Observing Winds in Collision with RXTE Weight-loss Secrets of the Biggest Stars Michael F. Corcoran...

Post on 30-Dec-2015

213 views 0 download

Tags:

Transcript of Observing Winds in Collision with RXTE Weight-loss Secrets of the Biggest Stars Michael F. Corcoran...

Observing Winds in Collision with RXTE

Weight-loss Secrets of the Biggest Stars

Michael F. Corcoran (CRESST/USRA/NASA-GSFC)

With: B. Ishibashi (Nagoya), T. Gull (NASA-GSFC), A. Damineli (IAGUSP), K. Hamaguchi (CRESST/UMBC/NASA-GSFC), A. F. J. Moffat (U. Montreal), E. R. Parkin (ANU), A. M. T Pollock (ESA), J. M. Pittard (U. Leeds), Atsuo Okazaki (Hokkai-Gakuen University), S. P. Owocki (U. Delaware), C. M. P. Russell (U. Delaware)

Overview

• Introduction: The complicated mass of massive stars• Colliding Wind Binaries: A shocking way to study mass loss• RXTE Breakthroughs

– WR 140 – Eta Carinae

• Conclusion

Introduction: M ~ M

Mass is the fundamental stellar parameter, but as it increases it becomes (observationally) less well constrained; especially near the upper mass limit (~150 solar masses, Figer 2005)

Moffat 1989

R145 in 30 Dor: M sin3i= 116+48 M

(i =38° => M=300+125 M!)Schnurr et al 2009

Problems In Measuring Mass Loss• Winds: Smooth wind vs. clumped? (factor of a few overestimate);

spherical or not?• LBV eruptions: timescales & rates?• SN explosion: core & remnant amounts?

X-rays as a Probe of Colliding Wind Systems

Solving the X-ray Lightcurve:

• wind terminal velocities (i.e. escape velocities) through temperatures; wind acceleration

• shock conditions & adiabatic/radiative cooling; flow dynamics from X-ray lines

• density information via absorbing column measurements

• D–1 variation – orbital elements and masses

Wind velocities of 1000’s of km/s X-ray emission

“Canonical” Examples of Colliding Wind Binaries

Two important “colliding wind binary shock laboratories”:

• WR 140, a carbon-rich Wolf-Rayet star + a “normal” O-type companion in a peculiar orbit: 8 year period, eccentricity = 0.88

• Eta Carinae: an LBV + unseen companion, in a peculiar orbit: 5.5 year period, eccentricity ~ 0.9

(Space Oddity: Both went through periastron passage within 5 days of each other in January 2009: next occurrence AD 12,620.808)

RXTE has been key in defining the X-ray emitting shock properties vs. time

Parameter WR 140 Eta Car

Sp. Type WC7+O4-5I LBV+O4?

Period (days) 2897 2024

e 0.89 ~0.9

a (AU) 15 15

Vwinds (km/s) 2860, 3100 500, 3000

Mass loss rate (M/yr) 3.3 x 10-5, 10-6 few x 10-4, 10-5

Masses (M) 16, 41 90, 30?

References Fahed et al. 2011; Russell et al. 2011; Monnier et al. 2011; Dougherty et al 2011

Hillier et al. 2001;Corcoran et al. 2010; Damineli et al. 2008; Parkin et al. 2011; Russell et al. 2011

WR 140 + Eta Car: Dynamical Sisters

X-ray derived

WR 140: “Canonical1” Massive Long Period Binary

• Coordinated variability at wavelengths from cm to 10-8 cm

• all driven by wind-wind collision between the two stars

P. Williams

A Brief History of Eta Car

1820 1850 1880 1910 1940 1970 2000

50 yrs

V-Band Lightcurve

V9

75

31

–1

An Eruptive Variable – what caused the Eruption?

