Originally posted to flash-users but maybe it's more pertainent here.
attached mail follows:
Dear All,
I have come upon a difficulty with using the multigrid gravity solver in
isolated boundary conditions with FLASH 2.1. I originally found this on while
testing the solver with a polytrope setup. I have demostrated it here with a
low resolution version of the dust collapse problem. The issue is that the
gravitational potential given by the multigrid solver looks nothing like that
given by the multipole solver. I have attempted to include all the relevent
information in this post. Advice on how to configure or modify the multigrid
solver to give the correct solution would be greatly appreciated.
Perhaps, as others have asked similar questions to this before, it might be of
use if someone could document how to properly make the dust collapse problem
run with the multigrid solver as an example for proper usage.
-Colin McNally
McMaster University Physics & Astronomy
Included below are:
Config file for the problem
flash.par file from the multigrid run
logfile from the multigrid run
stdout from the multigrid run
Attached are
ASCII data slices of the gravitational potential given by the multipole
and multigrid solvers at the end of the first step.
(columns: 1 = X, 2=Y, 3=Z, 4=junk, 5=gpot)
------------------------------------------------------------------------
Config file
------------------------------------------------------------------------
# Configuration file for spherical dust collapse problem
# (Colgate, S. A., & White, R. H. 1966, ApJ, 143, 626)
REQUIRES mesh/amr/paramesh2.0/second_order
REQUIRES driver/time_dep
REQUIRES hydro
REQUIRES gravity/poisson/multigrid
REQUIRES materials/eos/gamma
# R_init Initial radius of cloud
# rho_0 Initial density of cloud
# T_ambient Initial ambient temperature (everywhere)
# x/y/zctr Coordinates of the center of the cloud
# delta_ref Refine a block if the max density
contrast is
# greater than this
# delta_deref Derefine a block if the max density
contrast is
# less than this
PARAMETER T_ambient REAL 1.
PARAMETER R_init REAL 0.05
PARAMETER rho_0 REAL 1.
PARAMETER xctr REAL 0.5
PARAMETER yctr REAL 0.5
PARAMETER zctr REAL 0.5
PARAMETER delta_ref REAL 0.1
PARAMETER delta_deref REAL 0.01
# For testing purposes
VARIABLE grvx NORENORM # x-component of grav. acceleration
VARIABLE grvy NORENORM # y-component of grav. acceleration
VARIABLE grvz NORENORM # z-component of grav. acceleration
------------------------------------------------------------------------
logfile
------------------------------------------------------------------------
FLASH log file: 07-02-2002 14:17.59 Run number: 1
===============================================================================
Number of processors: 4
Build stamp: Tue Jul 2 13:01:26 2002
OSF1 idra1 V5.1 alpha
Version: FLASH 2.1.
Build directory: /astro/colinm/dust_coll
Setup syntax: ./setup.py dust_coll -auto -3d
-objdir=/astro/colinm/dust_coll
===============================================================================
Comment: Dust collapse problem
===============================================================================
Runtime parameters: read from file "flash.par"
dtmin = 1.0000000000000E-10
small = 1.0000000000000E-10
tmax = 6.3700000000000E-02
zfinal = -2.0000000000000E+00
tinitial = 0.0000000000000E+00
dtmax = 1.0000000000000E-02
smalle = 1.0000000000000E-10
dtini = 1.0000000000000E-03
smallx = 1.0000000000000E-10
zinitial = -1.0000000000000E+00
smallt = 1.0000000000000E-10
smallu = 1.0000000000000E-10
smlrho = 1.0000000000000E-10
smallp = 1.0000000000000E-10
temp_factor = 1.0000000000000E-01
tstep_change_factor = 2.0000000000000E+00
wall_clock_time_limi = 6.0480000000000E+05
