The Dynamical Evolution of Planetary Nebulae · 2015. 12. 2. · 70 yr 535 yr 1290 yr 1690 yr 1925...

The Dynamical Evolution of

Planetary Nebulae

D. Schonberner

Leibniz-Institut fur Astrophysik Potsdam (AIP)

In cooperation with M. Steffen, R. Jacob, & R. L. M. Corradi

• The physical system & its modelling

• Expansion velocities

– PNe of the Galactic disk

– Metal-poor PNe

• Soft X-ray emission

• Summary

11th Pac. Rim Conf. on Stellar Astrophys., Hong Kong, Dec. 14–17, 2015 c© D. Schonberner, AIP 1/22

The physical system (1)

Planetary Nebula:Relic of AGB wind, re-shaped by the steadily changingradiation field & wind of the post-AGB (= central) starwhile evolving across the HRD towards the WD stage

NGC 6826 AGB

♦ 103 yrs

3 M�

0.6 M�

RGB

Double-shellburning

Last TP cycle:strong mass lossterminatesevolution!


The physical system (2)

A typical round/elliptical PN –

Central star: Teff ' 100 000 K

Size of PN: Rpn ' 0.2 pc

⇒ kinematic PN age: ' 8 000 yr

Size of halo: Rhalo ' 0.6 pc

⇒ kinematic halo age: ' 40 000 yr

Halo –

Record of very final loss of stellar matter,enriched by freshly synthesized elementsdredged-up from the stellar interior bymixing processes

Planetary Nebula proper –

c© R. Corradi

2 nested shells: Bright RIM & attached fainter, much more massive SHELL,enclosing a “hole/cavity” containing the central star, and expanding

into the halo

⇒ A dynamical system with time-dependent boundary conditions !


The physical system, schematic (3)

Physical structure of a Planetary Nebula

fast

wind

hot

Bubble

106 .. 108 K

inner wind shock

Contact discontinuity

Rim 104 K

PN shaped by

colliding winds

fast central star wind

ÃÄ slow AGB wind

photoionization

by UV radiation

of hot central star

Slow

AGB

wind

Slow

AGB

wind

Shell 104 K outer shock

oute

r sh

ock

c© M. Steffen, AIP


Historical summary (1)

Early attempts – Matthews 1966, Ferch & Salpeter 1975

Expansion of a shell with initially ad hoc constant densityand velocity into vacuum, stellar radiation field constant

⇒• No sharp outer boundaries, shells disperse

• Backfill if no stellar wind (or radiation pressure on grains) for support

Authors claim fair agreement with observations, thoughno observables have been computed for comparisons !



A breakthrough – Kwok, Purton & Fitzgerald 1978

Birth of the interacting stellar winds (ISW) theory⇒

• Fast, tenuous central-star wind sweeps-up slow, dense relic of AGB-wind

• Natural consequence is one nebular shell expanding into and engulfingupstream AGB-wind material (= halo, if ionised)

Kwok 1982/1983:

Stellar evolution considered, i.e. time-dependent stellar winds& radiation fields ⇒ Shells expanding with '20 km s−1 possible

BUT:Analytical solutions, thus hydrodynamics due to ionisation

could not be considered properly

Chu et al. 1987, based on deep CCD imaging of PNe:

PNe generally with double-shell structure (triple with halo)



Modern simulations –

Schmidt-Voigt & Koppen 1989:• 1st radiation-hydrodynamics (RHD) study with time-dependent

ionisation of H & He, recombination discussed• Ad hoc initial shells around evolving central-star models (Schonberner),

simple wind model• Ionisation + stellar wind ⇒ double-shell structure, as observed!

Marten & Schonberner 1991: Like above but with better wind model

Mellema 1994/1995 . . . (& coauthors):• Better numerics, evolving & non-evolving stars, 1D & 2D (!)

