Geneviève Parmentier - Heidelberg University · 14 l o g (d N / d t [M y r-1]) log ( age [yr] ) Bound Cluster Formation History and Cluster Dissolution Time-Scale in the LMC Geneviève

Star Clusters as Crucial Tracers of Galaxy Evolution

and Probes into Star Formation

Splinter Meeting@AG2010, Bonn, 15 September 2010

Geneviève Geneviève Geneviève Geneviève ParmentierParmentierParmentierParmentier

ArgelanderArgelander --Institut fInstitut füürr AstronomieAstronomieRheinischeRheinische Fr iedr ichFr iedr ich--WilhelmsWilhelms UniversitUniversitäät Bonnt Bonn

Heidelberg Joint Astronomical Colloquium, 19 October 2010

22

��

Ø �� Ø ��

��

��

�� !��

Ø ��"#$�%$��&�

Ø �� '��

Star Clusters: Fast Facts

Geneviève Parmentier - Bonn Argelander-Institut für Astronomie 2 /

33

Cluster mass vs cluster age for a sample of ≅1,000 clusters in the Large Magellanic Cloud

A Vital Diagnostic Tool: the Age-Mass Diagram


44

�� ( ��)��*��" )*&� ��( +�� )� ��"+ �)&�

1

2

log−

−

∝≡

∝

mmd

dN

mdm

dN

log(dN/dlogm)

log(m)

t=20Myr t=2Gyr(100×××× more SCs if

CFR=cst)

Age-Mass Diagram and Size-of-Sample Effect


55

Age-Mass Diagram - Do Not Trust Your Eyes: 1/2


Size-of-Sample Effect


66

log (m [m0]) log (m [m0])

dN

/dlo

gm

Large Magellanic Cloud cluster sample:

Fading limit

77


dN

/dlo

gm

Large Magellanic Cloud cluster sample:

Fading limit

Gyrtdis 14 =

Completenesslimit: Mv=-4.7

88



Evolutionary fading:for a given mass,

older clusters are fainter

Size-of-Sample Effect


99



Higher SFR

Improved detection

limit


1010

Complete sample

At a given age:the lower the mass, the more incomplete

the sample

100% completenesslimit: Mv=-4.7


1111

��

Ø �� Ø ��

��

��

�� !��

Ø ��"#$�%$��&�

Ø �� '��

Star Clusters: Fast Facts


1212

log (m [M0]) log (m [M0])

dN

/dlo

gm

Modelling of Secular Evolution: Fig.17 in Parmentier & de Grijs (2008),based on the models of Baumgardt & Makino (2003) and Lamers et al (2005)

75Myr

300Myr

1Gyr

5Gyr

Secular Evol.: Carving a Turnover in the ICMF


Quicker Slower

Vertical scaling is arbitrary

1313


dN

/dlo

gm

Gyrtdis 14 =

Completenesslimit: Mv=-4.7

1414

log

(d

N/d

t[M

yr-1

])

log ( age [yr] )

Bound Cluster Formation History and Cluster Dissolution Time-Scale in the LMC


t4dis = 1Gyr

t4dis = 8Gyr

Secular evolution modelling Bound Cluster Formation History from the observed cluster age distribution

Parmentier & de Grijs 2008

Predicted Bound CFR

Observed age distribution

50Myr – 1.5GyrMass ratio : 0.50Number ratio: 0.03

1515

,�!��-� ��

"�� .�#�/��0� � �� &

1�!��-� ��

"�� .�!�2%�� +�)&

.�!�$�low- and high-mass clusters equally

contribute to the total mass (star clusters) 3 4 5 6 logm

m2× dN/dm

α−∝ mdm

dN : Function Mass Law-Power


Cluster Mass Functions: mass versus number

2 3 4 5 6 log(m)

Numbers: 90% 9% 0.9% 0.09%

Masses: 25% 25% 25% 25%

1616

log

(d

N/d

t[M

yr-1

])

log ( age [yr] )

Gas-Embedded Cluster Formation History: the next step ...


Violent relaxation

Secular evolution modelling:

(Bound) CFR

Violent relaxation modelling:

(Gas-embedded) CFR SFR

�

1717

Violent Relaxation (VR): Observable SignaturesAnd Prime Parameters


Intra-Cluster Gas-Expulsion and Violent Relaxation

Observable Imprints upon Star Cluster Systems :Cluster mass distribution, Cluster age distribution, Cluster radius distribution,Ratio of the total mass in clusters to

the total stellar mass in gas-embedded clusters

Geyer & Burkert (2001)

time

Effects of Gas expulsion - VIOLENT RELAXATIONCluster expansionCluster infant weight-loss and infant mortality

Prime parameters: (e.g. Baumgardt & Kroupa 2007)

- SFE in cluster-forming molecular cores

- Gas expulsion time-scale: ττττGExp /ττττcross

- Impact of external tidal field (environment)

See also Adams (2000), Vesperini et al (2009), …

1818

log m [[Mo]

Lo

g[d

N/d

log

m]

CMF3Myr ( = -2.0)

CMF50Myr ( = -2.0)

Violent Relaxation: Cluster Mass Functions


Infant mortality/

weight-loss

1

2

log−

−

∝≡

∝

mmd

dN

mdm

dN

)()( Exp Gas at VRof end clusterboundcluster mFm ×=

��

��

��

Time-Evolution of Cluster Mass Functions: What observers tell modellers …

1919

SFE and Cluster Mass Functions

��.��

��.��

coreboundcluster mSFESFEFm ××= )()( VRof end

Instantaneousgas removal,

weak tf impact

Fb

ou

nd

SFE


��

��

( )ε SFEboundF

2020

GExp/ cross and Cluster Mass Functions


log ( m [Mo] )

dN

/dlo

gm

Parmentier, Goodwin et al. (2008)

