[cig-commits] r4302 - geodyn/3D/MAG/trunk/doc
wei at geodynamics.org
wei at geodynamics.org
Tue Aug 15 22:52:47 PDT 2006
Author: wei
Date: 2006-08-15 22:52:47 -0700 (Tue, 15 Aug 2006)
New Revision: 4302
Modified:
geodyn/3D/MAG/trunk/doc/MAGDOC
geodyn/3D/MAG/trunk/doc/testlist
Log:
update testlist, fix typo in MAGDOC
Modified: geodyn/3D/MAG/trunk/doc/MAGDOC
===================================================================
--- geodyn/3D/MAG/trunk/doc/MAGDOC 2006-08-15 22:49:33 UTC (rev 4301)
+++ geodyn/3D/MAG/trunk/doc/MAGDOC 2006-08-16 05:52:47 UTC (rev 4302)
@@ -197,8 +197,8 @@
kbots: thermal boundary condition at ICB. As above.
ktopv: velocity condition at CMB. 1-free, 2-rigid.
kbotv: velocity condition at ICB. As above.
-ktopb: =1 for insulating inner core =2: ideally conducting inner core
-kbotb: =1 for insulating mantle =2: not implemented !
+ktopb: =1 for insulating mantle =2: ideally conducting inner core
+kbotb: =1 for insulating inner core =2: not implemented !
imagcon: <0 imposed poloidal field (l=1,m=0) at ICB
>=0 imposed toroidal field (l=2,m=0) at ICB
>=10 imposed toroidal field (l=2,m=0) at both CMB and ICB
Modified: geodyn/3D/MAG/trunk/doc/testlist
===================================================================
--- geodyn/3D/MAG/trunk/doc/testlist 2006-08-15 22:49:33 UTC (rev 4301)
+++ geodyn/3D/MAG/trunk/doc/testlist 2006-08-16 05:52:47 UTC (rev 4302)
@@ -19,7 +19,6 @@
*run restart a few times, each time rename d7.bench1 to "initfile"
compare output with published benchmark
-
2. produce unformated data for magsym.pro data process
set ngfom=2 for unformated file
@@ -28,66 +27,131 @@
3. dtmax, the maximum(initial) time step
set dtmax= 0.25e-4 = 0.000025 in bench1
dtmax < 0.25*Ekman (Ekman=1.0e-3)
- is bench1 unstable? when dtmax=0.25e-3
-
+ dtmax=0.25e-3
+
4. courfac determines dt limit, set to 3, 4 in bench1 with dtmax=1.5e-4
larger courfac gives smaller dt
reduce courfac would reduce run time, program running more efficent. It is a balanced act
-5. radratio ri/ro =0.35 (bench1, bench0)
- set to: =0.8
- =0.1
+5. radratio ri/ro =0.35 (bench1, bench0)
+ set to:
+ =0.8 MAG runs ok
+ =0.99 MAG runs ok
+ =0.1 MAG runs only executed 23 statement
+ =0.05 MAG runs only executed 23 statement
+ =0.005 MAG runs only executed 23 statement
+ =0. MAG hangs, dose not kick out the program. It has to be terminated manually
+
-
6. Rayleigh number: Ra = 1.0e5=100000 in bench1
- set to 3.0e4
- 6.0e4
- 3.0e5
-
-7. Prandtl number Pr stable, Pr=1.0 in bench1
+ set to
+ 3.0e4
+ 6.0e4
+ 3.0e5
+ 6.0e5
+
+7. Prandtl number Pr stable, Pr=1.0 in bench1. (dynamo models)
+ set Pr = 0.1 (Iron Core)
+
8. Magnetic Prandtl number Prmag = 0.5 -10, Prmag=5.0 in bench1
set Prmag = 10 for test
+ MAG runs ok
-9. bpeak = -5 in bench1, poloidal dipole
-
+9. bpeak is the peak value of magnetic field imposed by boundary condition at ICB.
