root/trunk/tsuite/auto/dynamics_orion_flow.in

Revision 2539, 5.5 kB (checked in by peter, 4 weeks ago)

Merge the newsolvers branch.

This is the first merge from this branch implementing the new electron temperature and
density solvers, as well as a new method for tracking the convergence of the line optical
depths through different iterations.

In iter_track.h and iter_track.cpp the new classes iter_track and iter_track_basic are
implemented. The iter_track class iterates for the zero-point of a function based on the
van Wijngaarden-Dekker-Brent method, and is used for heavy-duty tracking of convergence
towards the root, complete with a full history of function evaluations. The
iter_track_basic class is a light-weight version that is based on a bisection search for
the zero-point and it does not keep track of the function evaluations.

The new electron density and temperature solvers are based on the iter_track class and as
a result now always bracket the solution before converging onto the solution. This should
improve the quality of the solution. The temperature solver now checks the accuracy of the
electron temperature in addition to the quality of the heating-cooling match. The old
solvers have been removed.

This merge also implements a new method based on the iter_track_basic class to track
convergence of the line optical depths through different iterations. This should improve
convergence when oscillations in the optical depth occur.

In cddefines.h new versions of the safe_div function are added that do not crash on 0/0,
but return a user-supplied argument instead.

The command PUNCH TEGRID has been removed since it is largely redundant. Use the command
PUNCH TEMPERATURE HISTORY instead. The commands SET TEMP/EDEN SOLVER have been removed
since they no longer serve any purpose (only one solver remains).

