Plan
- I discovered a mistake in the AGB star model (in the code). It is relevant to the optical depth. I will fix it this week.
- Due to the mistake, the mass loss rate was not correct. I think the mass loss rate is crucial in the formation of circumbinary disk. Liljegren et al. (2016) did a detailed study on AGB wind recently. I will try their model. is somewhat small. If there could be a AGB wind, the formation of the circumbinary disk will be much more easier.
- I will work with Jonathan to add the self-consistent mass loss calculation in AstroBEAR.
- Run 4 low res simulations (3au, 4au, 6au, 10au). May not finish all of them since it takes 24hrs to finish a 3au simulation.
On the separation we may choose
There are two radius in this simulation with respect to the AGB star: dust sublimation radius, L1 point radius.
The dust sublimation radius is calculated by:
where
represent the photon absorption efficiency. is the AGB star's luminosity. is the Stefan-Boltzmann constant. is the dust sublimation temperature.A major question is, what is the value of
? I choose because some dust only absorb optical and ultraviolet wavelength photon e.g. MgSiO3. However, iron based silicates can absorb NIR photon efficiently. On the other hand, the self-extinction can redden the SED a lot. At last, a cool atmosphere should be assumed in the first place, e.g. 3000 K or so.Usually the dust formation radii is around
(I saw it somewhere). If I choose the photosphere of the AGB star to be around then the dust formation radii can be aroundLet me list my chose of parameters:
This will give
If we choose
then the radii of L1 points are:Therefore, case a is the only one that may be different from the others. But it is not necessary that case one must have different mass transfer behavior. Because the separation is still larger than the dust sublimation radii, in the current model, the radiative force on the fluid may blow away the disk or any structure around the secondary. If one want to see the disk around the secondary, the radiative ray tracing algorithm should be 3-D. Besides, the dust sublimation around the secondary should also be considered.
Difference between iron deficit and iron rich amorphous dust can be found here.
The AGB star is put into the simulation. In high resolution simulation, a circumbinary disk can form with
separationMeeting update
Paper editing. Just a couple of points:
Is it ever okay to use just the word 'mach'?
e.g. "The mach of the flows ranged between X and Y".
Also — for the methods section, why was II questioned? And also, is my wording about the box size confusing?
Meeting Update --06/23/16
- Wire Turbulence
- Variance:
- density PDF
From Gaussian fit, Federrath 15, Mach number can be calculated for different values of
, Using relation with 1/3≤b⇐1, according tob | 1 | 2/3 | 0.53 | 1/3 |
Mach | 0.80 | 0.53 | 1.0 | 1.60 |
- Other Variance and PDF plots, see Figures
- OH231 for Bruce
- Updated code with low density in nozzle area.
- test result which looks good to Bruce.
Update 6/23
- Ran planetary simulations with no rotation and an anisotropic temperature profile with various parameters. See my page for density movies. Changing lambda performs as expected - with lambda = 5, there is a marked increase in the strength of the wind, and with lambda = 15 there is very little to no wind. Changing the mass of the planet didn't result in quite a clear changes, but overall it doesn't appear to have affected the wind very much. Hotter planet appears to have a stronger outflow, while the cooler planet doesn't differ qualitatively from the original run. The last simulations I've performed so far were decreasing the ambient temperature, and this doesn't appear to affect the wind very much in terms of density. Momentum plots, however, appear to show outflows that are stronger by a factor of two.
- Also researched justification for using a fluid approximation for charge exchange. In Christie et al., they calculate the mean free path (sound speed over rate of reaction) for charge exchange and note that it is less than the planetary radius (in regions dominated by the planetary wind) to justify using the fluid approximation, with mixing and exchange primarily at turbulent boundaries. In Murray-Clay et al., they justify the fluid approximation in general by comparing the scale height H to the mean free path of a particle, with the fluid approximation holding for H > lambda_{mfp} at the sonic point (in other words, the exobase [where H = lambda_{mfp}] is above the sonic point). Thus, H > R_{p} as well, and therefore greater than mfp of charge exchange near the planet.
- Finally, working my way through Zel'dovich and Raizer for hydrodynamics. Beginning viscosity and heat conduction at the moment.
Current total quotas for afrank group
Total 27.5 TB with 97USD per TB per year or ~2667.5USD per year
BlueHive | BG/Q | |
Eddie | 4TB | 4TB |
Erica | 0TB | 12.5 TB |
Zhuo | 1TB | 0 TB |
afrank_lab | 6TB | 0 TB |
Total | 27.5 TB |
Mach Stems: Cooling Strength ==> Effective Gamma
Below is a new set of 2-D runs designed to explore how the strength of radiative cooling affects Mach stem formation and size. An approximation of the critical angle for Mach stem formation depends only on the adiabatic index gamma. We hypothesize that the cooling strength implies an "effective" gamma such that very weak cooling implies a gamma of 5/3 and very strong cooling implies a gamma of 1.
