Simulating the Bullet Cluster Is there Evidence for a LongRange NonGravitational Force

1
Simulating the Bullet Cluster Is there Evidence
for a Long-Range Non-Gravitational Force?

• Glennys R. Farrar
• Center for Cosmology and Particle Physics
• New York University
• Based on
• GRF Rachel Rosen astro-ph/0610298, Phys
  Rev Lett in press
• Volker Springel GRF astro-ph/0703232

2
Long Range 5th Force

• Natural in many theories, e.g., superstrings or
  matter in a shadow universe.
• 5th force is NOT MOND.
• 5th force cant couple to ordinary matter
  (Eotwash limits)
• 5th force may not simply decrease with distance,
  e.g., may be screened in dense concentrations of
  matter (chameleon models)
• Strategy measure the mass of a DM concentration
  and the acceleration it produces -- Is the
  acceleration consistent with gravity alone?
• If 5th force exists, DM coupling to ordinary
  matter must be very weak (radiative corrections
  Eotwash)

3
Possible Benefitsof an attractive 5th force

• Enhances structure formation for a given density
  contrast. This helps because
• Large Scale Structure simulations fit power
  spectrum with ?8 0.9, but ?8 0.75 from WMAP3
  (further exploration needed)
• 10 times more super-clusters observed in SDSS
  than in LCDM simulations (Einasto et al,
  astro-ph/06)
• Voids in LCDM simulations not empty enough (?) --
  5th force helps (GRF P.J.E.Peebles Ap. J.
  6401(2004). D. Croton GRF in prep)
• Too much late accretion in LCDM simulations
  (Nusser, Gubser, Peebles astro-ph/0412586)
• Baryons are less tightly bound, explaining
• Fewer DM substructures in galaxies than predicted
  by LCDM simulations

4
IE0657-56 bullet cluster, Z 0.3

• X-ray observations
• (Markevitch astro-ph/0511345)
• Mach 3.0 0.4 Velocity
• V rel_DM 4740- 550 630 km/s
• Tclus 14 keV

5
Statistical Approach(Hayashi and White MNRAS
370, L38 (2006))

• HW
• Parameterize distribution of vsubcl/v200 in
  Millenium Run.
• HW used early estimate
• vsubcl / v200 1.9
• probability ? 1/500 to find subcluster as fast as
  bullet.
• Latest values of vsubcl and v200
• vsubcl / v200 2.7
• probability ? 10-6 to find subcluster as fast as
  bullet

6
Expected Velocity from Gravitational Acceleration
(GRF Rachel Rosen, astro-ph/0610298, PRL in
press)

• Simplest assumptions gives Fiducial model
  vrel 2950 km/sec
• Consider alternate density profiles, initial
  conditions, mass accretion histories
  Conspiracy model (uncertainties conspire to
  maximize velocity)
• vrel 3435 km/sec
• Other uncertainties reduce the predicted velocity
• If measurements and their interpretations are
  correct, we seem to need an attractive 5th force
  on ranges Mpc with strength GN

7
Simulation of Bullet DM gas V. Springel
GRF astro-ph/0703232
8
Default Toy Model V. Springel GRF
astro-ph/0703232

• Spherical NFW Profiles, truncated at virial
  radius, (parameters from Clowe, private
  communication)
• Main Cluster M200 1.5 1015 M? c 2.0
  rvir 2300 kpc
• Subcluster M200 1.5 1014 M? c 7.2
  rvir 1070 kpc
• v 0 at 8 when touch, vsubcl 1870 km/s (C.M.)
• GADGET2 N-Body/SPH (Springel et al 05)
• 3-D (essential to get gas dynamics right)
• 4.4 106 particles (mgas 1.3 108 M?, mDM 6.2
  108 M?)
• Rerun best-fit case with 8x more particles -- no
  visible change.

9
(No Transcript)
10
X-ray Surface Brightness
11
Luminosity-weighted projected Temperature
12
Best Matching Snapshot with Basic Modelt 1340
Myr (180 Myr after core passage)
Data from Markevitch, our contours (private
communication)
13
t 1340 Myr (180 Myr after core passage) Mach
2.8
14
Positions of Landmarks as a function of time
Velocities of Landmarks as a function of time
15
Using different initial concentrations
16
Best-fit model compared to data
17
Off-Center Collision (12.5 kpc offset)
Plane of sky
45o to plane of sky
Perpendicular to plane of sky
18
Add fifth force (no other changes here)
? 1
? 0.3
Clear potential to constrain 5th force (V.
Springel GRF -- stay tuned)
19
(No Transcript)
20
Conclusions

• Bullet Cluster
• Relative DM velocities 2600 260 km/s .
• Agrees well with fiducial prediction of GRFRAR
  (2950 km/s)
• NO NEED for FIFTH FORCE
• Custom Simulations of Merging Clusters
• Must include DM and gas dynamics! (Naïve
  Markevitch et al estimate vrel Mach ? cs
  overestimates vrel, DM by 70)
• Fully realistic simulations essential (2D
  simulation of Milosavljevic et al astro-ph/070319
  16 instead.)
• Custom simulations will be a very powerful tool.
• Many observables to fit. Precision observations
  crucial!
• .

21
Keep Observing and Testing!

• kpc scales Milky Way dwarfs (oblate-prolate
  paradox!)
• Mpc scales cluster dynamics -- peculiar
  velocity weak lensing
• Cosmological scales -- LSS simulations