamics p.1/14

Fluid Dynamics. Massimo Ricotti. University of Maryland. Fluid Dynamics p.1/14 Fluid Dynamics p.1/14 Fluid Dynamics Massimo Ricotti ricotti@astro.umd.edu University of Maryland Fluid Dynamics p.2/14 The equations of fluid dynamics are coupled PDEs that form an IVP (hyperbolic). Use

More information

Gamma-rays from Earth-Size dark-matter halos

Gamma-rays from Earth-Size dark-matter halos Gamma-rays from Earth-Size dark-matter halos Tomoaki Ishiyama, Jun Makino, Toshikazu Ebisuzaki, and Veniamin Berezinsky Presentation by: JM Bottom line Microhalos (mass earth mass) do survive to the present

More information

Chapter 1. Introduction

Chapter 1. Introduction Chapter 1 Introduction Many astrophysical scenarios are modeled using the field equations of fluid dynamics. Fluids are generally challenging systems to describe analytically, as they form a nonlinear

More information

arxiv:astro-ph/ v1 19 Nov 1999

arxiv:astro-ph/ v1 19 Nov 1999 Where are the First Stars now? Simon D.M. White & Volker Springel Max-Planck-Institute for Astrophysics, Garching bei München, Germany arxiv:astro-ph/9911378v1 19 Nov 1999 Abstract. We use high-resolution

More information

A Trillion Particles: Studying Large-Scale Structure Formation on the BG/Q

A Trillion Particles: Studying Large-Scale Structure Formation on the BG/Q A Trillion Particles: Studying Large-Scale Structure Formation on the BG/Q David Daniel, Patricia Fasel, Hal Finkel, Nicholas Frontiere, Salman Habib, Katrin Heitmann, Zarija Lukic, Adrian Pope July 10,

More information

2.5.1 Static tides Tidal dissipation Dynamical tides Bibliographical notes Exercises 118

2.5.1 Static tides Tidal dissipation Dynamical tides Bibliographical notes Exercises 118 ii Contents Preface xiii 1 Foundations of Newtonian gravity 1 1.1 Newtonian gravity 2 1.2 Equations of Newtonian gravity 3 1.3 Newtonian field equation 7 1.4 Equations of hydrodynamics 9 1.4.1 Motion of

More information

The Milky Way and Near-Field Cosmology

The Milky Way and Near-Field Cosmology The Milky Way and Near-Field Cosmology Kathryn V Johnston (Columbia University) Collaborators (theorists): James S Bullock (Irvine), Andreea Font (Durham), Brant Robertson (Chicago), Sanjib Sharma (Columbia),

More information

The effect of star formation algorithms on the morphology of isolated galactic disc models

The effect of star formation algorithms on the morphology of isolated galactic disc models The effect of star formation algorithms on the morphology of isolated galactic disc models by Maan H Hani A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF BACHELOR OF SCIENCE

More information

1 Ordinary differential equations

1 Ordinary differential equations Numerical Analysis Seminar Frühjahrssemester 08 Lecturers: Prof. M. Torrilhon, Prof. D. Kressner The Störmer-Verlet method F. Crivelli (flcrivel@student.ethz.ch May 8, 2008 Introduction During this talk

More information

Feedback from massive stars in dwarf galaxy formation

Feedback from massive stars in dwarf galaxy formation Highlights of Spanish Astrophysics VII, Proceedings of the X Scientific Meeting of the Spanish Astronomical Society held on July 9-13, 2012, in Valencia, Spain. J. C. Guirado, L. M. Lara, V. Quilis, and

More information

Dark Energy and Massive Neutrino Universe Covariances

Dark Energy and Massive Neutrino Universe Covariances Dark Energy and Massive Neutrino Universe Covariances (DEMNUniCov) Carmelita Carbone Physics Dept, Milan University & INAF-Brera Collaborators: M. Calabrese, M. Zennaro, G. Fabbian, J. Bel November 30

More information

3.4 Numerical orbit integration

3.4 Numerical orbit integration 196 Chapter 3: The Orbits of Stars about the center of the galaxy. For small-amplitude librations, the libration frequency is p, consistent with our assumption that the oscillation frequency is of order

More information

Galaxy Bias from the Bispectrum: A New Method

Galaxy Bias from the Bispectrum: A New Method Galaxy Bias from the Bispectrum: A New Method Jennifer E. Pollack University of Bonn Thesis Supervisor: Cristiano Porciani Collaborator: Robert E. Smith Motivation 1) Main Objective of state-of-the-art

More information

telescopes resolve it into many faint (i.e. distant) stars What does it tell us?

telescopes resolve it into many faint (i.e. distant) stars What does it tell us? The Milky Way From a dark site the Milky Way can be seen as a broad band across the sky What is it? telescopes resolve it into many faint (i.e. distant) stars What does it tell us? that we live in a spiral

More information

Stellar-Dynamical Systems

Stellar-Dynamical Systems Chapter 8 Stellar-Dynamical Systems A wide range of self-gravitating systems may be idealized as configurations of point masses interacting through gravity. But in galaxies, the effects of interactions