Schonberner et al. 1997: & follow-up Papers I–VIII, 2004–2013• 1st consistent RHD simulation up the AGB & across the PN region

towards the white-dwarf stage, with a detailed mass-loss prescription• Time-dependent ionisation, recombination, heating & cooling for

Villaver et al. 2002:• Consistent RHD simulations based on mass-loss & AGB/post-AGB

models of Vassiliadis & Wood (1994)• No time-dependent ionisation, heating or cooling


Simulations (1a)

Consistent modelling the evolution of Star & wind envelope

5.5 5.0 4.5 4.0 3.5log Teff/K

2.0

2.5

3.0

3.5

4.0

log

L/L O •

ages in 103 yrs

025

10

15

0 5 10 15Post-AGB age [103 yr]

-2

0

2

4

log

L/L O •

L <912

Lbol

Lwind

10-1110-10

10-9

10-8

10-7

10-6

dM/d

t [M

O • /y

r] Reimers

Pauldrach et al.

0 5 10 15Post-AGB age [103 yr]

101

102

103

104

105

V [k

m/s

]

0.595 M�

1D-hydrodynamics of circumstellar envelope with time-dependent physics

• ionisation, recombination, heating, cooling for 9 elem., 12 ion. stages

• inner boundary condition (ri = 5×1014 cm):

– Star radiates as a black body with Teff(t)

– V∞(t), ρi(t) ∼ M(t)/r2i /V∞(t) from the wind model

Comput. of observables: line strengths & profiles, intensity distributions,X-ray emission (heat conduction incl.)


Simulations (1b)

5.5 5.0 4.5 4.0 3.5log (Teff / K)

2.5

3.0

3.5

4.0

log (

LC

S / L

O • )

0.696 MO • ,

0.625 MO • ,

0.605 MO • ,

0.595 MO • ,

0.585 MO • ,

0.836 MO • ,

0.565 MO • ,

70 yr

535 yr1290 yr1690 yr1925 yr

90 yr

2930 yr

time marks @ {0.3, 1, 3, 10, 30}x103 yr

The post-AGB tracksused for the simulations –

Post-AGB evolutionextremely sensitiveto remnant masses interms of luminosity &

time scale

Limiting luminosity fora '0.6 M� remnantwithin 30 000 years:'125 L�

RHD simulations for

• various post-AGB remnamts: 0.565. . . 0.696 M�• various initial envelope config., e.g. ρ ∝ r−α, α = 2, . . .3.5,

& by hydrodynamics of dusty AGB winds


Simulations (2a)

Match between models & real objects – No fits aimed at!

Young model:0.595 M�,age = 3364 yr,Teff = 39 300 K

rim

shell

Shell (or I-shell) bright in in Hα or [N II]

Rim (or W-shell) bright in [O III]

PN structure mainly determined by ionisation !

Winds interaction still not dominant !

NOTE: V[N II] > V[O III]

for young, optically thick PNe !


Simulations (2b)

Doppler-velocities V[N II] vs. V[O III] Jacob et al. 2013 (Paper VIII)

-2 0 2 4

0

1

2

3

4

V [N

II] /

V [O

III]

6 7 8

MV

0

1

2

3

4

0

1

2

3

4

0

1

2

3

4

0

1

2

3

4

0

1

2

3

4

0

1

2

3

4

0

1

2

3

4

0

1

2

3

4

0

1

2

3

4

0

1

2

3

4

0

1

2

3

4

0

1

2

3

4

0

1

2

3

4

0

1

2

3

4

0

1

2

3

4

0

1

2

3

4

0

1

2

3

4

0

1

2

3

4

0

1

2

3

4

0

1

2

3

4

0

1

2

3

4

0

1

2

3

4

0

1

2

3

4

0

1

2

3

4

0

1

2

3

4

0

1

2

3

4

0

1

2

3

4

0

1

2

3

4

0

1

2

3

4

0

1

2

3

4

0

1

2

3

4

0

1

2

3

4

0

1

2

3

4

0

1

2

3

4

0

1

2

3

4

0

1

2

3

4

0

1

2

3

4

0

1

2

3

4

0

1

2

3

4

0

1

2

3

4

0

1

2

3

4

0

1

2

3

4

0

1

2

3

4

0

1

2

3

4

0

1

2

3

4

0

1

2

3

4

0

1

2

3

4

0

1

2

3

4

0

1

2

3

4

0

1

2

3

4

0

1

2

3

4

0

1

2

3

4

0

1

2

3

4

0

1

2

3

4

0

1

2

3

4

0

1

2

3

4

0

1

2

3

4

0

1

2

3

4

0

1

2

3

4

-2 0 2 4

0

1

2

3

4

V [N

II] /

V [O

III]