eint_switch = 1.0000000000000E-04
cfl = 8.0000000000000E-01
rieman_tol = 1.0000000000000E-05
vgrid = 0.0000000000000E+00
cvisc = 0.0000000000000E+00
epsiln = 3.3000000000000E-01
omg2 = 1.0000000000000E+01
omg1 = 7.5000000000000E-01
trstrt = 3.5800000000000E-02
tplot = 1.0000000000000E+03
wall_clock_checkpoin = 4.3200000000000E+04
gamma = 1.6666667000000E+00
ymin = 0.0000000000000E+00
xmin = 0.0000000000000E+00
xmax = 1.4000000000000E+09
zmin = 0.0000000000000E+00
ymax = 1.4000000000000E+09
zmax = 1.4000000000000E+09
refine_filter = 1.0000000000000E-02
refine_cutoff = 8.0000000000000E-01
derefine_cutoff = 2.0000000000000E-01
mgrid_jacobi_weight = 6.6666666666670E-01
mgrid_max_iter_chang = 1.0000000000000E-03
mgrid_max_residual_n = 1.0000000000000E-06
mgrid_smooth_tol = 1.0000000000000E-06
zctr = 7.0000000000000E+08
delta_ref = 1.0000000000000E-01
delta_deref = 1.0000000000000E-02
xctr = 7.0000000000000E+08
rho_0 = 1.0000000000000E+09
t_ambient = 1.0000000000000E+00
yctr = 7.0000000000000E+08
r_init = 6.5000000000000E+08
nend = 5000
ichem = 0
ishkbn = 0
itemp_limit = 0
ipart = 0
igrav = 1
grav_boundary = -1
irenorm = 0
nriem = 100
igodu = 0
iplm = 0
rolling_checkpoint = 10000
cpnumber = 0
wr_integrals_freq = 1
memory_stat_freq = 100000
ptnumber = 0
nrstrt = 19
igeomz = 0
igeomx = 0
igeomy = 0
nblockz = 4
nblocky = 4
nblockx = 4
zlboundary = 0
xrboundary = 0
yrboundary = 0
zrboundary = 0
xlboundary = 0
ylboundary = 0
lrefine_min = 1
lrefine_max = 1
nrefs = 2
mgrid_solve_max_iter = 5000
mgrid_nsmooth = 4
mgrid_max_vcycles = 100
mpole_lmax = 0
log_file = "dustcoll.log"
run_comment = "Dust collapse problem"
run_number = "1"
grav_boundary_type = "isolated"
plot_var_5 = "none"
plot_var_7 = "none"
basenm = "dustcoll"
plot_var_8 = "none"
plot_var_6 = "none"
plot_var_4 = "none"
plot_var_3 = "temp"
plot_var_2 = "pres"
plot_var_1 = "grvx"
refine_var_3 = "none"
refine_var_2 = "none"
refine_var_1 = "dens"
xr_boundary_type = "outflow"
refine_var_4 = "none"
yr_boundary_type = "outflow"
bc_velocity_type = "outflow"
xl_boundary_type = "outflow"
zr_boundary_type = "outflow"
yl_boundary_type = "outflow"
zl_boundary_type = "outflow"
shock_burning = T
conserved_var = T
restart = F
print_tstep_loc = F
oldvisc = T
leveque = F
ppm_modifystates = F
corners = F
msgbuffer = T
monotone = T
mgrid_print_norm = F
===============================================================================
Known units of measurement:
cm = 1.0000000000000E+00 cm
s = 1.0000000000000E+00 s
K = 1.0000000000000E+00 K
g = 1.0000000000000E+00 g
esu = 1.0000000000000E+00 esu
m = 1.0000000000000E+02 cm
km = 1.0000000000000E+05 cm
pc = 3.0856775807000E+18 cm
kpc = 3.0856775807000E+21 cm
Mpc = 3.0856775807000E+24 cm
Gpc = 3.0856775807000E+27 cm
Rsun = 6.9600000000000E+10 cm
AU = 1.4959787066200E+13 cm
yr = 3.1556925200000E+07 s
Myr = 3.1556925200000E+13 s
Gyr = 3.1556925200000E+16 s
kg = 1.0000000000000E+03 g
Msun = 1.9889225000000E+33 g
amu = 1.6605402100000E-24 g
C = 2.9979245800000E+09 esu
Known physical constants:
Newton = 6.6725985000000E-08 cm^3 s^-2 g^-1
speed of light = 2.9979245800000E+10 cm s^-1
Planck = 6.6260755400000E-27 cm^2 s^-1 g
electron charge = 4.8032068150000E-10 esu
electron mass = 9.1093897540000E-28 g
proton mass = 1.6726231100000E-24 g
fine-structure = 7.2973530764000E-03
Avogadro = 6.0221367360000E+23
Boltzmann = 1.3806581200000E-16 cm^2 s^-2 g K^-1
ideal gas constant = 8.3145119843000E+07 cm^2 s^-2 g K^-1
Wien = 2.8977562400000E-01 cm K
Stefan-Boltzmann = 5.6705119000000E-05 s^-3 g K^-4
pi = 3.1415926535898E+00
e = 2.7182818284590E+00
Euler = 5.7721566490153E-01
===============================================================================
Multifluid database contents:
Full name Abbrev A Z N E Eb gamma
EOS?