More information

formation of the cosmic large-scale structure

formation of the cosmic large-scale structure formation of the cosmic large-scale structure Heraeus summer school on cosmology, Heidelberg 2013 Centre for Astronomy Fakultät für Physik und Astronomie, Universität Heidelberg August 23, 2013 outline

More information

Galaxy Formation: Overview

Galaxy Formation: Overview Galaxy Formation: Overview Houjun Mo March 30, 2004 The basic picture Formation of dark matter halos. Gas cooling in dark matter halos Star formation in cold gas Evolution of the stellar populaion Metal

More information

Two Phase Formation of Massive Galaxies

Two Phase Formation of Massive Galaxies Two Phase Formation of Massive Galaxies Focus: High Resolution Cosmological Zoom Simulation of Massive Galaxies ApJ.L.,658,710 (2007) ApJ.,697, 38 (2009) ApJ.L.,699,L178 (2009) ApJ.,725,2312 (2010) ApJ.,744,63(2012)

More information

Fitting the NGC 1560 rotation curve and other galaxies in the constant Lagrangian model for galactic dynamics.

Fitting the NGC 1560 rotation curve and other galaxies in the constant Lagrangian model for galactic dynamics. Fitting the NGC 1560 rotation curve and other galaxies in the constant Lagrangian model for galactic dynamics. 1, a) E.P.J. de Haas Nijmegen, The Netherlands (Dated: April 24, 2018) The velocity rotation

More information

Type Ia Supernova Observations. Supernova Results:The Context. Supernova Results. Physics 121 December 4, fainter. brighter

Type Ia Supernova Observations. Supernova Results:The Context. Supernova Results. Physics 121 December 4, fainter. brighter Physics 11 December 4, 009 Today Supernovae revisited Galaxy Rotation Curves Dark Matter & Dark Energy Scaling Factor a(t) Course Evaluations Type Ia Supernova Observations Distant supernovae are than

More information

Galaxy Hydrodynamic Simulations and Sunrise Visualizations

Galaxy Hydrodynamic Simulations and Sunrise Visualizations Galaxy Hydrodynamic Simulations and Sunrise Visualizations Joel Primack, UCSC Daniel Ceverino, HU Madrid Avishai Dekel, HU & UCSC Sandra Faber, UCSC Anatoly Klypin, NMSU Patrik Jonsson, Harvard CfA Chris

More information

Collisionless Boltzmann Eq (Vlasov eq)" S+G sec 3.4! Collisionless Boltzmann Eq S&G 3.4!

Collisionless Boltzmann Eq (Vlasov eq) S+G sec 3.4! Collisionless Boltzmann Eq S&G 3.4! Collisionless Boltzmann Eq (Vlasov eq)" S+G sec 3.4 When considering the structure of galaxies cannot follow each individual star (10 11 of them), Consider instead stellar density and velocity distributions.

More information

Veilleux! see MBW ! 23! 24!

Veilleux! see MBW ! 23! 24! Veilleux! see MBW 10.4.3! 23! 24! MBW pg 488-491! 25! But simple closed-box model works well for bulge of Milky Way! Outflow and/or accretion is needed to explain!!!metallicity distribution of stars in

More information

CS520: numerical ODEs (Ch.2)

CS520: numerical ODEs (Ch.2) .. CS520: numerical ODEs (Ch.2) Uri Ascher Department of Computer Science University of British Columbia ascher@cs.ubc.ca people.cs.ubc.ca/ ascher/520.html Uri Ascher (UBC) CPSC 520: ODEs (Ch. 2) Fall

More information

Understanding isolated and satellite galaxies through simulations

Understanding isolated and satellite galaxies through simulations Understanding isolated and satellite galaxies through simulations Kenza Arraki! Blue Waters Graduate Fellow! New Mexico State University! Anatoly Klypin! Daniel Ceverino! Sebastian Trujillo-Gomez! Joel

More information

2008 ASPEN WINTER WORKSHOP The first 2 billion years of Galaxy Formation. Gustavo Yepes Universidad Autónoma de Madrid

2008 ASPEN WINTER WORKSHOP The first 2 billion years of Galaxy Formation. Gustavo Yepes Universidad Autónoma de Madrid 2008 ASPEN WINTER WORKSHOP The first 2 billion years of Galaxy Formation Gustavo Yepes Universidad Autónoma de Madrid MNCP The MareNostrum Numerical Cosmology Project International collaboration to perform

More information

Smoothed Particle Hydrodynamics (SPH) a meshfree particle method for astrophysics *

Smoothed Particle Hydrodynamics (SPH) a meshfree particle method for astrophysics * Smoothed Particle Hydrodynamics (SPH) a meshfree particle method for astrophysics * Supplement 3: Example simulations using SPH and advanced graphics David Hobbs Lund Observatory ASTM17 Galaxy collision

More information

Galaxy Formation! Lecture Seven: Galaxy Formation! Cosmic History. Big Bang! time! present! ...fluctuations to galaxies!

