Topic Options
#15361 - 08/28/07 12:27 AM restart dynamics
Manori Offline
Forum Member

Registered: 09/24/04
Posts: 251
Loc: oh.USA
Dear charmmers,

I have used a restat file to restart the dynamics. The dynamics at the end of the 1st run is not the same as the dynamics at the begining of the 2nd run (restart run). I think they both should be the same. But the numbers (the temperature ,potential energy, total energy etc.) different. Please explain me a possible reason for this difference. I have cpoied some parts of two out put files.

Thank you very much

Manori


end of first run (start run) data


DYNA> 10000 10.00000 -50520.03149 11988.11353 -62508.14502 296.41042
DYNA PROP> 16.96142 -50514.51219 12016.04252 5.51929 9929.32890
DYNA INTERN> 499.69612 1298.59808 197.26483 906.23210 82.43915
DYNA CROSS> -225.42707
DYNA EXTERN> 4853.11816 -53034.66686 0.00000 0.00000 0.00000
DYNA IMAGES> 233.08008 -4739.42894 0.00000 0.00000 0.00000
DYNA EWALD> 649.68820-405825.17108 392209.30340 0.00000 0.00000
DYNA CONSTR> 387.12880 0.00000 0.00000 0.00000 0.00000
DYNA PRESS> 416.72141 -7036.27400 -149.24370 350.43741 191457.74647
DYNA XTLE> -59654.23032 555.81739 -9128.55783 135.26752
---------- --------- --------- --------- --------- ---------
Crystal Parameters : Crystal Type = HEXA
DYNA A = 51.17900 B = 51.17900 C = 84.40318
DYNA Alpha = 90.00000 Beta = 90.00000 Gamma = 120.00000
DYNA PIXX = 38.62 PIYY = -162.44 PIZZ = 1175.13
DYNA PIXY = 14.72 PIXZ = 386.51 PIYZ = 632.72
DYNA Gradient Norm = 29.99045



begining of second run(restart run) data
NUMBER OF DEGREES OF FREEDOM = 59025
DYNA DYN: Step Time TOTEner TOTKe ENERgy TEMPerature
DYNA PROP: GRMS HFCTote HFCKe EHFCor VIRKe
DYNA INTERN: BONDs ANGLes UREY-b DIHEdrals IMPRopers
DYNA CROSS: CMAPs
DYNA EXTERN: VDWaals ELEC HBONds ASP USER
DYNA IMAGES: IMNBvdw IMELec IMHBnd RXNField EXTElec
DYNA EWALD: EWKSum EWSElf EWEXcl EWQCor EWUTil
DYNA PRESS: VIRE VIRI PRESSE PRESSI VOLUme
DYNA XTLE: XTLTe SURFtension XTLPe XTLtemp
---------- --------- --------- --------- --------- ---------
DYNA> 0 10.00000 -50907.16029 11988.11353 -62895.27381 204.41146
DYNA PROP> 14.01938 -50833.20086 12209.99182 73.95943 -77888.27242
DYNA INTERN> 499.69612 1298.59808 197.26483 906.23210 82.43915
DYNA CROSS> -225.42707
DYNA EXTERN> 4868.43892 -53048.01370 0.00000 0.00000 0.00000
DYNA IMAGES> 217.75933 -4726.08209 0.00000 0.00000 0.00000
DYNA EWALD> 649.68820-405825.17108 392209.30340 0.00000 0.00000
DYNA PRESS> 1895.80791 50029.70704 -678.96052 20832.77047 191457.74647
DYNA XTLE> -59654.23032 555.81739 -9128.55783 135.26752
---------- --------- --------- --------- --------- ---------
Crystal Parameters : Crystal Type = HEXA
DYNA A = 51.17900 B = 51.17900 C = 84.40318
DYNA Alpha = 90.00000 Beta = 90.00000 Gamma = 120.00000
DYNA PIXX = 20656.11 PIYY = 20749.77 PIZZ = 21092.43
DYNA PIXY = 356.70 PIXZ = 98.37 PIYZ = 217.26
DYNA Gradient Norm = 74.26851

Top
#15362 - 08/28/07 03:35 AM Re: restart dynamics [Re: Manori]
lennart Online   content

Forum Member

Registered: 09/25/03
Posts: 4741
Loc: ~ 59N, 15E
You should post also the relevant input - it is very helpful if you can construct very brief input files that exhibit the same behavior. See also the files in the support/forms directory of your CHARMM installation for help on formulating questions.

The setup of a restart run has to be the same as for the start, eg regarding constraints or boundary conditions; an inappropriate setting of CUTNB,CTFONB,INBFRQ,CUTIM,IMGFRQ may also cause a difference between nonbonded energies at the and of a run and the restart of the next piece, when lists have been regenerated. You supervisor should know more about these things.
_________________________
Lennart Nilsson
Karolinska Institutet
Stockholm, Sweden

Top
#15363 - 08/28/07 10:51 AM Re: restart dynamics [Re: lennart]
rmv Online   content

Forum Member

Registered: 09/17/03
Posts: 8373
Loc: 39 03 48 N, 77 06 54 W
Also, if you change any one of CHARMM release, compiler, hardware platform, or number of processors, the repeated run would be expected to diverge.

Top

Moderator:  John Legato, lennart