Panchromatic (?) Variability: Eta Car

• Periodic spectral variations in He I 10830A (Damineli 1996) and X-rays

• strict period => gravitational dynamics

Radio variability (R. Duncan, S. White et al 1995)

X-rays (Corcoran et al. 1996)

1992 1993

Har

d

Sof

t

WR 140 + Eta Car: 3D Hydro models

Simulations by E. R. Parkin, J. M. Pittard

WR 140 Eta Carinae

Two fast winds at different Mdots fast+slow wind at different Mdots

Structured Mass Loss: Eta Car

Bow shock shapes primary wind

Variable (NH)

Coriolis distortions in wind near periastron passage effects photoionization of wind/nebula by the companion

Thermalization of KE at WWC produces X-rays sensitive to orbit

“Flashlight Effect”: variable illumination of the circumstellar material (See poster by T. Gull)

Density profile

wind structure in orbital plane

Okazaki et al. 2008

WR 140: Historical X-ray Variability

EXOSAT � ROSAT + ASCA

EXOSAT: 1983 – 1986(Williams et al. 1990)

ROSAT: 1990-1999 ASCA: 1993-2000

Courtesy A. Pollock

X-ray Spectrum: Eta Car

Hot Component 50 MK

Fe XXV

RXTE Breakthroughs:Snapshot observations hampers detailed understanding

RXTE Observations of WR 140 and Eta Car provided:• Strongest evidence that Eta Car is a binary system• Adequate time sampling to separate phase-locked and secular trends in both systems• Detailed lightcurves to compare to models• New physical insights: clumping, radiative braking• Realistic wind structure and the “flashlight” effect • Surprises

RXTE PCU observations:Eta Car: 1500 ksec Feb 9, 1996 - Dec 28, 2011WR 140: 500 ksec Mar 3, 2001 - Dec 23, 2011

RXTE: WR140

“Historical”

RXTE lightcurve in the 2-10 keV band

RXTE: Eta Car

2-10 keV

Comparison of Periastra: WR140

Eta Car

…what happened in 2009?

Eta Car Flare Behavior

... 1 AU Clumps?

Color Change: WR 140

Color Change Near Minimum: WR 140H

ardn

ess

Rat

io =

(7.5

-3 k

eV)/

(7.5

+3

keV

)

Color Change Near Minimum: Eta Car

3D Lightcurve Models: WR 140

Modelling by C. Russell

=270-

Okazaki et al. 2008

Parkin et al. 2008

3-D SPH + Isothermal point source emission

3-D Hydro + extended emission via 2-D hydro

Duration of X-ray minimum suggests collapse of shock(radiative braking? radiative inhibition?)

3D Lightcurve Models: Eta Car

Eta Car: Extended Emission?

Post-Periastron “Hot Bubble” due to fast companion wind colliding with swept-up wall of the LBV wind reduces minimum depth, duration... (Russell et al. 2011)

What Shortened the 2009 Minimum?

Minimum depends on: intrinsic 2-10 keV emission of shock near periastron (radiative instabilities important) amount of absorption around the “bubble”

A decrease in the LBV’s mass loss rate could make the shock less radiative (more stable) and less absorbed shorter X-ray Minimum?

Or something Else?

SummaryStudies of the colliding winds in systems like WR140 and Eta Car in X-rays (and UV, optical, IR & radio) provide unique information regarding mass loss in extremely massive stars, and information on behavior of astrophysical shocks

• Detailed modeling of the wind-shock boundary shape and flow

• Relative mass-loss pressure ratio (in-situ density diagnostics; clumping independent?)

• Dynamical studies (and periastron triggers?)

• Dust generation from shocked gas

• Detailed lab shock-physics experiments with varying conditions/parameters

Thanks to the RXTE Team for 16 Great Years!

esp. Evan Smith, heroic scheduler

Colliding Winds in WR 140Wind-Wind collisions in WR 140 allow us to probe time-variable shock physics under conditions of densities and temperatures which are difficult to reproduce on Earth

WR140: Our Shock Physics Laboratory

Courtesy P. Williams

Initial X-ray Results • Bright, variable X-ray source (unusual for a single massive star, even more unusual for a single WR star)

• Variable X-ray spectrum: Changes in emission measure of the hot gas, absorption to the hot gas

• Hard source: kT ~ 3-4 keV (also unusual for single massive star)

Need Detailed Monitoring: the Rossi X-ray Timing Explorer

Direct Imaging

VLBA imaging courtesy S Dougherty

What RXTE Sees:

Optical: Crowded Stellar Field

Dominated by WR140 CW emission above 2 keV

RASS 2x2 degree image of WR 140, 0.5-2 keV(with RXTE 5%, 30% & 80% contours)

X-ray: WR 140 dominant source

RXTE InstrumentationRXTE has 3 instruments:

• The Proportional Counter Array (PCA): • a set of five collimated Xenon-filled proportional counter units• 2-70 keV; 1600 cm2

• Most useful for WR 140

• The High Energy X-ray Timing Experiment (HEXTE)• two clusters of 4 NaI/CsI scintillator detectors• 15-250 keV; 1600 cm2

• The All-Sky Monitor (ASM)• 3 shadow cameras; 90 cm2

Outstanding Issues• Cause of the “Event” (eclipse; cooling/“discombobulation”; phase-

dependent mdot? Jet formation?)

