circulation flows fabrizio brighenti (bologna) david buote (uc irvine) cooling flows with bubble...

Circulation Flows

Fabrizio Brighenti (Bologna)

David Buote (UC Irvine)

Cooling flows with bubble return!

Bill Mathews (UC Santa Cruz)

O’Sullivan et al. 2001

X-ray Luminosity of Elliptical Galaxies

Observed SNIa rate in E galaxies SNu = 0.16 per LB = 1010 per 100 yrs

Is almost certainly too high

(Cappellaro et al. 1999)

ROSAT

O’Sullivan et al. 2001

X-ray Luminosity of Elliptical Galaxies

Range of Lx/LB determined by extent of circumgalactic gas

Mathews & Brighenti 1998Lx/LB = (rex/re)0.6

O’Sullivan et al. 2001

Optically Dark Groups & Elliptical Galaxies

Filled circles: Optically dark galaxies/groups aka “Overluminous Elliptical Galaxies” (OLEG) “Fossil Groups”

Vikhlinin et al. 1999Ponman et al. 1994

NGC 5044

Optically Dark Groups with Mvir known from X-ray Observations

LB ~ Mvir may result from hierarchical assemblySeveral (all?) dark groups are baryonically “closed” like rich clusters:fb = Mbary/Mtot ~ 0.16 (WMAP)

NGC 6482

Caon et al. 2001

Warm gas in NGC 5044 -- Stellar Ejecta?

H + [NII] very disturbedwith crazy velocity fieldscale > SNIa remnantsejecta receives momentum

6 kpc

stellar isophotes

Extended Dusty Core in NGC 5044 -- Stellar Ejecta?

B-Iimage

12 x 12 kpcGoudfrooij 1991

Van Dokkum & Franx 1995Verdoes Kleijn et al. 1999

~50-60% of Normal Ellipticals and ~90% of Radio-Jet Ellipticals have Dusty Cores

HSTimages

Mathews & Brighenti 2003

Accelerated Cooling in Dusty Stellar Ejecta

Even dusty gas at 107 K cools very rapidly

Cooled gas still contains dust

Reliable minimum gas flow to black hole

Cooling at 1 kpc in NGC 4472

no dust

Buote, Lewis, Brighenti, Mathews 2003

XMM & Chandra Observations of NGC 5044

150 kpc 20 kpc

In pressure equilibrium |/|~|T/T|Scale of hot bubbles >> size of SNIa remnantsFilling factor f ~ 0.5 in r < 20kpc

XMM image is smooth beyond ~30 kpc

Buote, Lewis, Brighenti, Mathews 2003

Gas Temperature Profile in NGC 5044r (kpc) r (kpc)

Multiphase temperature Tc ~ T* ≤ T ≤ Th

but no gas with T ≤ Tc

(dM/dt)cool < 0.4 Msun /yr expected: ~5 Msun /yr

2T -- a better fit to data:1T fit to data:

Sun et al. 2003

Gas Temperature Profiles in Groups & ClustersGroups Clusters

Allen et al. 2001

dT/dr > 0 at small radii

Buote, Lewis, Brighenti, Mathews 2003

2T Multi-phas Emission in NGC 5044r (kpc) r (kpc)

Cool

Cool phase dominates inr ≤ 30 kpc

Filling factor of cool gas is f ~ 0.5 in r < 20 kpc

Global Properties of NGC 5044 E/group

r

(kpc)

Mgas

(Msun)

M*

(Msun)

Mtot

(Msun)

M Fe,gas

(Msun)

robs 327 13x1011 5.8x1011 2.0x1013 3.5x108

rvir 870 45x1011 7.4x1011 3.9x1013 9.0x108

Mbary/Mtot MFe/LB

5044 group 0.14 0.006

Rich clusters 0.13 - 0.17 0.015

ReE = 10 kpc LB,E = 4.5x1010 ∑LB,dwarfs = 10x1010

Buote, Brighenti & Mathews 2004

160

missing iron~WMAP baryons

Global Energetics of NGC 5044 E/groupEnergy in cavities Ecav = PfV = 1 x 1058 erg

Total SN energy Esn = 8 x 1060 erg

Gas binding energy Ebind = Eth = ∫thdV = 2 x 1061 erg

Black hole mass Mbh = 7.6x10-5 M*1.12 = 6 x108 Msun

Haring & Rix 2004

Black hole energy Ebh = .1 Mbh c2 = 1 x 1062 erg

to retain gas: the efficiency of black hole heating is < 0.02 power to maintain low density phase: PfV/tbuoy ~ 1043 erg/sec

~ Lx,bol = 6 x 1042 erg/sec

=> dMbh/dt = 4 x 10-3 Msun/yr

Circulation FlowsConstruct flows that simultaneously move in both radial directions with no net cooling or radial mass flow: cooling inflows balanced by bubble outflows

This is not convection as in stellar interiors, the S variations are more extreme Successful circulation flows: must look like cooling flows with dT/dr > 0 near center but with no cooling below ~Tvir/3 must reproduce observed iron abundance profiles to achieve this must recirculate both mass and thermal energy out from the center of the flows

Mathews et al. 2003

Simple Steady State Circulation Flows

Can low-density,heated bubbles carry enough gas upstream to balance the cooling inflow mass flux?

