[gmx-users] NAN at high temperatures and high pressures

[David van der Spoel]

Popov Konstantin wrote:
> Hello,
> 
> I run simulations with polyethylene molecules with harmonic potential for C-H bonds and Morse potential for C-C bonds. The system consists of 1000 carbon atoms and 2002 hydrogen atoms. The simulation runs with PBC at constant density (~0.8 g cm-3) and no pressure coupling. At the temperature 2900 K about 25% of simulations end up with NAN (not a number) records in their .log and .gro files. Simulations for the same system at the same temperature in vacuum don't have this problem. Simulations run with the same density but lower temperature (2500 K) don't have this problem either.
> Below is a fragment of a .log file when reasonable numbers turn to NANs:
> 
> 
>            Step           Time         Lambda
>           14780       14.78000        0.00000
> 
>    Energies (kJ/mol)
>            Bond          Morse          Angle Ryckaert-Bell.        LJ (SR)
>     2.62634e+04    1.84883e+04    5.57324e+04    6.74611e+03    5.36911e+03
>    Coulomb (SR)      Potential    Kinetic En.   Total Energy    Temperature
>     6.11740e+02    1.13211e+05    1.12145e+05    2.25356e+05    2.99729e+03
>  Pressure (bar)
>     1.49525e+04
> 
>            Step           Time         Lambda
>           14790       14.79000        0.00000
> 
> Group rest with mass  1.40287e+04, Ekrot  8.82852e+00 Det(I) =  1.54313e+12
>   COM:      1.56520       1.58669       1.58357
>   P:       -0.00208       0.00121       0.00119
>   V:       -0.00000       0.00000       0.00000
>   J:      560.09625    -303.20575      13.91409
>   w:        0.02418      -0.01353       0.00097
> Inertia tensor (3x3):
>    Inertia tensor[    0]={ 1.13658e+04, -2.64121e+02,  1.90793e+02}
>    Inertia tensor[    1]={-2.64121e+02,  1.17820e+04, -2.87465e+02}
>    Inertia tensor[    2]={ 1.90793e+02, -2.87465e+02,  1.15395e+04}
>    Energies (kJ/mol)
>            Bond          Morse          Angle Ryckaert-Bell.        LJ (SR)
>     2.48452e+04    1.71570e+04    5.62121e+04    6.85057e+03    5.75860e+03
>    Coulomb (SR)      Potential    Kinetic En.   Total Energy    Temperature
>     6.03772e+02    1.11427e+05    1.13892e+05    2.25319e+05    3.04398e+03
>  Pressure (bar)
>     2.24204e+04
> 
>            Step           Time         Lambda
>           14800       14.80000        0.00000
> 
>    Energies (kJ/mol)
>            Bond          Morse          Angle Ryckaert-Bell.        LJ (SR)
>             nan            nan            nan            nan    0.00000e+00
>    Coulomb (SR)      Potential    Kinetic En.   Total Energy    Temperature
>     0.00000e+00            nan            nan            nan            nan
>  Pressure (bar)
>             nan
> 
> Does anybody have any idea what might cause this?
> 


Time step too large? Since your atoms move faster you have to decrease 
the time step. At 3000 K the T is 10 times higher, ergo motions will be 
three times faster, and hence your time step has to be three times 
smaller. I would suggest 1 fs at 300 K and 0.25 fs at 3000 K.

> Thanks
> 
> K. Popov
> _______________________________________________
> gmx-users mailing list    gmx-users at gromacs.org
> http://www.gromacs.org/mailman/listinfo/gmx-users
> Please search the archive at http://www.gromacs.org/search before posting!
> Please don't post (un)subscribe requests to the list. Use the 
> www interface or send it to gmx-users-request at gromacs.org.
> Can't post? Read http://www.gromacs.org/mailing_lists/users.php


-- 
David van der Spoel, Ph.D., Professor of Biology
Molec. Biophys. group, Dept. of Cell & Molec. Biol., Uppsala University.
Box 596, 75124 Uppsala, Sweden. Phone:	+46184714205. Fax: +4618511755.
spoel at xray.bmc.uu.se	spoel at gromacs.org   http://folding.bmc.uu.se