6 7 8

MV

0.585 MO • , α=2.5

0.595 MO • , α=3

0.595 MO • , Hydro

0.605 MO • , Hydro

0.625 MO • , α=2

0.696 MO • , α=2

Cool luminous stars hot, faint starsOpt. thick PNe opt. thick PNe

Objects from thelocal sample of

Frew 2008

Open symbols:D ≤ 1 kpc

Filled symbols:D > 1 kpc


Simulations (3)

Match between models & real objects – No fits aimed at!

Middle-aged model:0.595 M�,age = 6100 yr,Teff = 80 200 K

density / velocity

0 2 4 6 8r [1017 cm]

0

500

1000

1500

2000

2500

n [c

m−3

]

0

20

40

60

80

V [k

m/s

]

surface brightness

0

0.5

1.0

I/Im

ax

[O III]

line profiles

0

0.5

1.0

I/Im

ax

[O II

I] 5

007

Å

0 2 4 6 8r [1017 cm]

0

0.5

1.0

I/Im

ax

[N II]

−40 −20 0 20 40V [km/s]

0

0.5

1.0

I/Im

ax

[N II

] 65

83 Å

rim

shell

halo

RPNC D

Mrim = 0.07 M�, ISW !Mshell = 0.40 M�, ionisation !

0.0

0.2

0.4

0.6

0.8

1.0

norm

alis

ed in

tens

ity

[NII] 6583Å

[OIII] 5007Å

−40 −20 0 20 40V [km/s]

0.0

0.2

0.4

0.6

0.8

1.0

norm

alis

ed in

tens

ity

−40 −20 0 20 40V [km/s]

Double-shell structure: rim & shell,despite smooth initial density profile!

Distinct velocities for rim & shell

Again NOT compatible with the Interacting Stellar Winds (ISW) theory!


Expansion velocities (1)

What is the true expansion velocity of a planetary nebula?

The position of the shell’s shock defines the nebular radius, RPN

⇒ This shock’s propagation speed, RPN, is the true PN expansion velocity

However,

the shock velocity cannot be measured spectroscopically!

Measurable “expansion” velocities are:

1. Peak separation of Doppler split emission lines (= Vrim)

2. Half width of emission lines of spatially unresolved objects

3. Post-shock velocity Vpost = RPN/F , with F > 1, Corradi et al. 2007

Velocities to 1. & 2. may depend (!) on ion used, correspondence to RPN ?A typical mean value used in statistical studies is 20–25 km s−1

Only velocity to 3. is physically sound, independent (!) of the ion used,correction F depends on shock property, = 1.25± 0.05 Jacob et al. 2013

Spectroscop. measured velocity smaller than the true velocity!


Expansion velocities (1a)

Kinematical age 6= post-AGB age, agePN ' tpost−AGB − ttrans

Predictions by our model PNe:

0.81.0

1.2

1.4

1.6

1.8

2.0

2.2

po

st-

AG

B a

ge

/

(R

OU

T /

VP

OS

T)

0.585 MO • , α=2.5

0.595 MO • , α=3

0.595 MO • , Hydro

0.605 MO • , Hydro

0.625 MO • , α=2

0.0

0.5

1.0

1.5

2.0

2.5

3.0

po

st-

AG

B a

ge

/

(R

CD /

VR

IM )

0.00.2

0.4

0.6

0.8

1.0

1.2

1.4

po

st-

AG

B a

ge

/

(R

OU

T /

VR

IM )

4.6 4.8 5.0 5.2log Teff [K]

0.6

0.7

0.8

0.9

1.0

1.1

1.2

po

st-

AG

B a

ge

/

(R

OU

T /

VA

V )

Shell radius/post-shock velocity:

Reasonable, except youngest PNe

Rim velocity/rim velocity:

Underestimated, except youngest PNe

Shell velocity/rim velocity:

Overestimated, except oldest PNe

Shell velocity/average velocity:Gesicki et al. 1998

Reasonable, velocity field must be known!



