Light-cone cosmological simulations

[sketos]

Hello,

I would like to understand in detail what are the Light-Cone cosmological simulations and the difference between N-body cosmological simulations.
If anyone is familiar with the topic i am open to an explanation or perhaps he could recommend me a review or an article about the subject.

Thanks!

 

[Bobbywhy]

You may find answers in these two papers:

The MICE Grand Challenge Lightcone Simulation I: Dark matter clusteringAuthors: P. Fosalba, M. Crocce, E. Gaztanaga, F. J. Castander
(Submitted on 5 Dec 2013)
From Abstract: “We present a new N-body simulation from the MICE collaboration, the MICE Grand Challenge (MICE-GC), containing about 70 billion dark-matter particles in a (3 Gpc/h)^3 comoving volume. Given its large volume and fine spatial resolution, spanning over 5 orders of magnitude in dynamic range, it allows an accurate modeling of the growth of structure in the universe from the linear through the highly non-linear regime of gravitational clustering.”
http://arxiv.org/abs/1312.1707

arXiv:astro-ph/9701195v3 29 Jul 1997
Astrophys.J.Supplement 111, 73, 1997
ADAPTIVE REFINEMENT TREE – A NEW HIGH-RESOLUTION
N-BODY CODE FOR COSMOLOGICAL SIMULATIONS
Andrey V. Kravtsov, and Anatoly A. Klypin, Astronomy Department, New Mexico State University, Box 30001, Department 4500, Las Cruces, NM 88003-0001, USA and Alexei M. Khokhlov, Laboratory for Computational Physics and Fluid Dynamics, Code 6404, Naval Research Laboratory, Washington, DC 20375, USA

From Abstract: “We present a new high-resolution N-body algorithm for cosmological simulations. The algorithm employs a traditional particle-mesh technique on a cubic grid and successive multilevel relaxations on the finer meshes, introduced recursively in a fully adaptive manner in the regions where the density exceeds a predefined threshold.”
http://arxiv.org/pdf/astro-ph/9701195.pdf?origin=publication_detail

 

[Mordred]

N-body simulations are a mathematical algorithm to simulate a large number of particle interactions with a reduced number of pixels. Obviously simulating all the particles in the universe is too great a number. The algorithm depends on the process you wish to simulate, and uses a variety of matrix schemes, representative formulas of fundamental processes such as 2 body interactions due to gravity to reduce the number of calculations. Similar to a 3d graphics engine (though only in a VERY rough sense). Here is a site for some N-body algorithms.

http://www.cs.hut.fi/~ctl/NBody.pdf

and yes there is an available textbook on the subject.

https://www.amazon.com/dp/0521121531/?tag=pfamazon01-20

 

[Mordred]

there are some open source codes, you can look at.

http://www.artcompsci.org/
http://www.ast.cam.ac.uk/~sverre/web/pages/nbody.htm coincidentally this is the site of the author of that textbook I mentioned. I do own a copy, however its extremely tricky to understand lol
http://www.mpa-garching.mpg.de/gadget/
http://bima.astro.umd.edu/nemo/

edit: forgot to mention a FORTRAN compiler is often used, the book is designed for that compiler, however c and C++ can also be used.

 

[Majid.babaei]

MD simulation

Hi friends,

I am a newcomer to fortran. I am working on md simulation and I want to write a code for a rectangular nanopore.
I would appreciate if somebody can help me to write a fortran code for this shape.

 

[Mordred]

isn't a nanopore a small round hole? This forum has a programming sub-forum,

https://www.physicsforums.com/forumdisplay.php?f=165

I've seen FORTRAN questions posted there often, so you should be able to get help there. Its been years since I did FORTRAN. I code C++, among other languages,

 

[Majid.babaei]

Thank you very much Dear Mordred and Bobbywhy

 

[Mordred]

as a side note, the reason I looked into N-body was that I had hoped to see some of the more efficient ways to program. Unfortunately I noticed numerous instructions that could be performed faster, that are rarely used.

take for example. 4/2 most N-body codes simply divide. however Computers are binary machines. a bit shift left is the same as divide by 2 in binary. A bit shift right is the same as multiply by two.

In binary for the values 2,4,16,32,64,... the bit division often works faster than a divide or multiply function.

same with the use of jump instructions in a program. A jump uses 8 to 10 clock cycles. However a call to a subroutine only requires 4.

then I also rarely see pointers used. pointing to an address location on a stack uses less memory than a variable.

Just a side note.

 

[sketos]

Bobbywhy i had already checked THE MICE LIGHTCONE simulation but there is not much on the subject of light cones.

I am interested in a more detailed explanation of what light-cones are, how we can create halo "mock" catalogues, how light cones are created from N-body simulations (especially this one),... etc.

I am not interested in fortran codes either... but thanks anyway guys!

 

[Mordred]

ah , then we don't need N-body simulations to explain what light cones are. N-body simulations don't create light cones, they merely simulate one.

A light cone is the path that photons follow from a single,(source), and a single moment in time and traveling in all directions would take through spacetime. http://en.wikipedia.org/wiki/Light_cone

Its also often called the region of causal connection. as it describes all the possible locations
light can ever reach from a single source.

this Lineweaver and Davies has some good coverage

Expanding Confusion: Common Misconceptions of Cosmological
Horizons and the Superluminal Expansion of the Universe.

http://www.mso.anu.edu.au/~charley/papers/DavisLineweaver04.pdf

edit:
lol coincidentally my signature has a light cone calculator, written by a PF member.

http://www.einsteins-theory-of-relativity-4engineers.com/LightCone7/LightCone.html
the user manual and step by step on its uses is
http://cosmocalc.wikidot.com/start

if you look at the pinned thread
https://www.physicsforums.com/showthread.php?t=634757

you will see numerous examples and explanations

 

[sketos]

how exactly do the N-body simulations, simulate a light-cone?? ( i am more interested in this topic)

Is there a difference between light-cone simulations and N-body simulations??

 

[Mordred]

we seem to be traveling in a circle here.

Please read what a light cone means in Cosmology. A light cone is the path that photons follow from a single,(source), and a single moment in time and traveling in all directions would take through spacetime.

Then read the lineweaver and Davies article. What influences the paths that light can take? The answer has to do with how the universe expands or contracts.
What determines that expansion history? The relations between matter(gravity/positive pressure) and the cosmological constant. (negative pressure)

N-Body simulations is an algorithm that simulates the relations between matter (gravity as positive pressure) and the cosmological constant (negative pressure).

this relation also determines the Universes topology or geometry. More details here.
http://cosmology101.wikidot.com/universe-geometry

In a perfectly flat universe, the path of two parallel light beams will remain parallel, in a curved geometry, the light paths will either converge, or diverge.

If you want to understand how an N-body simulations work for cosmology applications you must first understand how Cosmology defines and describes the evolution of the Universe. See Einstein field equations or the FLRW metrics. Without that knowledge you cannot design a corresponding algorithm

Light cone simulations is simply the name given to identify the type of N-body simulation. N-body simulations can simulate any form of interactions if enough computer processing power is available. An N-body simulation of the interactions between two gravitational bodies is called P-P (point to point) simulation for example.