  • Property svn:eol-style set to native
Line 
1title Orion nebula blister with wind
2c
3c commands controlling continuum =========
4blackbody 40000
5phi(h) 13.0
6brems 6
7phi(h) 10
8c
9c commands for density & abundances =========
10hden 4
11abundances hii region no grains
12grains orion no qheat single
13c
14c commands controlling geometry  =========
15wind -7 km/s advection
16c this accounts for matter beyond end of the model
17double optical depths
18sphere
19// >>chng 06 jul 05, stop on AV rather than temperature */
20// >>chng 08 jul 01, from 15 to 5
21stop AV 5
22stop temperature linear 5 
23c
24c other commands for details     =========
25iterate 4
26magnetic field -4
27cosmic rays background
28c this has no effect on dynamics, but is to desaturate the lines as per flow
29turbulence 8 km/sec no pressure
30c try to speed up simulation without hurting dynamics
31init "fast.ini"
32c use Bakes & Tielens heating
33set PAH Bakes
34* init "c84.ini"
35* failures 5
36c
37c commands controlling output    =========
38c want to print line intensities are surface brightness, per arcsec^2
39print lines surface brightness arcsec
40print diffuse continuum
41print line faint -1.5
42punch overview "dynamics_orion_flow.ovr" no hash
43punch hydrogen 21 cm "dynamics_orion_flow.21cm" no hash
44punch dr "dynamics_orion_flow.dr" no hash
45punch continuum last "dynamics_orion_flow.con" units microns no hash
46punch asserts last "dynamics_orion_flow.asr" no hash
47punch pressure "dynamics_orion_flow.pre" no hash
48punch heating "dynamics_orion_flow.het" no hash
49punch cooling "dynamics_orion_flow.col" no hash
50punch molecules "dynamics_orion_flow.mol" no hash
51punch wind "dynamics_orion_flow.wnd" last
52c
53c commands giving the asserts    =========
54// >>chng 06 mar 11, from -11.241 to -11.175, chng small H from 4 to 5 lev
55// >>chng 07 jun 25, from -11.175 to -11.328, old h-like topoff added indirect to n=4, new only to n=6, huge effect on 4861.
56// >>chng 07 oct 26, from -11.328 to -11.350, very close to botch after recent changes
57assert line "totl" 4861 intensity -11.350
58//
59assert line "CaBo" 4861 1.0
60//
61// >>chng 05 oct 01, from 0.103 to 0.082, RP update He levels
62// >>chng 07 dec 29, from 0.082 to 0.857, ddrift up
63// >>chng 08 apr 30, from 0.0857 to 0.0908 H, He topoff change
64assert line "he 1" 5876 0.0908
65//
66assert line "+Col" 5876 0.127
67//
68assert line "He 1" 3889 0.0801
69assert line "+Col" 3889 0.105
70//
71assert line "He 1" 7065 0.0273
72assert line "+Col" 7065 0.0331
73//
74// >>chng 08 apr 30, from 0.913 to 0.842 H, He topoff change
75assert line "totl" 10830 0.842
76//
77// >>chng 06 mar 11, from 0.241 to 0.263, chng small H from 4 to 5 lev
78// >>chng 06 nov 29, from 0.263 to 0.221, Badnell DR turned on
79// >>chng 08 apr 30, from 0.221 to 0.250 H, He topoff change
80assert line "totl" 2326 0.250 error 0.1
81//
82// >>chng 05 oct 01, from 0.200 to 0.223, RP update He levels
83// >>chng 07 oct 22, from 0.223 to 0.246, resolve l-levels of h-like sequence
84assert line "totl" 1909 0.246 error 0.1
85//
86// >>chng 05 oct 01, from 0.505 to 0.480, RP update He levels
87// >>chng 05 dec 20, from 0.480 to 0.516, RP change default helium levels
88// >>chng 06 nov 29, from 0.516 to 0.449, Badnell DR turned on
89// >>chng 08 apr 30, from 0.449 to 0.497 H, He topoff change
90assert line "n  2" 6584 0.497 error 0.1
91//
92// >>chng 05 dec 20, from 0.621 to 0.655, RP change default helium levels
93// >>chng 06 nov 29, from 0.655 to 0.610, Badnell DR turned on
94// >>chng 08 apr 30, from 0.610 to 0.668 H, He topoff change
95assert line "totl" 3727 0.668 error 0.1
96//
97// >>chng 05 oct 01, from 3.707 to 3.871, RP update He levels
98// >>chng 07 dec 06, from 3.871 to 4.133, changes in h-like collisions
99assert line "o  3" 5007 4.133
100//
101// >>chng 06 jul 06, from 0.159 to 0.175, update Ne cs, CO chem now
102// includes dynamics but no cross talk between CO and H2 in H2 part
103// >>chng 07 dec 14, from 0.175 to 0.167 RP update is seq
104// >>chng 08 nov 04, from 0.167 to 0.1746, line RT bug fixes (r2429-35)
105assert line "ne 2" 12.81m 0.1746
106//
107// >>chng 05 dec 08, from 0.342 to 0.309 update to Elwert Solomon rate
108// >>chng 06 jul 06, from 0.309 to 0.484, sim goes much deeper, CO chem now
109// includes dynamics but no cross talk between CO and H2 in H2 part
110// >>chng 06 nov 29, from 0.484 to 0.561, Badnell DR turned on, no co in advec
111// >>chng 07 may 08, from 0.561 to 0.543, dyn logic changed to set stop
112// thickness after relaxing 2 iter, increased to 4 interation in this
113// sim to get more dynamics logic tested
114// >>chng 08 jul 01, from 0.561 to 0.0161, Av from 15 to 5
115assert HHeicf 0.0161 error 0.02
116//
117// some dynamic stuff
118assert velocity -0.130
119//
120// assert dynamics convergence error < xx
121//
122// >>chng 05 mar 28, from 80 to 50, impr mol converg
123// >>chng 08 jul 01, from 50 to 65 (58) Av from 15 to 5
124// >>chng 08 dec 08, from 65 to 57.33, upper limit -> equals (r2515)
125// >>chng 08 dec 13, from 57.33 to 100.8, merging newsolvers branch
126assert itrzn 100.8 error 0.02
127//
128// >>chng 06 jul 06, from 900 to 1000 sim goes much deeper, CO chem now
129// includes dynamics but no cross talk between CO and H2 in H2 part
130// >>chng 08 dec 08, from 1000 to 578, upper limit -> equals (r2515)
131assert nzone 578 error 0.01
132//
133c dynamics_orion_flow.in
134c class dynamics hii pdr
135c ========================================
136c
137
138This is a model similar in spirit to the blister geometry H+ region
139model computed by Baldwin et al. (1991), but with a D-critical flow.
140Many physical processes have been disabled to make this simulationn faster.
141Grain physics is not done so the gas temperature is incorrect.  The
142main purpose is to do a quick test of the dynamical flow with grain
143opacities included.  The slow directory contains a full simulation
144of a flow like Orion.
Note: See TracBrowser for help on using the browser.