[lammps-users] A problem about "fix shake command"

Dear Steve

Recently, I was confused by a problem.

Problem description: I run a simulation.The system was consisted of water molecules and a three-layer platinum

plane which lies at the bottom of the simulatin region along z-axis direction.

The accuracy and iteration in fix shake were 1E-6 and 2E20 respectively. when I run the original

simulation ,the deta of shake statistics became very small after 2 thousand step (timestep=3D1.5fs)

with nve/limit command,which were small enough to continue my following simulation with nve command. But ,when I double the

region size,the statistics can approach to be small enough to continue the simmulation. Would you please do me a favour to give

me some suggestions?

Ps:the encloures are my input and data script.

Best wishes


water.data.txt (110 KB)

in.water (1.6 KB)

Does it run correctly without fix shake (flexible water) and with
out fix nve/limit? I.e. just run with fix nve. If so, I would turn
on one feature at a time and make sure you understand what it is
doing. Also, visuzliaing the system may help you understand if
part of it is freezing and part is heating up, which can be the
case with different kinds of thermostatting.