Download SIMECA code Recent Results : Exploiting the VLTI

Recent results from the SIMECA code
and the VLTI observations
Anthony Meilland and Philippe Stee
Observatoire de la Côte d’Azur
I. Active Hot Stars
II. The SIMECA Code
III. Modelling The VLTI data
α Ara with MIDI
α Ara with AMBER
MWC297 with AMBER
HD50013 with AMBER
IV. Conclusion
I. Active Hot Stars
What Are Active Hot Stars ?
-Hot :
Spectral Type O B A  Teff > 8000 K
-Active :
Emission lines, IR excess ,Envelope or disc, gaz or\and dust,
Pulsations and other Variability
I. Active Hot Stars
A huge variety of phenomena !
-Fast Rotation : 60-80% of the critical velocity
(nearly 100% for Achernar)
-Stellar Wind : Radiatively driven, with high velocity
-Binarity : Interaction with a companion, mass transfer
-Pulsation : Non radial, many modes measured
-Magnetism : few hundred Gauss recently measured
I. Active Hot Stars
A huge variety of geometry and Kinematics !
-Envelope shape : Spherical, flattened, thick or thin disc, ring, jets
-Opacity : optically thin, thick or between
-Inhomogeneities : outbursts, blobs, arms, holes …
-Rotation law : Keplerian, angular momentum conservation …
-Radial velocity : None, expansion, accretion or both …
-Polar component of the velocity : Wind Compressed disc
I. Active Hot Stars
A huge variety of stars !
-Be : Main sequence stars, ionised hydrogen disc, stellar wind
-Herbig Ae\Be : Young stars, dust accretion disc, stellar wind
-B[e] : super-giant stars, stellar wind, dust disc
-Wolf Rayet : Strong mass loss, No photospheric line (optical thick
wind)
…And even some more violent objects like the monster Eta Car !!
II. The SIMECA code
SIMulation d’Etoiles Chaudes Active
Developed by Stee (1995)
Stee and Bittar (2001)
Stee and Meilland (2004)
A physical model :
Hydrodynamics (CAK wind model)
and radiative transfer (in Sobolev approximation)
in a rotating and expanding gaz envelope
Made for interpretation of observations :
Compute photometric (SED), spectroscopic (line profiles)
and interferometric (intensity mapsvisibility curves) observables
II. The SIMECA code
Star and envelope physical parameters :
Enter parameters :
Temperature
Stellar radius
Photospheric density
Rotational velocity
Inclination
Equatorial terminal velocity
Polar terminal velocity
Polar mass flux
Equatorial/polar mass flux ratio
H/H+He
“Free parameters”
m1: Exponent of the mass flux variation law
m2: Exponent of the velocity variation law
II. The SIMECA code
Starting with the basic hydrodynamic equations :
-Continuity equation
-Mass conservation equation
-No energy conservation (we don’t know the heating processes)
-Perfect gaz equation
Enter parameters :
Hydrodynamic :
ρ, Vr, VФ, T
With few hypothesis :
-axial symmetry (no azimuthal dependency)
-Stationarity
-Temperature depending only on r
-No polar component of the velocity in the envelope
We obtain in the envelope the distribution of : :
-Density
-Radial and azimuthal velocity
-Temperature
II. The SIMECA code
We start with at the LTE (Level 1 to 7 + continuum)
Using the Sobolev approximation (high velocity gradient) we obtain the
statistic equilibrium equation :
Enter parameters :
Hydrodynamic :
ρ, Vr, VФ, T
Statistic equilibrium:
n1,…,n7,ne at LTE
n1,…,n7,ne non-LTE
 i 1
 
ni   Aik  ki  Bic ic    nk Aki  ik  ne2Ci (Te )
 k 1
 k i 1
Aik, Bic et Ci : absorption, spontaneous emission and recombination
coefficient
βik : Escape probability (depend on the velocity gradient)
We calculate the level population from this equations and the previous
calculated values..
