[gmx-users] weird mdrun...using gromacs 3.3.3

[Naga Rajesh Tummala]

Hi Justin,

I tried "unconstrained_start = yes". It did not work.
I freeze CNT, CMT groups rather than constraints.

To try from scratch, I deleted all the groups apart from CNT and CMT, then
solvated the box with water.
I ran a short simulation for 5000 steps, it runs fine. But when I tried to
resubmit the simulation from the confout.gro generated by GROMACS at the end
of the 5000 step simulation, the simulation crashes.

Below are the energies from the log file for 5000th step in the first
simulation and first step in the second simulation

Step           Time         Lambda
           5000       10.00000        0.00000

   Energies (kJ/mol)
           Bond        LJ (SR)  Disper. corr.   Coulomb (SR)   Coul. recip.
    1.22013e+02    1.14904e+05   -1.22580e+03   -7.15625e+05   -1.80614e+04
      Potential    Kinetic En.   Total Energy    Temperature Pressure (bar)
   -6.19887e+05    1.01181e+05   -5.18706e+05    2.98975e+02    4.13525e+01

The below is the energy as soon as the simulation starts.

Energies (kJ/mol)
           Bond        LJ (SR)  Disper. corr.   Coulomb (SR)   Coul. recip.
    1.48991e+02   -1.09357e+34   -1.22580e+03   -7.15625e+05   -1.80568e+04
      Potential    Kinetic En.   Total Energy    Temperature Pressure (bar)
   -1.09357e+34            inf            inf            inf            inf

The only thing I changed is to use the final configuration from the first
simulation.

Can anyone guess the reason for such weird behavior ??

Thanks
Rajesh
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-users/attachments/20090908/69817b46/attachment.html>