How to kill a giant Dust-driven stellar winds on the AGB Susanne - PowerPoint PPT Presentation

How to kill a giant Dust-driven stellar winds on the AGB Susanne Höfner Department of Physics and Astronomy Uppsala University

Large-scale structure of the wind V-band image of IRC+10216 show- A fascinating picture of ing shell-like stellar mass loss and the cosmic matter cycle structures in the at work ... circumstellar en- velope (90''x 90'') Mauron & Huggins (2010) GALEX images of IRC+10216 (left: composite NUV+FUV, right: FUV) showing wind - ISM interaction (field of view 62'x 62') Sahai & Chronopoulos (2010)

Dust-driven winds: micro-physics Stellar photon hits dust grain dust grain absorbs & emits or scatters photon dust gains momentum dust collides with gas particle gas gains momentum dust loses momentum gas particle collides with other gas particle momentum redistribution in gas phase

Dust: condensation distance simple estimate for grain temperature: - radiative equilibrium - Planckian radiation field, geom. diluted - dust opacity approximated by power law condensation distance: T grain = T c condensation temperature (material property)

Dust: condensation distance Grain opacity full-drawn: lab data T * = 3000 K dashed: power law fit

Dust: condensation distance T * = 3000 K Example: am. carbon R c = 2-3 R *

Dust: condensation distance How to bridge the gap between the stellar photosphere and the condensation distance ? R c Early ideas: - acoustic wave pressure (e.g. Pijpers & Hearn 1989) - Alfvén waves (e.g. Hartmann & MacGregor 1980) ... inspired by hydrostatic atmosphere models (steep density profile) and stationary wind models (time-independent dynamical structures) But: AGB atmospheres are highly dynamical due to convection and pulsation ...

Dust: condensation distance How to bridge the gap between the stellar photosphere and the condensation distance ? R c stellar pulsation → propagating atmospheric shock waves → extended atmosphere, periodic “ballistic” motion of mass layers

Dust: condensation distance How to bridge the gap between the stellar photosphere and the condensation distance ? R c stellar pulsation → propagating atmospheric shock waves → extended atmosphere, periodic “ballistic” motion of mass layers

Effects of pulsation and shocks movement of mass shells in a detailed dynamical model atmosphere outer atmospheric layers: about 70 (50) percent of the time in the upper half (quarter) of the trajectory ⇒ levitation, extended cool atmosphere formation and propagation of shock waves pulsation

Extended dynamical atmospheres Spatial structure of a detailed dynamical model (incl. frequency- dependent RT and dust formation) top: density and gas temperature bottom: partial pressures of various molecules Note the effects of shock waves. Nowotny et al. (2010)

Dynamics of mass layers dust-driven outflow “pulsation-enhanced radial distance dust-driven wind” scenario mass layer accelerated by radiation pressure on dust condensation distance mass layer accelerated dust-free pulsating by passing atmosphere shock wave time

Dynamics of mass layers dust-driven outflow observed radial distance velocities indicate maximum mass layer levitation accelerated distances by radiation of a few pressure stellar radii on dust condensation distance mass layer E pot accelerated dust-free pulsating by passing atmosphere E kin shock wave time

Properties of wind-driving grains ● Form close to the star (at about 2-4 R * ), i.e. condensation distance < levitation distance ● Have high radiative cross sections (efficient Stellar spectrum radiative acceleration of individual grains) Planck function ● Consist of abundant elements (sufficient total momentum transfer to the gas)

Properties of wind-driving grains ● Form close to the star (at about 2-4 R * ), i.e. condensation distance < levitation distance ● Have high radiative cross sections (efficient Studying the range of viable radiative acceleration of individual grains) condensation temperatures and NIR optical grain ● Consist of abundant elements (sufficient total properties momentum transfer to the gas) using a dynamical model with detailed radiative transfer Bladh & Höfner (in prep.)

Properties of wind-driving grains ● Form close to the star (at about 2-4 R * ), i.e. condensation distance < levitation distance ● Have high radiative cross sections (efficient Studying the range of viable iron radiative acceleration of individual grains) MgFeSiO 4 condensation temperatures and NIR optical grain ● Consist of abundant elements (sufficient total d n properties i w o n momentum transfer to the gas) using a dynamical model with detailed radiative transfer Mg 2 SiO 4 Bladh & Höfner (in prep.)

Radiative acceleration: basics radiative / gravitational acceleration: κ H L * > 1 Γ = _______________________________ < 1 4 π c G M * critical value = 1 ⇒ critical flux mean opacity: κ crit = 4 π c G M * / L *

Properties of wind-driving grains ● Form close to the star (at about 2-4 R * ), i.e. condensation distance < levitation distance ● Have high radiative cross sections (efficient radiative acceleration of individual grains) ● Consist of abundant elements (sufficient total momentum transfer to the gas)

Chemistry: the simple picture available for M-type S-type C-type silicate grains available for carbon grains Abundance of C and O CO CO CO C/O ... changes during AGB evolution

Properties of wind-driving grains ● Form close to the star (at about 2-4 R * ), i.e. condensation distance < levitation distance ● Have high radiative cross sections (efficient radiative acceleration of individual grains) ● Consist of abundant elements (sufficient total momentum transfer to the gas) Promising candidate for C/O<1: Mg 2 SiO 4

Radiative pressure on Mg 2 SiO 4 ● Force: radiation pressure (dust) ● Conditions: standard scenario (e.g. Woitke 2006): set by shocks (pulsation) absorption by small - levitation grains (compared to λ ) - temporal variations Γ << 1: insufficient “pulsation-enhanced radiative pressure! log Γ dust-driven wind” stellar flux maximum ↔ κ H

Radiative pressure on Mg 2 SiO 4 ● Force: radiation pressure (dust) Γ > 1 ● Conditions: set by shocks (pulsation) alternative scenario: µ m-sized grains - levitation (scattering) - temporal variations “pulsation-enhanced log Γ dust-driven wind” stellar flux maximum ↔ κ H

Dust-driven winds for C/O < 1 opacity of Mg 2 SiO 4 as a function of viable scenario for wavelength and grain radius winds of pulsating Γ > 1 M-type AGB stars black contour marks where radiative alternative scenario: pressure exceeds µ m-sized grains gravity for a typical (scattering) detailed models with AGB star frequency-dependent radiative transfer and non-equilibrium log Γ dust condensation (Mg 2 SiO 4 ) + observations (Olofsson et al. 2002, stellar flux maximum Gonzalez Delgado 2003) o models (Höfner 2008) wind driven by Fe-free, micron-sized silicate grains Höfner 2008 comparison of mass loss rates and (A&A 491, L1) wind velocities with observations

Dust-driven winds for C/O < 1 } models Testing the scenario: NIR photometry (Bladh et al., in prep.)

Radiative pressure: M- vs. C-stars C/O < 1: Fe-free silicates scattering does the trick Γ > 1 for grain radius > 0.1 micron C/O > 1: amorphous carbon absorption dominates, even small grains can drive winds Γ > 1

Summary: dust-driven winds ● C/O > 1: winds driven by carbon grains, good agreement of detailed non-grey models with observations ● C/O < 1: non-grey effects on temperature → Fe-free silicates close to the star → too low radiative pressure for small grains (absorption) → winds possibly driven by Fe-free silicates > 0.1 µ m (scattering) ● More observational constraints for models!

How to kill a giant ● Collect suitable stones (lots of them ...) ● Accelerate them to high velocity ● Aim high and let them collide with the upper end of the giant ... Sounds familiar?