I noticed that the results I get by using a restart file and without it (which produced the restart file) are different. For example:
In one case, without using a restart file, I get these results for the radius of gyration:
but, when running by a restart file from step 8000000, I get the following results:
I’m using version lammps/28Jun14. The simulation is at a constant temperature of 290. I use this input file for restarting:
boundary p p p
pair_style soft 10
pair_style lj/class2/coul/long 10
pair_modify tail yes table 12
kspace_style ewald 1e-4
neighbor 1 bin
neigh_modify delay 0 every 1 check yes page 1000000
comm_modify mode single cutoff 10 vel no
fix Main all nvt temp 290 290 100 tchain 1
compute g Chain gyration
fix gr Chain ave/time 1 500 500 c_g file tmp.gr
thermo_style custom step temp press vol density etotal ke pe time
thermo_modify lost warn flush yes
restart 500000 restart-main.*
Can you please help me find the problem? Thank you
There is no problem here and what you see is perfectly normal. At 8000500 step you have the same results, then they slowly diverge - which is a well documented bevaviour of MD. Please pick up a MD textbook for more details.
Thank you for your responses. Unfortunately, I could not find the reasons in manual, and this is why I sent email to get some help
The fact is that sometimes, the results achieved by using a restart file are significantly different from those obtained without it, or are even contrary to each other. I think this may be due to the instability of the system or my simulation.
Thank you for your responses. Unfortunately, I could not find the reasons
in manual, and this is why I sent email to get some help
please think about this statement for a little bit. you are essentially
telling the people WHO WROTE THE MANUAL, that what they say they did, and
*know* they did, isn't there. how likely is that?? how much more likely is
that you didn't look properly. answers don't always appear in the way that
you want them to be and jump right at you, but you need to *work* to obtain
The fact is that sometimes, the results achieved by using a restart file
are significantly different from those obtained without it, or are even
contrary to each other. I think this may be due to the instability of the
system or my simulation.
this very topic has been discussed many times on the mailing list, and is
also discussed in the text book literature, and is something that any
practitioner of MD should have some fundamental knowledge about. so please
make an effort and do your homework on that.
you will find much less friendlier answers then when insisting on thing not
being the way they are.