[CPMD-list] MAXIT

Ari P Seitsonen Ari.P.Seitsonen at iki.fi
Fri Dec 12 15:38:21 CET 2003


Dear Sergei,

  Your simulation is diverging, it simply doesn't make sense to try to 
continue the run anyway! Maybe the timestep is too large, the fictitious 
mass too small, you didn't use "RESTART VELOCITIES", your system is 
metallic or something like that. How did the previous run end, was it an 
optimisation or already MD? Can you send the input file on the mailing list?

    Greetings,

         apsi

Sergei Lisenkov wrote:

>Dear CPMD users,
>
>I have met an error in MD run. From manual I don't know how to solve it:
>
>....
>(K+E1+L+N+X)           TOTAL ENERGY =         -305.18262043 A.U.
> (K)                  KINETIC ENERGY =          307.83507678 A.U.
> (E1=A-S+R)     ELECTROSTATIC ENERGY =         -303.98985038 A.U.
> (S)                           ESELF =          319.15382432 A.U.
> (R)                             ESR =           13.10960226 A.U.
> (L)    LOCAL PSEUDOPOTENTIAL ENERGY =         -242.09751838 A.U.
> (N)      N-L PSEUDOPOTENTIAL ENERGY =           38.65432287 A.U.
> (X)     EXCHANGE-CORRELATION ENERGY =         -105.58465132 A.U.
>          GRADIENT CORRECTION ENERGY =           -1.09383941 A.U.
>
>       NFI    EKINC   TEMPP           EKS      ECLASSIC          EHAM     DIS    TCPU
>         1343.03311 23665.6    -305.18262    -298.66287      44.37024   0.002   26.09
> FILE ./TRAJECTORY EXISTS, NEW DATA WILL BE APPENDED
> FILE ./TRAJEC.xyz EXISTS, NEW DATA WILL BE APPENDED
> FILE ./MOVIE EXISTS, NEW DATA WILL BE APPENDED
> RORTHO:DIFGAM= ???????????????????????  ITER =           31  >           30
>        MAXIMUM NUMBER OF ITERATIONS EXCEEDED
>
>
> PROGRAM STOPS IN SUBROUTINE RORTOG|   [PROC=   0]
>
>And another thing. This calculations was done with Troul.Martins pseudo. I tried also a Vanderbil potential, and I got:
>(K+E1+L+N+X)           TOTAL ENERGY =          274.90586009 A.U.
> (K)                  KINETIC ENERGY =          222.98588130 A.U.
> (E1=A-S+R)     ELECTROSTATIC ENERGY =          173.95634078 A.U.
> (S)                           ESELF =          319.15382432 A.U.
> (R)                             ESR =           13.05706950 A.U.
> (L)    LOCAL PSEUDOPOTENTIAL ENERGY =          -77.79274991 A.U.
> (N)      N-L PSEUDOPOTENTIAL ENERGY =           23.86876994 A.U.
> (X)     EXCHANGE-CORRELATION ENERGY =          -68.11238202 A.U.
>          GRADIENT CORRECTION ENERGY =           -0.22392827 A.U.
>
>       NFI    EKINC   TEMPP           EKS      ECLASSIC          EHAM     DIS    TCPU
>         1137.98105  1171.3     274.90586     275.22854     413.20959   0.000   27.45
> FILE ./TRAJECTORY EXISTS, NEW DATA WILL BE APPENDED
> FILE ./MOVIE EXISTS, NEW DATA WILL BE APPENDED
>         2498.01012  3789.1    -164.32603    -163.28215     334.72797   0.000   27.33
>         3449.64236  3870.6    -137.66681    -136.60049     313.04187   0.000   27.27
>         4297.53068  5889.6      96.62376      98.24631     395.77698   0.000   27.38
>         5400.23621  9141.1     -63.55281     -61.03448     339.20173   0.001   27.28
>         6436.88253 11604.6      62.38252      65.57953     502.46206   0.002   27.35
>         7487.16340 13737.2       8.22484      12.00937     499.17276   0.003   27.22
>         8482.57376 14948.3       1.25990       5.37808     487.95184   0.004   27.60
>         9437.18224 17654.0      54.20901      59.07260     496.25484   0.006   27.25
>        10440.60420 21855.3      39.80961      45.83064     486.43484   0.008   27.31
>        11457.87573 25703.0      18.48448      25.56552     483.44125   0.011   27.37
>        12445.60822 29257.3      32.02978      40.09001     485.69823   0.015   27.42
>        13433.76948 33646.8      44.99931      54.26884     488.03832   0.020   27.31
>        14443.40019 38383.0      31.58492      42.15926     485.55944   0.026   27.33
>...
>
>You can see that process is disconverged. Why? 
>
>Thanks,
>Sergey
>_______________________________________________
>CPMD-list mailing list
>CPMD-list at cpmd.org
>http://www.cpmd.org/mailman/listinfo/cpmd-list
>  
>





More information about the CPMD-list mailing list