Galaxy Formation! Lecture Seven: Galaxy Formation! Cosmic History. Big Bang! time! present! ...fluctuations to galaxies! Galaxy Formation Lecture Seven: Why is the universe populated by galaxies, rather than a uniform sea of stars? Galaxy Formation...fluctuations to galaxies Why are most stars in galaxies with luminosities

More information

ASOHF: a new adaptive spherical overdensity halo finder. S. Planelles and V. Quilis

ASOHF: a new adaptive spherical overdensity halo finder. S. Planelles and V. Quilis A&A 519, A94 (2010) DOI: 10.1051/0004-6361/201014214 c ESO 2010 Astronomy & Astrophysics ASOHF: a new adaptive spherical overdensity halo finder S. Planelles and V. Quilis Departament d Astronomia i Astrofísica,

More information

Elliptic Problems / Multigrid. PHY 604: Computational Methods for Physics and Astrophysics II

Elliptic Problems / Multigrid. PHY 604: Computational Methods for Physics and Astrophysics II Elliptic Problems / Multigrid Summary of Hyperbolic PDEs We looked at a simple linear and a nonlinear scalar hyperbolic PDE There is a speed associated with the change of the solution Explicit methods

More information

Modeling and Experimentation: Compound Pendulum

Modeling and Experimentation: Compound Pendulum Modeling and Experimentation: Compound Pendulum Prof. R.G. Longoria Department of Mechanical Engineering The University of Texas at Austin Fall 2014 Overview This lab focuses on developing a mathematical

More information

Number of Stars: 100 billion (10 11 ) Mass : 5 x Solar masses. Size of Disk: 100,000 Light Years (30 kpc)

Number of Stars: 100 billion (10 11 ) Mass : 5 x Solar masses. Size of Disk: 100,000 Light Years (30 kpc) THE MILKY WAY GALAXY Type: Spiral galaxy composed of a highly flattened disk and a central elliptical bulge. The disk is about 100,000 light years (30kpc) in diameter. The term spiral arises from the external

More information

The Los Cabos Lectures

The Los Cabos Lectures January 2009 The Los Cabos Lectures Dark Matter Halos: 3 Simon White Max Planck Institute for Astrophysics Shapes of halo equidensity surfaces Group Jing & Suto 2002 Galaxy δ 100 2500 6250 Shapes of halo

More information

The Current Status of Too Big To Fail problem! based on Warm Dark Matter cosmology

The Current Status of Too Big To Fail problem! based on Warm Dark Matter cosmology The Current Status of Too Big To Fail problem! based on Warm Dark Matter cosmology 172th Astronomical Seminar Dec.3 2013 Chiba Lab.M2 Yusuke Komuro Key Word s Too Big To Fail TBTF Cold Dark Matter CDM

More information

Active Galaxies and Galactic Structure Lecture 22 April 18th

Active Galaxies and Galactic Structure Lecture 22 April 18th Active Galaxies and Galactic Structure Lecture 22 April 18th FINAL Wednesday 5/9/2018 6-8 pm 100 questions, with ~20-30% based on material covered since test 3. Do not miss the final! Extra Credit: Thursday

More information

Computing the Universe, Dynamics IV: Liouville, Boltzmann, Jeans [DRAFT]

Computing the Universe, Dynamics IV: Liouville, Boltzmann, Jeans [DRAFT] Computing the Universe, Dynamics IV: Liouville, Boltzmann, Jeans [DRAFT] Martin Weinberg June 24, 1998 1 Liouville s Theorem Liouville s theorem is a fairly straightforward consequence of Hamiltonian dynamics.

More information

The Universe in a Supercomputer

The Universe in a Supercomputer The Universe in a Supercomputer Volker Springel Max-Planck-Institute for Astrophysics The universe in a supercomputer: From the first quasars to the large-scale galaxy distribution Volker Springel Simon

More information

The Evolution of SPH. J. J. Monaghan Monash University Australia

The Evolution of SPH. J. J. Monaghan Monash University Australia The Evolution of SPH J. J. Monaghan Monash University Australia planetary disks magnetic fields cosmology radiation star formation lava phase change dam break multiphase waves bodies in water fracture

More information

EXASCALE COMPUTING. Implementation of a 2D Electrostatic Particle in Cell algorithm in UniÞed Parallel C with dynamic load-balancing

EXASCALE COMPUTING. Implementation of a 2D Electrostatic Particle in Cell algorithm in UniÞed Parallel C with dynamic load-balancing ExaScience Lab Intel Labs Europe EXASCALE COMPUTING Implementation of a 2D Electrostatic Particle in Cell algorithm in UniÞed Parallel C with dynamic load-balancing B. Verleye P. Henry R. Wuyts G. Lapenta

More information

MOND and the Galaxies

MOND and the Galaxies MOND and the Galaxies Françoise Combes Observatoire de Paris With Olivier Tiret Angus, Famaey, Gentile, Wu, Zhao Wednesday 1st July 2009 MOND =MOdified Newtonian Dynamics Modification at weak acceleration

More information