1 / 44

An excessive core collapse and its limitation in cosmological simulations

An excessive core collapse and its limitation in cosmological simulations. Weike Xiao xwk @mail.tsinghua.edu.cn. N-body Simulation Method. Have got huge success in cosmological studies For LCDM model: Nice agreement with observations: Power spectrum Correlation function

jolivas
Download Presentation

An excessive core collapse and its limitation in cosmological simulations

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. An excessive core collapse and its limitation in cosmological simulations Weike Xiao xwk@mail.tsinghua.edu.cn

  2. N-body Simulation Method • Have got huge success in cosmological studies • For LCDM model: Nice agreement with observations: Power spectrum Correlation function Mass function Halo model … …

  3. Some problem on small scale • DM halo density profile • NFW model: Navarro,Frenk & White (1996,97) the Halo(10^4 particle)density profile, can be discribed with an uniform func: • Higher resolution results (10^6): Moore 1999, Ghigna 2000,Klypin 2001 • More higher results even < 1 Power 2002, Stoehr 2003. • “the cusp problem”

  4. Gentile et al.对dwarf galaxy DDO 47

  5. W.J.G. de Blok et al. LSB galaxies (with two different methods)

  6. “the substructure problem” • Less than 50 dwarf galaxies in the local group was found. • Similar results in neighbor clusters • Simulation results predict 10-100times of substructures (10^10-11 M_sun)

  7. Explanations • LCDM model should be changed: SIDM WDM Modified Newtonian Dynamics(MOND) • Effects of baryons on small scales: feedback model, SBH…. • Delay of Observations: small dark halo gravitational lensing method…

  8. The particle number N in simulations • Started 1970th, 400 particles • Yasushi Suto found: • At Dec. of 2348, • Astro/ph-0311575

  9. The Particle number in simulations

  10. The limitation of N: • To keep the mean density, large particle mass: • It’s acceptable for large scale problems. each particle  one galaxy ( or one Halo)

  11. But for small scale topics: • We got the statistical property of the “Huge mass particle system” from simualtion • But in particle physics:  GeV candidates

  12. Dark Matter Particle Candidates

  13. A Physical Bias Exist • For a given box size L:

  14. The problems • Do they have the same dynamic and stitistical property? • How will it affect the simulation results? • At what time should it be considered? • any relationship with the “cusp problem” and the ”substructure problem”? • How to avoid the result of such a bias?

  15. From the viewpoint of each particle: • Different density fluctuation • The two-body relaxation

  16. Define the “mean free path” Ls: • For a virialized spherical system:

  17. The relaxation time scale: • For a given halo, Ls ~ N. • If the AE bias calls:

  18. equivalent scattering cross section • Follow the way of SIDM: • In simulation: • Near the value needed in SIDM • an excessive cross section exist!

  19. How will it affect the whole system? • Excessive scatter cross section can make E and L transition more effectively • Relaxation  Ei and Li be changed • Evaporation : Ei > 0  fly away • 2T+U=0 • Negative thermal capacity • Core collapse

  20. Core collapse process of one globular cluster (M. Freitag 2001)

  21. Time scale of core collapse • Based on Fokker-Planck eq., • Follow the way of Spitzer & Hart(1971)

  22. What can we see? • For the same halo, (given M,rh) different m: • For one simulation, (given m) different halo:

  23. If tcc < tu  introduce an excessive core collapse of the halo! • For one halo of the Milky Way, need more then 10^4 particles in sim. • The nowaday simulation, low z 10^6~10^8

  24. Evolution scenario at the halo center (Toshiyuki Fukushige, 2003)

  25. But in cosmological sim • CDM model, hierarchical clustering small halos, at high z, 10~10^2 particles For people use Kny when generate the initial conditions

  26. For small halos in early univ • An excessive core collapse was introduced! • They will have a too cuspy center • How will it affect the simulation result?

  27. Major merger M1<~10 M2 • Center part will soon merge • Cusp center can be kept, densiter profile keeps to be the deeper one only core+core = > core • Stelios Kazantzidis et al 2006 • Boylan-Kolchin et al.2004 • Astro-ph/0301271

  28. Minor merger M1 >> M2 • core center small halo, will soon melt together with large halo • Cusp center small halo, the cusp center won’t be demaged, be a new substructure

  29. How will the results be affected? • Init small halos will experience an excessive core collapse, got an unexpeted cusp center • In merging process, they can be kept: Major merger: the new halo will get a too cusp center .cusp problem Minor merger : the cusp can be kept as one more unexpected substructure of the large halo too many substruc: substructure problem

  30. How to avoid it? • One reliable simulation result should avoid such excessive core collapse • To avoid the excessive core collapse for the init small halos • Make sure they can include enough particles

  31. Limitation for generiting the initial condition • The mass within one shortest wavelenth: • The half mass radio: • The mass of each particle

  32. For nowaday simulations • Ng for FFT trasition: (Ng/2)^3=N K_max == K_Nyquist • There will must be many small halos with 10~10^3 particles experence the excessive core collapse  excessive cusp problem & substructure problem.

  33. How to avoid it? • Sensitive to : N, L, Kmax • 1. Increase N, but Kmax should be kept! • 2. When generating the initial condition, cut P(k) with K<Kcut • 3. Smaller L • resimilation: L’<L keeping : K<Kcut. L’<L  K’cut>Kcut. different with HDM model

  34. Expecting numerical testing What can we expect • When N increased, but still use(k_max=k_Nyquist), substructure will increase, cusp still exists. • To keep Kmax & L unchanged, N’=N*(2Pi)^3~250 N, we can avoid the excessive core collapse • N,Lunchanged, to use K<Kcutfor P(k)_ini

  35. Thank you!

  36. Halo Model

  37. Ant-Elephant bias Ant-Elephant (AE) Bias

More Related