Post-shock velocity of the shell – after Corradi et al. 2007

Example NGC 6862, Inflection point of the (outer) shell line profile

good measure of Vpost:

0.0

0.2

0.4

0.6

0.8

1.0

norm

alis

ed in

tens

ity

[NII] 6583Å

[OIII] 5007Å

−40 −20 0 20 40V [km/s]

−1.0

−0.5

0.0

0.5

1.0

norm

alis

ed d

eriv

ativ

es

−40 −20 0 20 40V [km/s]

Model predictions:

4.6 4.8 5.0 5.2log Teff [K]

1.0

1.1

1.2

1.3

1.4

1.5

RO

UT /

VP

OS

T

. 0.585 MO • , α=2.5

0.595 MO • , α=3

0.595 MO • , Hydro

0.605 MO • , Hydro

0.625 MO • , α=2

Robust value of F = Rout/Vpost:

(optically thin models):

1.25±0.05



Results for 22 PNe – Tautenburg/NTT/NOT/CAT

Schonberner et al. 2014

4.5 4.6 4.7 4.8 4.9 5.0log Teff [K]

0

10

20

30

40

50

60

70

V [km

/s]

Vpost

Vrim

Vpost

Vrim

Vpost

Vrim

Vpost

Vrim

IC289

IC289

IC289

IC289

IC289

IC289

IC289

IC3568

IC3568

IC3568

IC3568

IC3568

IC3568

IC4593

IC4593

IC4593

IC4593

IC4593

IC4593

IC4593

IC4593

IC4593

IC4593

M1-4

6M

1-4

6

M2-2

M2-2

M2-2

M2-2

M2-2

M2-2

M2-2

M2-2

M2-2

NG

C2022

NG

C2022

NG

C2022

NG

C2022

NG

C2022

NG

C2022

NG

C2022

NG

C2022

NG

C2610

NG

C2610

NG

C2610

NG

C2610

NG

C3242

NG

C3242

NG

C3242

NG

C3242

NG

C3242

NG

C3242

NG

C3242

NG

C3242

NG

C3242

NG

C3242

NG

C3242

NG

C3242

NG

C3242

NG

C5882

NG

C5882

NG

C6543

NG

C6543

NG

C6543

NG

C6543

NG

C6629

NG

C6629

NG

C6629

NG

C6629

NG

C6826

NG

C6826

NG

C6826

NG

C6826

NG

C6826

NG

C6826

NG

C6826

NG

C6826

NG

C6884

NG

C6884

NG

C6884

NG

C6884

NG

C6891

NG

C6891

NG

C6891

NG

C6891

NG

C6891

NG

C6891

NG

C7009

NG

C7009

NG

C7009

NG

C7009

NG

C7354

NG

C7354

NG

C7354

NG

C7354

NG

C7354

NG

C7354

NG

C7354

NG

C7354

NG

C7354

NG

C7354

NG

C7354

NG

C7354

NG

C7354

NG

C7354

NG

C7354

NG

C7354

NG

C7354

NG

C7354

NG

C7354

NG

C7662

NG

C7662

NG

C7662

NG

C7662

NG

C7662

NG

C7662

Tc1

Tc1

Vy2-3

Vy2-3

Vy2-3

Vy2-3

IC418

IC418

IC2448

IC2448

NG

C1535

NG

C2610

NG

C7662

My60

My60

My60• Vpost > Vagb (' 10 km s−1),

rapid increase of Vpost

from ≈ Vagb to ' 40 km s−1

• For logTeff <∼ 4.7:

Vrim < Vagb,

then increase of Vrim to

≈ 25–30 km s−1

• Vpost/Vrim can be as large as 6 !