• Stability of the bow shock

• Interactions near periastron

• Geometry of the inner/outer wind

• Density profile of the inner wind of Eta Car

• Radial velocities & mass ratio

Goals: 3-D Reconstruction of mass outflows, ejecta, photon fields Use the orbit of the companion as a probe of the fundamental parameters of Eta Car

RXTE ObservationsRXTE started observing 2-10 keV emission from Eta Car shortly after launch in Feb 1996

Continued monitoring with daily/weekly/monthly cadence since then

Daily monitoring near X-ray minima

The Campaign

P=2024 d; e~0.9; a~15 AU; i~50o

Ishibashi et al. 1998; Corcoran et al. 2001

Swift

Simulations by A. Okazaki

The Event

A (Sea) Change in Mass Loss RateSTIS, Gemini reveal weakening of emission lines in 2009 (Mehner et al. 2010)Large decrease in Mdot from LBV?

Decrease in Mdot may also explain the decrease in duration of X-ray minimum ( See C. Russell, Poster P5.20; Kashi & Soker 2009)

3-D Spectro-Models: Geometry of Mass Loss

Courtesy T. Madura, Phd thesis

Synthetic Slit Spectra from 3-D SPH Sims

ω = 270°

ω = 90°

ω = 180°

ω = 0°

STISMODEL

Courtesy Tom Madura

A Coupled Problem:• Evolution effects mass loss• Mass loss effects evolution

Vink et al. 2001(rotation? magnetic fields?)

LBV Nebulae

Post RSG?

Smith & Owocki 2006

Brandner et al.

D. F. Figer (UCLA) et al., NICMOS, HST, NASA,

Barlow et al. 1994

Weight Loss Secrets of the Stars• Mass is lost due to

– radiatively driven stellar winds – Transfer/Roche Lobe leaks– Eruptions– Explosions…

Mass Loss History

For the most massive star,mass loss makes it difficult to measure masses

Colliding Winds: A Simplified Approach

Early Work: Cherepashchuk (1976), Prilutskii & Usov (1976), Usov (1992), Steven, Blondin, Pollock (1992)

for Adiabatic shock, Lx 1/D(a, e)

Temps of 107K: Thermal X-ray Emission

cooling parameter

>1:adiabatic<1:radiative

Colliding Winds as a Means to Mdot

Stellar winds will hit something:• Interstellar medium

• another star

• wind from another star

Colliding wind Binaries provide:

– in-situ probe of mass loss

– “Clumping-free” estimator? (Pittard 2007)

– way to probe the stellar parameters

– “laboratory” physics of astrophysical hypersonic shocks

“Reality is Complicated”– Hideki Yukawa

Importance of radiative instabilities at wind-wind interface (thin shell, etc, Parkin et al. 2009, Pittard & Corcoran 2002)

Results

• Point source approximation near periastron can reproduce minimum

• More realistic distribution of hot gas along the shock boundary (“extended emission model”) does not reproduce the X-ray minimum as well…

• radiative cooling near periastron vs. adiabatic cooling near apastron?

• shift in X-ray temperature near periastron?

X-ray Value Non-X-ray ValueMass Loss Rate M/yr 1.2x10-6 & 3.8x10-5 same

Terminal Velocity (km/s) 3200 & 2860 same

Escape Velocity (km/s)§ 1280 & 1144 same

§assuming V∞=2.5 Vesc

Eta Carinae

Car is one of the most luminous (massive) stars in the Galaxy (most luminous, massive star within 3 kpc.)

Hubble Mosaic of the Carinae Nebula optical & X-ray

Results: Eta Car

• Point source approximation near periastron can reproduce depth of minimum

• More realistic distribution of hot gas along the shock boundary (“extended emission model”) does not reproduce the X-ray minimum as well…

• radiative cooling near periastron vs. adiabatic cooling near apastron?

• shift in X-ray temperature near periastron?

X-ray Value Non-X-ray Value

Mass Loss Rate 2.5x10-4 & 10-5 10-3 & ?

Terminal Velocity 500 & 3000 500 & ?

Escape Velocity 200 & 1200 200 & ?