-------------------------------------------------------------------------------
fluid 1 1 1.00 1.00 0.00 0.00 1.00 1.67
1.00
-------------------------------------------------------------------------------
===============================================================================
flash: initializing for dust collapse problem.
sound speed = 11771.7969108304
central dens = 1000000000.00000
central pres = 8.314511984300000E+016
external dens = 100000000.000000
external pres = 8.314511984299999E+015
[07-02-2002 14:19.33] *** checkpointed: dustcollhdf_chk_0000
[07-02-2002 14:19.33] *** wrote 64 blocks.
[07-02-2002 14:19.33] *** plot file: dustcollhdf_plt_cnt_0000
[07-02-2002 14:19.34] *** wrote to dustcollhdf_plt_cnt_0000
[07-02-2002 14:19.34] step 1 t= 0.000000E+00 dt= 1.000000E-03
mg_solve: using relaxation solver for coarse grid
init_mpole: using 113 radial samples
moment array: 452 items
[07-02-2002 14:20.21] *** plot file: dustcollhdf_plt_cnt_0001
[07-02-2002 14:20.21] *** wrote to dustcollhdf_plt_cnt_0001
------------------------------------------------------------
stdout
-------------------------------------------------------------
idra1> prun -n4 ./flash2
flash: initializing for dust collapse problem.
INITIAL TIMESTEP = 1.000000000000000E-003
*** Wrote output to dustcollhdf_chk_0000 ( 64 blocks ) ***
number of variables found for plotfile storage = 3
grvx pres temp
*** Wrote output to dustcollhdf_plt_cnt_0000 ***
n t dt | dt_hydro
1 2.0000E-03 2.0000E-03 | 1.368E-01
*** Wrote output to dustcollhdf_plt_cnt_0001 ***
2 6.0000E-03 4.0000E-03 | 4.568E-02
*** Wrote output to dustcollhdf_plt_cnt_0002 ***
3 1.4000E-02 8.0000E-03 | 1.977E-02
*** Wrote output to dustcollhdf_plt_cnt_0003 ***
4 3.0000E-02 9.6825E-03 | 9.682E-03
*** Wrote output to dustcollhdf_plt_cnt_0004 ***
5 4.9365E-02 6.0467E-03 | 6.047E-03
*** Wrote output to dustcollhdf_plt_cnt_0005 ***
6 6.1458E-02 4.7064E-03 | 4.706E-03
*** Wrote output to dustcollhdf_plt_cnt_0006 ***
7 7.0871E-02 2.0557E-03 | 2.056E-03
*** Wrote output to dustcollhdf_plt_cnt_0007 ***
*** Wrote output to dustcollhdf_chk_0001 ( 64 blocks ) ***
This archive was generated by hypermail 2b30 : Thu Jul 04 2002 - 12:04:41 CDT