Magnetically Confined Wind Shock model
From MassiveWiki
(diff) ←Older revision | Current revision | Newer revision→ (diff)
The Magnetically Confined Wind Shock (MCWS) model was orignially advanced by Babel & Montmerle (1997a), to explain the surprisingly-hard X-ray emission from the Ap/Bp star IQ Aur. The model envisages that a stellar wind upflow from opposite footpoints of closed magnetic loops collides to form strong, stationary shocks near the loop apex. The post-shock material, typically at temperatures of tens of millions kelvin, emits X-rays as it cools via radiative recombination.
A follow-up paper (Babel & Montmerle 1997b) suggested that a similar model could explain the 15.42-day periodic, hard X-ray emission of the O star θ1 Ori C. Although no magnetic field had ever been detected in the star at the time of writing, the authors conjectured that the presence of an oblique dipolar field could lead - via an MCWS - to the hard X-ray signature modulated on the &theta1's rotation period. Spectropolarimetric observations by Donati et al. (2002) were able to confirm the existence of the posited field.
Contents |
Geometry of the θ1 Ori C MCWS
Donati et al. (2002) concluded that the geometry of the magnetic field in θ1 Ori C was essentially dipolar, with an intensity 1,100 G, an obliquity β=42° to the rotation axis, and a vieing inclination i=45°.
This combination of parameters means that the star's magnetic field, over one rotation cycle, alternates between being seen pole-on and equator-on (see animation, right). Marc (Gagné et al 2005) took advantage of this fortuitous configuration when planning the Chandra observations of the star.
MHD simulations
To test the MCWS model for θ1 Ori C, Asif and Stan conducted 2-D (meridional plane) magnetohydrodynamic (MHD) simulations of wind outflow in the presence of a dipole magnetic field, with parameters appropriate to the star. The MHD models are primarily discussed in ud-Doula & Owocki (2002) and Owocki & ud-Doula (2004) (although the stuff on the non-isothermal models hasn't reached the journals yet; see Owocki et al. 2005). Generally speaking, the MHD simulations of θ1 Ori C predict X-ray properties that are in good agreement with those observed by Chandra (see Gagné et al. 2005). Selected animations based on the simulations are shown below.
| Density | Temperature | Speed | |
|---|---|---|---|
| Full Size |  |  |  |
| Zoomed |  |  |  |
ζ Cas 2D-MHD Simulations
ζ Cassiopiea is a B2 IV-V star with a dipole magnetic field of Bpole=335 G (see Neiner et al. 2003). Here are the movies of the first attempt of simulating the magnetic wind of ζ Cas. The simulation crashed at around time=155 ksec when a small vacuos region appeared near the equator. Typically, I add small amount of material near the stellar surface to keep the Alfven speed at bay, but this one occurred fairly high up.
Parameters used
I used R= 6*RSun, M=8 MSun and L=2.1e37 ergs/sec for this simulation. Terminal velocity for the non-magnetic version is roughly 1500 km/s (approx twice the escape speed, as should be the case for the CAK winds with finite disk correction taken into account). I used CAK δ=0.1 to keep the terminal velocity this low. Mass loss rate for the non-magnetic spherical wind is 3.3e-10 MSun/yr. The polar magnetic field is assumed to be B0=335 G.
Movies
This shows log10 of Temperature. Note, that the whole magnetosphere is pretty 'hot', and a cool disk is barely apparent. Here, we have to distinguish two different types of winds: 1) high density-- they behave much like isothermal winds since their column density is high and cooling is efficient. Theta 1 Ori C and Zeta Pup had these types of winds. 2) low density-- these behave more like adiabatic winds. Because of low densities the cooling seems fairly inefficient, and magnetosphere gets bloated with high T material. However, bear in mind that what you see below could be the result of initial condition. i will try to rerun this and 'flush' the initial hot gas and see what happens. As it stands now, it appears that Zeta Cas has the second type of wind.
This one is log10 of density.
This is the raw data in ASCII format that we could use with Steve's IDL code:
Input file for Zeus-3D, just in case..
