Update 10/2

Line transfer

Stellar gravity and Coriolis test went well.

Parameters are the same for linked line transfer run, except:

Extent(37, -10, -10, 57, 10, 10)
Physical Extent(5.55x1011, -1.5x1011, -1.5x1011, 8.55x1011, 1.5x1011, 1.5x1011) cm

Side view

http://www.pas.rochester.edu/~adebrech/PlanetIonization/visit0005.png

​mesh

Top view

http://www.pas.rochester.edu/~adebrech/PlanetIonization/visit0007.png

​mesh

Contour key:

PurpleMach 1 surface
Black
GreenRib
RedRP
YellowRob
Brownish-red

Next

Need to test bigger box on BlueStreak (and perhaps get OpenMP working?). Since line transfer step happens at maximum resolution, the resolution of the box plays only a small role until a large portion becomes resolved. Pace is ~50 frames in 20 hours.

Fixed boundary

The equilibrium state is too close to the corrected Parker solution for any evolution to be seen. At 10-11 CT, the wind is the same as at 1 CT. Nothing more to be done here.

http://www.pas.rochester.edu/~adebrech/PlanetIonization/visit0009.png

1D comparison

Physical Parameters

4x10-13 g/cc
0.7 MJ
1x103 K
1.4 RJ
lRotateF
lStellarGravityF
lRampFluxF
Flux1x1013 phot/cm2/s

Simulation Parameters

Resolution2048
Zones/RP~450
Extent(-4.5,0)
Physical Extent(-4.5x1010, 0) cm
frames/CT1
TimeScale9.68 hours
lScale1x1010 cm

lScale = RP

Got a run that seems to have reached steady state. Stays well below supersonic, though. Still need to get lineouts to compare with Ruth's '09 paper, but the problem seems to be that as the wind launches, the surface moves out so that the wind is launching from what seems to be a very different planet.

http://www.pas.rochester.edu/~adebrech/PlanetIonization/1D_wind0150.png

Next

Comparison of the ionization diagnostic variables should show why the wind isn't staying sufficiently ionized for an insignificant optical depth. Results should lead toward either futzing with the code or altering the simulation parameters.

Comments

No comments.