+ bpeak = -5 in bench1, with imagcon =0, initial magnetic field is poloidal dipole
+ set bpeak > 0 for test, imagcon=0, initial magnetic field is toroidal
+ set:
+ bpeak = 5.0
+ bpeak = 0.5
+
10. boundary conditions:
- ktops = 1 fixed temp at CMB
- ktops = 2 fixed radial heatflow @CMB
- kbots = 1 fixed temp at ICB
- kbots = 2 fixed radial heatflow at ICB
- ktopv = 1 velocity condition is free at CMB
- kbotv = 2 velocity condition is rigid at CMB
- ktopb = 1 for insulation inner core
- ktopb = 2 for ideally conduction inner core
- kbotb = 1 for insulating mantle (fixed)
- imagcon < 0
- bnch1
+ ktops = 1 fixed temp at CMB
+ ktops = 2 fixed radial heatflow @CMB
+ kbots = 1 fixed temp at ICB
+ kbots = 2 fixed radial heatflow at ICB
+ ktopv = 1 velocity condition is free at CMB
+ kbotv = 2 velocity condition is rigid at CMB
+ ktopb = 1 for insulating mantel
+ ktotb = 2 for ideally conduction inner core
+ kbotb = 1 for insulating inner core (fixed)
+ imagcon < 0
+ bnch1
ktops 1 2 1 2
kbots 1 2 2 1
ktopv 2 1 1 2
kbotv 2 2 1 1
ktopb 1 2
- kbotb 1 1
- x y y y y y y
+ kbotb 1 1
+ x y y y y y y
When ktops and kbots both set to 2, i.e. fixed radial heatflow at both CMB and ICB, MAG crashes.
11. Movie file generation:
- imovopt: 100 -> mm.
- 010 -> ma.
- 001 -> me.
- 0001-> me.
- 1111-> me. ma. mm. cc.
- 1001-> cc. me.
- 1000-> cc.
+ imovopt:
+ 0100 -> mm.
+ 0010 -> ma.
+ 0001 -> me.
+ 1111 -> me. ma. mm. cc.
+ 1001 -> cc. me.
+ 1000 -> cc.
+ 5000 -> cc.
+
+ cc file was created with imovopt=1###, but all cc files were empty
+ in bench1:
+ iframes = 0
+ tmovstart = 0
+ tmovsteps = 0
+ did not test above three parameters.
+
+ In amhd.f, moved call cmbcoeff(kcv) out of if statement like following:
+
+ if((time/tscale.ge.tmovnext).and.(imovct.le.iframes)) then
+c if(kc.eq.nn) then
+c if(imovopt.ge.1000) then
+c kcv=imovopt/1000
+c call cmbcoeff(kcv)
+c endif
+c imovct=imovct+1
+c tmovnext=tmovnext+tmovstep
+c endif
+ kcv=imovopt/1000
+ call cmbcoeff(kcv)
+ kvp=mod(imovopt,1000)/100
+ if(mod(imovopt,10).ge.1) call moveout(kc)
+ if(mod(imovopt,100).ge.10) call movaout(kc)
+ if(mod(imovopt,1000).ge.100) call movmout(kc,kvp)
+ endif
+
+ cc file was empty. Need to exam cmbcoeff.f
+
+12. Hyperdiffusivities:
+ bench1:
+ difamp: = 2. (amplitude of hyperdiffusivities)
+ ldif: = 1 (hyperdiffusivites applied for harmonic degrees l>= ldif)
+ ldifexp:= -1 (exponet for increase of hyperdiffusities with l)
+ test:
+ difamp: =4 and Ra=10e6 (increases the convergence)
+ ldifexp:=-2 (damping fact)
+13. re-scale parameters:
+ bench1
+ amps 1
+ ampw 1
+ ampz 1
+ ampb 1 2 -1
+ ampj 1 2 -1
+
+
+To Do:
+ IDL:
+ 1. magsym.pro to read in unformated data
+ 2. Add color bar in display
+ 3. add error handling in magsym while read in data file
+ MAG:
-
- IDL:
-
- 1. magsym.pro to read in unformated data
\ No newline at end of file
+ 1. add more field in output files l. ls.
More information about the cig-commits
mailing list