��

��

�

cross

GExpboundF

ττ

ε SFE ,

Constant core radius:More massive cores have- a deeper potential well- a slower gas-expulsion t-s- can survive despite a low

SFE of, say, 20%

��

��

�� crosscross��

��

2121

Tidal Field Impact and Cluster Mass Functions


��

��

� −

tidal

masshalf

cross

GExp

cross r

rt SFE ,,,

ττ

ετ

Half-mass radius rhalf-mass rcore

Circular velocity of iso-T potential VcGalactocentric distance Dgal

Embedded cluster mass mecl

rtidal

Weak t.f. impact

Strong t.f. impact

coreeclgalc

ecltidal mSFEmD

V

mGr

radiustidalLimiting

. with,

=��

�

�

��

�

�= 3/2

3/1

22

:

2222

Half-mass radius—to—tidal radius ratio


30.01

=��

��

�

pc

rcore

δcoremasshalfcore mrr ∝∝ −

2/1

101.0

1 ��

�

�

��

�

�=��

�

��

�

o

corecore

M

m

pc

r

3/1−− ∝ δcore

tidal

masshalf mr

r

3/13/1coreecltidal mmr ∝∝

3/1

104.0

1 ��

�

�

��

�

�=��

�

��

�

o

corecore

M

m

pc

r

4100

2323

Fbound and Tidal Field Impact


30.01

=��

��

�

pc

rcore

2/1

101.0

1 ��

�

�

��

�

�=��

�

��

�

o

corecore

M

m

pc

r

3/1

104.0

1 ��

�

�

��

�

�=��

�

��

�

o

corecore

M

m

pc

r

4100

Parmentier & Kroupa (in press)

2424

The mcore - rcore Diagram as a Diagnostic Tool


30.01

=��

��

�

pc

rcore

2/1

101.0

1 ��

�

�

��

�

�=��

�

��

�

o

corecore

M

m

pc

r

3/1

104.0

1 ��

�

�

��

�

�=��

�

��

�

o

corecore

M

m

pc

r

External tidal field strength: (Vc, Dgal) rh/rt in [log(r),log(m)]

rh/rt < 0.05: tf does not matter(does not necessarily imply weak tf !)

rh/rt > 0.15: cluster survivability

demands long ττττGExp /ττττcross and high SFE

tf impact impedes cl. survivability

As if no external tf

core

rcore

core

2525Geneviève Parmentier - Bonn Argelander-Institut für Astronomie 25 /

Cluster infant weight-loss is mass-independent, the shape of the cluster mass function does not evolve during VR

3/1

104.0

1 ��

�

�

��

�

�=��

�

��

�

o

corecore

M

m

pc

r

log(m)

log (dN/dlogm)

CMF

Ecl.MF

Fbound

Constant Volume Density Cores: mass-independent infant weight-loss

core

4100

Tidal Field Impact and Cluster Mass Functions:Probing the cluster-forming core mass-radius relation

2626

Tidal Field Impact and Cluster Mass Functions


log ( m [Mo] )

dN

/dlo

gm

SFE=0.40Dgal=5kpc

2/1

101.0

1: �

�

�

�

��

�

�=��

�

��

�Σ

o

corecorecore M

m

pc

r

Fbound

Constant Surface Density Cores: When more massive

means more vulnerable ...

2727

log

(d

N/d

t[M

yr-1

])

log ( age [yr] )

Galaxy Star Formation Histories: evena long journey starts with one single footstep ...


Secular evolution modelling:

(Bound) CFR

Violent relaxation modelling:

(Gas-embedded) CFR SFR

��

��

� −

tidal

masshalf

cross

GExpbound r

rF SFE ,,

ττ

ε

2828

Core mass-radius relations: observations


Density profiles: (rcore,mcore) @ 1E4 cm-3

Mueller+02

?

Core mass-radius diagram:imprint of

- measurement method,or of

- cluster-formation physics?

Limited baseline in cluster-forming core mass

Lack of high-mass data

core or core ?

when picking-up obs dataread (part of) the paper

2929

Core mass-radius relations: so what ... ?


- Red circles: C18O cores showing signs of SF activity

C18O

Car GMC (Yonekura+05)Orion B (Aoyama+01)

3030




C18O

- Most of them also host a clump at 1E5 cm-3 (H13CO+)

SF takes place in regions where the density is higher than a threshold, i.e. only in the densest regions of C18O cores


3131

* ** ** *




- Most of them also host a clump at 1E5 cm-3 (H13CO+)

SF takes place in regions where the density is higher than a threshold, i.e. only in the densest regions of C18O cores

C18O

Why ? Efficient decay of turbulence … (Klessen 2003)

Consequence: the local SFE must be measured over the cluster volume, not over the whole C18O core

H13CO+ observations suggest:Constant volume density cluster-forming cores, which lead to mass-independent infant weight-loss (t.f. impact),as suggested by observations


Parmentier & Kroupa (in press)

3232

Conclusions


log

(d

N/d

t[M

yr-1

])

log ( age [yr] )

Observed age distribution of

LMC star clusters

SFH

Properties of young star cluster systems sharp insights into the clustered mode of star formation

star formation conditions determinewhat mass fraction clusters lose as they age

information needed to reconstruct galaxy SFHtime-variations ? (e.g. metallicity)

Most exciting years arestill to come: HERSCHEL, ALMA, …

“ Even a long journey starts with a one sinle step”

Oriental saying

Documents

Geneviève Parmentier - Heidelberg University · 14 l o g (d N / d t [M y r-1]) log ( age [yr] ) Bound Cluster Formation History and Cluster Dissolution Time-Scale in the LMC Geneviève