Furthermore, if a Mach stem forms, its size is limited to the smaller of the cooling length and the clump diameter. Thus, as cooling strength increases (decreasing cooling length), the size of the Mach stem should decrease.
Here are the important parameters:
vs = 50 km/s M = 5.2 T = 8322.56 K nclump/namb = 5000 tfinal = 100 yrs Rclump = 10 AU
Then, there are 5 more runs with M = 30 and T = 250.047 K.
dcool / Rclump | Ambient Density [cm^{-3}] | d / Rclump | Cooling Run | Effective Gamma Run |
---|---|---|---|---|
30 | 3.634 | 5 | 1.50 | |
10 | 11.16 | 4 | 1.30 | |
3 | 38.76 | 3 | 1.18 | |
1 | 123.30 | unstable | ⇐ 1.10 |
The effect of cooling is stronger than I had anticipated, and I'm not sure if we can use Pat's figure from his recent Mach stem paper to make quantitative comparisons. However, qualitatively, we are already proving our point, and more simulations with lowered gamma instead of cooling can lead to values for effective gamma.
Meeting Update --06/03/16
- Wire Turbulence
- OH231 module for Bruce
- The module is to simulate a soft-edged clump plow in an ambient gas left by a conical wind. The conical wind needs to be turned off after some time..
- Fixed the cooling issue for conical wind. ticket:445
- Need to set the nozzle area empty after the conical wind turned off — currently the code will stop pumping in material after the CW off but keeps updating the physical values and it will form a bubble. Clump could star from the nozzle area. Forcing the density & velocity in the nozzle to be low will cause the code choking. Haven't figured out a good way to do it.
round-nozzle of conical wind | |
current result with conical wind turned off | |
movie | bubble |
Update 6/3
- Read Tremblin and Chiang for computational charge exchange. The paper is a followup to earlier studies of charge exchange between planetary and stellar winds, which used Monte Carlo simulations of particles. Here they use the hydrodynamic equations (no magnetism, Coriolis and centrifugal forces, or tidal gravity); the isothermal planetary wind was initialized as 80% ionized, following Murray-Clay et al. The planetary wind incorporated photoionization/recombination and advection. To incorporate charge exchange, the hydrodynamic code was augmented with chemical reaction solvers, where beta is the reaction rate.
These equations take reverse exchange into account, so as not to overestimate neutral hydrogen (still slightly overestimated). They are very similar, but not identical to, the equations used by Christie et al.
The simulations appear to reproduce the observed absorption curves well, with asymmetry between the two sides of the Doppler shift.
- Used Jonathan's Matlab code to examine change of bow shock radius with magnetic fields. If sigma_{*} and sigma_{p} are equal, the bow shock radius is unchanged with or without magnetic field - ratio of radius to orbital separation, chi_{bow} = 0.240468. With sigma_{*} = 1, sigma_{p} = 0.1, chi_{bow} = 0.148204; sigma_{*} = 0.5, sigma_{p} = 0.1, chi_{bow} = 0.187300; sigma_{*} = 0.1, sigma_{p} = 0.5, chi_{bow} = 0.302483; sigma_{*} = 0.1, sigma_{p} = 1, chi_{bow} = 0.363674; and with sigma_{*} = 0.5 and sigma_{p} = 1, chi_{bow} = 0.297793 ≈ chi_{Coriolis}.
- Also attempted to recreate isotropic planetary wind with no rotation (Run5 from planet directory). Copied .data files, but clearly didn't turn out correctly. Need to figure out why.
Meeting update
- Melissa Morris reached out and told me she is submitting a proposal for the Emerging Worlds NASA project, and thinks I would be perfect for the project. If it goes through, she would like to offer me a postdoc at Suny Cortland. The project would be studying chondrule formation via large scale shocks in protoplanetary disks around young stars, using radiative transfer.
- Made a cv (find it here: http://www.pas.rochester.edu/~erica/cv.html)
- Poster got accepted (but not talk ) For upcoming "STAR FORMATION 2016" conference in Exeter
- Reorientation paper coming along nicely
- Spoke with Jonathan about beginning to build the outflow feedback routines. We said we can plan on that in the next week or 2.
- Submitted the shear flows paper to be published in Bo Reipurth's monthly Star Formation newsletter, cf.: http://www.ifa.hawaii.edu/~reipurth/newsletter.htm
- Thinking of reaching out to some professors at University of Missouri, St Louis, about meeting or giving a short talk while I am here for the week. It seems like 2 of them are sort of star-formationy. http://www.umsl.edu/~gibbe/Site/Research.html and http://www.umsl.edu/~wilkingb/