Theoretical expectations:

SHELL: RPN (or Vpost) depends on Te & upstream density gradiente.g. Chevalier 1997

RIM: Vrim is constraint by upstream pressure of ionised/heated shell) &hot bubble pressure (winds interaction) – e.g. Kwok et al. 1978

first deceleration/stalling due to high shell pressure,

then acceleration due to increasing wind power11th Pac. Rim Conf. on Stellar Astrophys., Hong Kong, Dec. 14–17, 2015 c© D. Schonberner, AIP 16/22


Kinematics of shells –

0

10

20

30

40

50

60

Vp

ost [

km

/s]

MCS = 0.595 MO • , nAGB ~ r - α, VAGB = 10 km/s

MCS = 0.595 MO • , nAGB ~ r - 3, VAGB = 20 km/s

MCS = 0.595 MO • , nAGB & VAGB : hydro sim.

α=2.00

α=2.25α=2.50

α=2.75

α=3.00

α=3.25

α=3.50

4.5 4.6 4.7 4.8 4.9 5.0 5.1log Teff [K]

0

10

20

30

40

50

60

Vrim [km

/s]

α=2.00

α=2.25

α=2.50α=2.75α=3.00

α=3.25

α=3.50

0

10

20

30

40

50

60

Vp

ost [

km

/s]




α=2.00

α=2.25α=2.50

α=2.75

α=3.00

α=3.25

α=3.50

4.5 4.6 4.7 4.8 4.9 5.0 5.1log Teff [K]

0

10

20

30

40

50

60

Vrim [km

/s]

α=2.00

α=2.25

α=2.50α=2.75α=3.00

α=3.25

α=3.50

0

10

20

30

40

50

60

Vpost [

km

/s]




α=2.00

α=2.25α=2.50

α=2.75

α=3.00

α=3.25

α=3.50

4.5 4.6 4.7 4.8 4.9 5.0 5.1log Teff [K]

0

10

20

30

40

50

60

Vrim [km

/s]

α=2.00

α=2.25

α=2.50α=2.75α=3.00

α=3.25

α=3.50

Vpost for a model grid withpower-law density profiles, ρ ∝ r−α,with α = 2 . . .3.5, Vagb = 10 km s−1,and for Mcspn = 0.595 M�:

PNe expand into (power-law)environments with initially α ≈ 2.8–3.4

Observed halo intensity distribution,SB ∝ r−γ ?

6 objects in common: 〈α〉 = 3.1± 0.1〈γ〉 = 5.2± 0.2 Schonberner et al. 2014

Since γ ' 2α− 1 ⇒ 〈α(γ)〉 ' 3.1

⇒ Observed post-shock velocities consistent with measured radialdensity profiles of halos

α > 2, ⇒ final mass loss on AGB “accelerated”! 〈RPN〉 ' 45 km s−1

(cf. Kwok, Volk & Hrivnak 2002 on density profiles of dusty envelopes around PPNe, α >∼ 2.5)


Expansion velocties (5)

Kinematics of rims –Sample objects with known Vcspn:

0 1000 2000 3000 4000VCSPN [km/s]

0

10

20

30

40

50

VR

IM [km

/s]

0.585 MO • , α=2.5

0.595 MO • , α=3

0.595 MO • , Hydro

0.605 MO • , Hydro

0.625 MO • , α=2

0

10

20

30

40

50

60

Vp

ost [

km

/s]




α=2.00

α=2.25α=2.50

α=2.75

α=3.00

α=3.25

α=3.50

4.5 4.6 4.7 4.8 4.9 5.0 5.1log Teff [K]

0

10

20

30

40

50

60

Vrim [km

/s]

α=2.00

α=2.25

α=2.50α=2.75α=3.00

α=3.25

α=3.50

〈Vrim〉 ' 19 km s−1

Vcspn ∝ R−0.5cspn , and for Lcspn = const ⇒ Teff ∝ R−0.5

cspn ⇒ Vcspn ∝ Teff

Increasing stellar wind power, Mcs×V 2cs/2,

⇒ increase of bubble pressure⇒ acceleration of rim, sweep-up of upstream (shell) matter

Theory of radiatively-driven stellar winds predicts correct rim velocities11th Pac. Rim Conf. on Stellar Astrophys., Hong Kong, Dec. 14–17, 2015 c© D. Schonberner, AIP 18/22


Metal-poor PNe – Final test against the “universality” of the ISW theory!

