Hi all,
i am willing to use an nve integrator in one of my problems. I was thinking if at the same time can i use both a temp rescale and a pressure rescale. i mean from the manual it says nothing about any restrictions,but is it physically correct or it belongs to what is called garbage in -garbage out?
Hi all,
hi nelson,
i am willing to use an nve integrator in one of my problems. I was thinking
if at the same time can i use both a temp rescale and a pressure rescale. i
mean from the manual it says nothing about any restrictions,but is it
physically correct or it belongs to what is called garbage in -garbage out?
well, i would go with temp/berendsen and pressure/berendsen.
temp/rescale literally gives your system a "kick" whenever it is
triggered and that may not be desirable.
the interesting question is why would you want to use these
over the npt integrator?
if you worry about ensemble fix npt is the thing to use.
if you want ease of use */berendsen is recommended.
if you want fast heat dissipation, temp/langevin with a suitable
friction term is a good choice. but you have to be careful with
low density gas phase systems, as this fix with these settings with
quickly convert all rotation of molecules quickly into vibrations.
am using this deliberately for my interactive MD experiments as
the friction term helps to make it easier to manipulate. free moving
larger objects stop after a bit and that is useful for the kind of
"playing around" purposes, but it is rather unphysical unless you
have a dense solvent.
cheers,
axel.
Dear Axel,
I have a gas,with the density given at T=300K and P=1atm let’s say. I put my molecules randomly ,in a box in a way that: from the density which is given from my data (which is in kg/m^3) ,i convert it to number density. Then i decided how many molecules i want to simulate and then i calculate the needed volume. so far i think i am correct.as i first step i tried to use an npt integrator,but i saw that my molecules are "behaving’’ like they want to go to a bigger volume.I can’t explain why, that’s i wanted to see if i use an nvt or nve with pressure and temp rescaling to 1atm and 300 K respectively. what’s your opinion on this?
P.S when i am saving something in a file like .lammpstrj is it possible to include in vmd the box limits so to see what’s going on? or is there an other way through for example a .xyz file? Because i used it and vmd says that it can’t be opened.
Thank’s again.
dear nelson,
Dear Axel,
I have a gas,with the density given at T=300K and P=1atm let's say. I put my
molecules randomly ,in a box in a way that: from the density which is given
from my data (which is in kg/m^3) ,i convert it to number density. Then i
decided how many molecules i want to simulate and then i calculate the
needed volume. so far i think i am correct.as i first step i tried to use an
npt integrator,but i saw that my molecules are "behaving'' like they want to
go to a bigger volume.I can't explain why, that's i wanted to see if i use
an nvt or nve with pressure and temp rescaling to 1atm and 300 K
respectively. what's your opinion on this?
it was mentioned here before, that if your system has already
the desired target density, then it is by far the best strategy to
start with fix nvt, or fix nve and temp/berendsen and no change
in the box volume until the system has equilibrated. you can
monitor the pressure in the thermo output for that. if it stabilizes
to fluctuate around a fixed value, then and only then i would switch
to fix npt or temp/berendsen plus press/berendsen. to start
with volume change when your system is not yet equilibrated
with artificially blow up your system and it will take much longer
to get it back to normal than if you force it to stay a "normal"
density and then enable volume changes later.
P.S when i am saving something in a file like .lammpstrj is it possible to
include in vmd the box limits so to see what's going on? or is there an
if you use the atom style dump or a properly configured custom dump
the lammps trajectory reader will read and apply the box, for as long
as it is orthogonal. support for non-orthogonal boxes is still lacking
and it ranks fairly low on my TODO list at the moment.
once you load the dump file, you should be able to show
the system cell with "pbc box" depending on your input,
you may need to set the center of the box differently.
other way through for example a .xyz file? Because i used it and vmd says
that it can't be opened.
xyz is the absolute lowest common denominator of file formats and
thus provides the least amount of information.
cheers,
axel.
I have read about that. I i chose to use only npt ,then the only difference would be that it will take me longer? i mean i simulated my system for 2ns ,and before that for 100ps for equilibration noticing that it has been equilibrated.
I have read about that. I i chose to use only npt ,then the only difference
would be that it will take me longer? i mean i simulated my system for 2ns
yes. at least in my limited experience. it may be _much_ longer
and thus a serious waste of (computer) time.
i don't see any reason why one would not first use nvt
or nve with fix langevin or temp/berendsen.
,and before that for 100ps for equilibration noticing that it has been
equilibrated.
cheers,
axel.
Fix langevin is a nice thermostat to use with fix nve. The only
barostsat to use with fix nve is fix press/berendsen. Fix npt
does its own integration.
Steve
Hi all,
i am willing to use an nve integrator in one of my problems. I was thinking
if at the same time can i use both a temp rescale and a pressure rescale. i
mean from the manual it says nothing about any restrictions,but is it
physically correct or it belongs to what is called garbage in -garbage out?------------------------------------------------------------------------------
_______________________________________________
lammps-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/lammps-usersHi all,
hi nelson,
- Show quoted text -
i am willing to use an nve integrator in one of my problems. I was thinking
if at the same time can i use both a temp rescale and a pressure rescale. i
mean from the manual it says nothing about any restrictions,but is it
physically correct or it belongs to what is called garbage in -garbage out?
well, i would go with temp/berendsen and pressure/berendsen.
temp/rescale literally gives your system a "kick" whenever it is
triggered and that may not be desirable.
the interesting question is why would you want to use these
over the npt integrator?
if you worry about ensemble fix npt is the thing to use.
if you want ease of use */berendsen is recommended.
if you want fast heat dissipation, temp/langevin with a suitable
friction term is a good choice. but you have to be careful with
low density gas phase systems, as this fix with these settings with
quickly convert all rotation of molecules quickly into vibrations.
am using this deliberately for my interactive MD experiments as
the friction term helps to make it easier to manipulate. free moving
larger objects stop after a bit and that is useful for the kind of
"playing around" purposes, but it is rather unphysical unless you
have a dense solvent.
cheers,
axel.
Would it be nice-correct to use at the same time nve with a thermostat and a barostat?
Would it be nice-correct to use at the same time nve with a thermostat and a
barostat?
yes. that is no problem. check the corresponding doc pages.
cheers,
axel