We iterate until the convergence of the values of the ni
II. The SIMECA code
Radiative transfer equation :
dI
  I   
dz
τ calculated by integration: dτ =-κ . dz (along the line of sight )
Enter parameters :
Hydrodynamic :
ρ, Vr, VФ, T
Statistic equilibrium:
n1,…,n7,ne at LTE
In the Continuum :
-Opacity of the envelope : Free-Free emission and electronic diffusion
-Emissivity of the envelope : Free-Free and Free-Bound
In the Lines :
- κ and ε expression for the selected transition
-Sobolev approximation
Intensity function of the spatial variables (perpendicularly to the line of sight)
for a transition (line) or function to the wavelength (continuum)
n1,…,n7,ne non-LTE
Transfer equation
in the contiuum
Transfer equation
In the lines
Transfer equation
In the continuum
II. The SIMECA code
In the line :
Enter parameters :
Calculation of the zone of projected iso-velocity  Doppler shift
-Spatial integration
Line profile
-Spectral integration (with a given spectral band)
Intensity maps in the line
In the continuum :
Hydrodynamic :
ρ, Vr, VФ, T
Statistic equilibrium:
n1,…,n7,ne at LTE
-Spatial integration
Spectral Energy Distribution
-Spectral integration
Intensity maps in the continuum
n1,…,n7,ne non-LTE
Transfer equation
in the contiuum
Transfer equation
In the lines
Transfer equation
In the continuum
Spectral Energy
Distribution
Line Profiles
Intensity maps
in the continuum
Intensity maps
In the lines
II. The SIMECA code
Enter parameters :
Hydrodynamic :
ρ, Vr, VФ, T
Statistic equilibrium:
n1,…,n7,ne at LTE
n1,…,n7,ne non-LTE
Transfer equation
in the contiuum
Transfer equation
In the lines
Transfer equation
In the continuum
Spectral Energy
Distribution
Line Profiles
Maps
in the continuum
Maps
In the lines
II. The SIMECA code
Actual and Future developments
Actual :
-More levels for the free-bound emission (for MIDI data)
-Ring and truncated disc model (evolution of the envelope)
-Interfacing with an accretion disc model (opacity of the dust)
-Asymmetry in the envelope
Future :
-Radiative transfer without Sobolev approximation (with Daniela Korkacova
code)
-Asymmetry in the envelope (Real 3D code without axisymmetry)
-Real dynamics
III. Modelling the VLTI data
The VLTI
4 Telescopes:
UT Fixed D=8,2m
4 Auxiliary Telescopes
AT moveables D=1,6m
- Baseline from few meters up to
200 m
- Good (u,v) plane coverage
(if you manage to have time and
telescopes!!)
Two Instruments
MIDI
AMBER
Mid-infrared ( 2 spectral bandwidths )
8-13 mm and 13-26 mm
2 telescopes
Visibility modulus and differential phase
Low spectral resolution ( R≈200)
Maximum spatial resolution of 12 mas @ 10 μm
Near infrared
1-2.5 mm
3 telescopes
Visibility modulus, differential phase, phase closure
Spectral resolution : R = 10000
Maximum spatial resolution of 2,5 mas @ 2 μm
Be studies:
- Observation of a Ara during SDT ( June 2003)
with HD 316285 (N=9.2, unresolved)
and d Cen (N=15, unresolved) in June 2004
with UT1-UT3 B =103m)
Be studies :
- Specially designed for stellar envelopes
- Kinematics studies
- Numerous faint objects
- Already guarantee time dedicated to Be stars
III. Modelling the VLTI data
α Ara with MIDI
Published In A&A in 2005
“First VLTI\MIDI observation of a Be star : α Arae”
Chesneau, Meilland, Rivinius, Stee et al.
B3Vne
mV=2.8
mK=3.8
Teff = 18000 K
R* = 4.8 Ro
M* = 9.6 Mo
Vsin i = 220km/s
Distance : 74 pc
Polarization : 172°
III. Modelling the VLTI data
α Ara with MIDI
VLTI data obtained in June 2003 and Spectrum from Brazil in august 2003
Visibilities as a function of l (8-13.5 mm) :
june, 16 : B=102 m , PA = 7°
june, 17 : B=79 m , PA = 55°
Spectral Energy Distribution :
Pa b line profile:
(8-13.5 mm)
(1,28 mm, transition 5-3)
III. Modelling the VLTI data
α Ara with MIDI
FEROS data obtained in may 1999 (Thomas Rivinius)
(transition 2-3)
(transition 2-4)
III. Modelling the VLTI data
α Ara with MIDI
Hα EW variations between 1978 and 2003
Physical Parameters determination
Ha line profile variation between 1978 & 1999
Timescale around 7 years
Circumstellar disk variations
between 1999 & 2003
Continuum supposed constant between
1999 & 2003
Two groups from non simultaneous
Spectroscopic and interferometric data
May 1999 & Summer 2003
Fit of the Ha & H lines (1999)
Few parameters variables:
density, wind velocity, envelope extension
Fit of the Paschen  (2003)
Agreement between observed and simulated visibilities?
α Ara’s SED
Results
α Ara’s distance
From Hipparcos :
74 pc
Parallax
From Cohen et al. 2001 :
122 pc
Fluxes & Color indices
Problem :
Mismatch between the two ditances determination.
Not possible for a B3Vne to have a radius less than 5 Ro
Maybe a (unseen) companion can produce a wrong Hipparcos parallax
or Error from Cohen et al. (2001) estimation
SIMECA :
Flux depends on the star radius and distance (Radius used : 4.8 Ro)
Distance obtained: 105 pc
74 pc
105 pc
Ha & H fits
Input Parameters
R*=5Ro
Teff = 18000K
phot=1.2 10-12
Vphot = 0.08
Vrot = 300 km/s
g = 0.86
fPole = 1.7 10-9
m1 = 0.3
m2 = 0.45
C1 = 30
Vpôle = 2000 km/s
Veq = 180 km/s
i = 45°
Nearly spherical
Inclination  Vrot
Strong polar wind
Low equatorial wind
Variations & parameters
Fit of the Paschen  line
and visibilities
Variations ?
density
Flattening
Envelope
Geometry
inhomogeneities
Winds
Troncated disk
 decreases by 25% (phot=0.97 10-12)
Rmax decreases by 18% (Rmax = 82.7R*)
Problems with the visibilities fits
Best Scenario:
Disk troncature by a
close companion
a Ara’s Binarity = z tau ?