Mathews et al. 2003

Simple Steady State Circulation Flow in NGC 4472

Red: cooling inflowGreen: bubble outflow

Steady circulation flows with no netmass flux are possible

Bubbles do not heat inflowing gas very much the emission-weighted <T> profile is that of the cooling inflow; but bubbles may contribute to the X-ray spectrum

Bubbles with larger mass mb require more heating at rh, but if mb is too large, there is no volume left for cool phase, f --> 0

Small bubbles move too slowly and also consume all available volume near rh, f--> 0

h = 3rh = 5 kpc

Buote, Lewis, Brighenti & Mathews 2003

Radial Abundances in NGC 5044 A measure of integrated historical stellar enrichment

are central abundance dips real?

iron silicon

large metal enhancements in r < 100 kpc much larger than stellar Re

r (kpc) r (kpc)

Buote, Lewis, Brighenti & Mathews 2003

More XMM-Chandra Abundances in NGC 5044

<zSi/zFe>em = 0.83 solar => 70-80% of iron from SNIa within 100 kpc

silicon/iron magnesium sulfur oxygen

Why do O and Mg vary differently?

r (kpc) r (kpc)

Buote, Lewis, Brighenti & Mathews 2003

XMM Iron Abundances in NGC 5044

Total iron mass within r = 100 kpc is ~ 108 Msun from all historic SNIae?

Iron in r < 100 kpc Iron in 100 < r < 300 kpc

zFe ~ 0.1 - 0.2 solar (where is the missing iron?)

Buote, Brighenti & Mathews 2004

De Grandi et al. 2004

Central Iron Abundance Peaks are Commonin group NGC 507 in 12 CF and 10 non-CF clusters

Kim & Fabbiano 2004

De Grandi et al. 2004

Central Iron Abundance Peaks are Commonin group NGC 507 in 12 CF and 10 non-CF clusters

Kim & Fabbiano 2004

about 200 kpc

“excess” iron mass in CF clusters correlates with LB of central E galaxy

Excess iron mass ~ total iron from all SNIae in central E

Mathews, Brighenti & Buote 2004

Time-dependant Cooling flows for NGC 5044 with f( r)

assume fixed filling factor profile f(r ) for inflow

begin with standard cooling flows for NGC 5044 with three f(r)

no heating -- only radiative cooling range of flow: rh = 5 < r < re = 500 kpc calculate for 10 Gyrs result: (dM/dt)cool(rh) ~ 6 Msun/yr

cooling flow is very insensitive to filling factor profile so choose constant ...profile with f(rh) = 0.5 as observed

Mathews, Brighenti & Buote 2004

Time-dependant Circulation flows for NGC 5044

Now assume no gas flows in past rh = 5 kpc

The incoming mass flux at rh and stellar mass loss are heated by AGN and instantaneously circulated outward according to dp/dVOnly the inflowing cool phase is computedCirculated gas may be heated further if h > 0

Ignore bubble drag momentum exchange

Mathews, Brighenti & Buote 2004

Time-dependant Circulation flows for NGC 5044

Normalized recirculation probability:

parameters are (m, n, rp,kpc, <h>)

Mathews, Brighenti & Buote 2004

Time-dependant Circulation flows for NGC 5044

Spatially concentrated recirculation of gas without additional heating (h = 0):

Flow begins at t = 2.7 GyrsAfter only ~ 1 Gyr, gas near rp cools

Dotted lines areNGC 5044 observations

unacceptable

Mathews, Brighenti & Buote 2004

Time-dependant Circulation flows for NGC 5044

Spatially extended recirculation of gas without additional heating (h = 0):

Temperature too low Density too highzFe peak too low and broad

Flow began at t = 2.7 GyrsFlow is shown at t = 8 Gyrs when catastrophic cooling occurred

unacceptable

Mathews, Brighenti & Buote 2004

Time-dependant Circulation flows for NGC 5044Flows with additional heating continue until t = 13.7 Gyrs without cooling

Spatially extended recirculation of heated gas (h = 1.6 and 1.9)

Luminosity of AGN in NGC 5044 is ~hLh = 4 1042 erg/sTemperature peak is reproduced Density is acceptableNo gas flows into originNo gas coolsIron abundance peak from SNIae contains ~108 Msun of iron!

All major attributes of 5044 are reproduced

Does the SNIa iron cool or mix into hot gas?SNIa with 1051 ergs and MFe = 0.7 Msun explodes in elliptical ISM: ne = 0.01 T = 107

equilibrium temperature profile after 5 x 104 years:

Star-ISM boundary at 20 pc

Diffusion zone

Cooling of an Iron-rich Plasma

Cooling plus Diffusion

To avoid cooling, Fe must mix with ~5 Msun in the ISM

If magnetic fields reduce the diffusion rate, the SNIa iron may cool

zFe

T

tcool

Four mixing times tm

105, 107, 2x107, 2x108 yrs

Van Dokkum & Franx 1995

~60 % of Ellipticals have Dusty Cores

HSTimages

Brighenti & Mathews 2002

Heated Bubbles have Adiabatically Cooled Rims

Gas adjacent to expanding bubbles is cooled by adiabatic expansion

Brighenti & Mathews 2002

Heated Bubbles have Adiabatically Cooled RimsSelf-similar flow around spherical piston expanding into isothermal gas of decreasing density

Gas temperature just beyond piston is lowered

M = Mach No. at shock