Repeat the two calculations of 19990831 43 and
19990902 44 for the
system with a smaller time step (
instead
of
) to see if the oxygen MSD plot comes out
reproducible.
This calculation repeats the 19990831 43 one.
In the two calcs being compared, the FIELD and CONFIG files are identical, while the following is the diff between the two CONTROL files:
dschsun1:72> diff -c CONTROL.20000103 CONTROL.19990831 *** CONTROL.20000103 Fri Jan 7 15:46:59 2000 --- CONTROL.19990831 Fri Jan 7 15:46:31 2000 *************** *** 1,9 **** Zr_{100}Ce^{4+}_{200}Ce^{3+}_{200}O_{900}(V_O^{\cdot\cdot})_{100} temperature 1773.0 pres 0.001013 kbar (0.001013 kbar=1.0 atm) ! steps 219921 ! equilibration steps 20000 ! timestep 0.0005 ps multiple timestep 1 scale 1 cutoff 13.8 # maximum allowed value from genlat --- 1,9 ---- Zr_{100}Ce^{4+}_{200}Ce^{3+}_{200}O_{900}(V_O^{\cdot\cdot})_{100} temperature 1773.0 pres 0.001013 kbar (0.001013 kbar=1.0 atm) ! steps 54981 ! equilibration steps 5000 ! timestep 0.002 ps multiple timestep 1 scale 1 cutoff 13.8 # maximum allowed value from genlat *************** *** 13,20 **** eps 1.0 ewald precision 1.0e-5 ensemble nve ! stats 4 ! trajectory nstraj 20001 istraj 80 keytrj 0 stack 10 job time 2000 close time 100 --- 13,20 ---- eps 1.0 ewald precision 1.0e-5 ensemble nve ! stats 1 ! trajectory nstraj 5001 istraj 20 keytrj 0 stack 10 job time 2000 close time 100 dschsun1:73>
i.e. I've reduced the time step by four times, while keeping the same number of dumped configurations to the HISTORY and STATIS files.
NOTE: this calc took 29836 UAC's on the Cray T3E at Cineca!!