Decrease of the envelope
extension
(4,5 times ≈ 22 R*)
+
Constant Flux
(density increase)
Period : 70 days
Orbital radius : 32 R*
Masse of the companion :
<2Mo
AMBER Observations : Pa & Brg
Baseline : 20 up to 100 m
Avoid P.A. ≈ 12°
III. Modelling the VLTI data
α Ara with AMBER
(Preliminary work)
Amber Br γ line profile
(not calibrated)
This emission line is
produced within the
Circumstellar envelope
III. Modelling the VLTI data
α Ara with AMBER
AMBER visibility
Theoretical visibilities
using the SIMECA code
III. Modelling the VLTI data
α Ara with AMBER
AMBER phase
Theoretical phase
using the SIMECA code
III. Modelling the VLTI data
MWC297 with AMBER
(work in progress)
MWC297 :
Herbig Be star
Strong hydrogen line in emission (Hα = 120 and Hβ=11)
Star + Accretion disc + Wind
Accretion disc + Star: Code by Fabien Malbet
Wind : Modified version of SIMECA
( with the opacity and emission from the accretion disc)
Data :
-Flux and visibility in the K band with Mid resolution
-Br γ line profile with quite high resolution (ISAAC)
-Hα and Hβ line profile from Drew’s 1999 article
-Magnitudes and ISO spectra
III. Modelling the VLTI data
MWC297 with AMBER
Problem : Where is the Wind ?
In the equatorial plan ? At the pole? Near the star ?
Quasi spherical wind, high velocity in the pole (600km\s), low velocity at the interface
with the disc (70km\s)
Br γ emission zone : starts at 8R ends at 50R 8 with a maximum around 27R
III. Modelling the VLTI data
MWC297 with AMBER
Problem : Differences between the 3 studied lines
Hα and Hβ profiles are very large (up to 600km\s)
Br γ profile is quite narrow (less than 200km\s)
They comes from slightly different regions
III. Modelling the VLTI data
MWC297 with AMBER
III. Modelling the VLTI data
HD50013 with AMBER (preliminary work)
Flux from SIMBAD :
Magnitudes U,B,V,R,I,J,H,K,L,M
+ UV Flux (0.2-0.4μm)
+ ISO Flux 10-30-60-100μm
+ Radio measurements
Star : B1.5IVe
= Planck function with :
Teff =22500K
Radius = 6 R
Distance = 242 parsecs
Classical Be star IR excess :
Beginning at 2μm
Spectral Energy Distribution (SED)
III. Modelling the VLTI data
HD50013 with AMBER
Fit of the SED with the SIMECA code
Total = Star + free-free + free-bound
Star = Planck
Free-Free and Free- Bound emission from the circumstellar envel
Inclination = 45°
Density at the photosphère = 10 -13 g.cm -3
Mass loss = 10-9 M\year
III. Modelling the VLTI data
HD50013 with AMBER
Line profiles
Hydrogen (+ He and Fe) lines in emission
=
Circumstellar matter
Ha
Ha
H
Asymmetric profiles
FeII l 5317 A
Wine bottle or double peaks
HeI l 5876 A
Long-term variations
H
Dachs et al. 1992, A&AS, 95, 437
Hg
From Lenorzer A. et al. 2002, A&A, 384, 473
Slettebak et al. 1992 ApJ Supp. 81, 335
III. Modelling the VLTI data
HD50013 with AMBER
Spectrum
Differential phase
Visibility Modulus
Closure phase
III. Modelling the VLTI data
HD50013 with AMBER
Visibility Modulus
HD 50013
Asymmetric Visibility modulus variation in the Br γ line
=
Red part of the emission in the line is more resolved than
in the continuum, but blue part is less resolved
=
Inhomogeneity in a rotating envelope
(cf “one-armed oscillations Berio et al. 1999) ?
=
Need of an asymmetric model
SIMECA simulation for a classical Be star
Visibility variation in the line (Hα)
for a classical Be star
for different rotational velocity law
(Constant, keplerian, angluar momentum conservation…)
III. Modelling the VLTI data
HD50013 with AMBER
Visibility Phase
HD 50013
SIMECA simulation for a classical Be star
No phase variation in the Br γ line
=
Circumstellar matter dominated by radial movement
=
No Rotation ? !
Phase variation in the line (Hα)
for different rotational velocity law
(Constant, keplerian, angluar momentum conservation…)
IV. Conclusion
-Need of lot of data to constrain models:
Simultaneous and time series with various timescales
each kind of data constrains some parameters
photometric  density, mass flux (and star parameters)
interferometric  geometry (+ kinematics if differential)
spectroscopic  kinematics ( + geometry )
- Need of two kind of model :
Physical but simple enough to be fast = SIMECA
+ non-axisymmetric + dynamics (for inhomogeneity)
More complex with less approximations (Sobolev)
to test the limits of the SIMECA code
 Slower  can’t be use easily to model observations