We expect: Schonberner et al. 2014

• RPN (& Vpost) up if metallicity down because of

higher electron temperatures

• Vrim down if metallicity down because of lower wind power

Observations with VLT/Argus IFU of 4 metal-poor PNe,BoBn 1, M2-29, PRMG 1, Ps 1/K648

0.0

0.2

0.4

0.6

0.8

1.0

norm

alis

ed in

tens

ity

BoBn1 [OIII] 5007Å

M2−29 [OIII] 5007Å

−50 0 50V [km/s]

0.0

0.2

0.4

0.6

0.8

1.0

norm

alis

ed in

tens

itiy

PRMG1 [OIII] 5007Å

−50 0 50V [km/s]

Ps1 [OIII] 5007Å

4.5 4.6 4.7 4.8 4.9 5.0 5.1log Teff [K]

0

10

20

30

40

50

60

70

V [

km

/s]

Vpost

Vrim

Vpost

Vrim

Vpost

Vrim

Vpost

Vrim

4.5 4.6 4.7 4.8 4.9 5.0 5.1log Teff [K]

0

10

20

30

40

50

60

70

V [

km

/s]

BoB

n1

M2−

29

PR

MG

1

Ps1


Soft X-ray emission from PNe (1)“Hot bubble” of shocked stellar wind – filling the central cavityc©Ruiz et al. 2013 0.3–2.0 keV

IC 418:

TX ' 3 MK

LX ' 9×1029 erg/s

NGC 2392:

TX = 2.1 MK

LX = 1.8×1031 erg/s

NGC 6826:

TX = 2.3 MK

LX = 2.0×1030 erg/s


Soft X-ray emission from PNe (2)

Heat conduction – Soker 1994, Steffen et al. 2008

Heat conduction reduces characteristic bubble temperature TX &increases mean bubble density (hence X-ray emission measure)

X−ray Temperature ( 6 − 41 Å) over wind velocity

0 1000 2000 3000 4000 5000 6000Vwind [km/s]

5.5

6.0

6.5

7.0

7.5

log

Tx

[K]

IC 418

NGC 2392

NGC 3242

NGC 6543

NGC 6826

NGC 7009

BD+30

NGC 40

NGC 5315

PN X−ray observationsM=0.696 Mo, HC2M=0.595 Mo, HC2M=0.565 Mo, HC2 T0

adiabaticpost-shock temp.

X−ray luminosity ( 6 − 41 Å) over total luminosity

5.5 5.0 4.5 4.0log(Teff) [K]

−9

−8

−7

−6

−5

−4lo

g(L X

/Lst

ar)

IC 418

NGC 2392

NGC 3242

NGC 6543

NGC 6826

NGC 7009

BD+30

NGC 40

NGC 5315

PN X−ray observationsM=0.696 Mo, HC2M=0.595 Mo, HC2M=0.565 Mo, HC2


SUMMARY

Structure, expansion, & X-ray emission of PNe in generalagreement with predictions of full RHD simulations:

• Double-shell structure a natural consequence of ionisationand winds interaction (no distinct mass-loss events necessary!)

• The expansion of a PN is driven by ionisation,not by the central-star wind alone (as the ISW theory assumes)

• Expansion speed of the nebula’s outer edge depends onelectron temperature & upstream density profile only,

with mean value ' 45 km s−1

• Expansion of the rim driven by central-star wind (ISW !),with mean value ' 20 km s−1

• Mass-loss rate increases along final AGB evolution α ' 3.1

• Metal-poor PNe expand faster (Te ↑), their bright rimslower (Lwind ↓) as compared to the metal-rich counterparts

• Winds interaction important in PNe with WR central stars

• Heat conduction responsible for soft X-ray emission


The Dynamical Evolution of Planetary Nebulae · 2015. 12. 2. · 70 yr 535 yr 1290 yr 1690 yr 1925...

Documents

Transcript of The Dynamical Evolution of Planetary Nebulae · 2015. 12. 2. · 70 yr 535 yr 1290 yr 1690 yr 1925...