Hi all,
I have data file for mt lammps input script . I have velocity for all particles there. I like to put momentum zero for my system… (befrore run ) … I am doing DPD simulation … (so momentum will be zero for all steps also …)
how do i do that ? Which .cpp file have to change for that …?
Thnaks in advance…
Hi all,
I have data file for mt lammps input script . I have velocity for all
particles there. I like to put momentum zero for my system.. (befrore run )
.. I am doing DPD simulation ..... (so momentum will be zero for all steps
also ...)how do i do that ? Which .cpp file have to change for that .....?
you only have to read the documentation. try: fix momentum
a.
The velocity all zero linear command will zero the linear momentum
after you read in the velocities.
The fix momentum command will zero if periodically during
a simulation.
Steve
Hello lammps-users
I am having problem simulating TIP3P water with PPPM. Attached is the
input and data file for my 100 molecules water simulation. I am using
Nose Hoover npt and i am seting Tdamp and Pdamp to 100 and 1000
respectively with Tstart=Tstap=300 k and Pstart=Pstop=1 atm. I notice that
the fix does not control the pressure properly, i get fluctuations up to
1000 atmospheres. I ran a similar system with more water and I get back
the density of water but i still get hight fluctaions in pressure. I tried
the drag option of the fix npt and tried the values drag 0.2,1.0 and 2.0
but still pressure fluctuates enormously.
Kindly check my input file. Am I missing some command here?
Thanks
Jan-Michael
Water.dat (17 KB)
in.H2O (1.01 KB)
100 waters is a tiny system. I would expect large
fluctuations in pressure. Maybe Paul can comment
on whether what you are seeing is typical.
One note on your input script: Since you are running
with PPPM, you shouldn't need a 15 Ang cutoff on
the short-range. You may get better performance with
a shorter cutoff.
Steve
Jan-Michael,
I agree with Steve that your system is tiny. What you're seeing is not unexpected. Try again with a much bigger system. Large pressure fluctuations are typical in MD simulations, and smaller systems exhibit larger fluctuations.
I also agree that normally, you'd want a smaller short-range cutoff --- usually around 10 angstroms.
Paul
Hi Paul and Steve, Thanks for your help. I also have another system with
3611 water molecules in them. I have the set points for the thermostat and
barostat to 300 K and 1 atm respectively. I ran up to 100 pico seconds
with an equilibration of 10 pico seconds before gathering data. These are
the average thermo properties:
P: 3.76896 +- 346.655 atm
T: 300.042 +- 4.51091 K
V: 111944 +- 404.135 A^3
The V corresponds to a density of 0.964 g/cc which is pretty much close to
water. I got the 300 K set point for the temperature however you can see
that the P fluctuation is high with a std. dev of 346 atm.
Is this normal? How many water molecules should I have so that these
fluctuations will decrease? I didn't see this problem when I was running
lj systems with the same number of particles in them. P was pretty much
constant in LJ/coarse grain systems.
I foresee that I cannot use the sxx,syy and szz atom stress tensor dumps
to compute surface energy in an oil and water interface system. Any
comment on this is highly appreciated.
Jan-Michael
Jan-Michael,
What you are seeing seems very reasonable to me. Your water box still seems small to me. Water will give larger fluctuations than an LJ fluid. A bigger water box should reduce the magnitude of the fluctuations. You might also play with the pressure control constants, although that won't make the fluctuations go away --- but might help some.
You oil/water interfacial system plan may still be salvageable, but you might want to reconsider the spatial/temporal scales that you'd originally planned on. If you're able to go big enough and simulate for long enough, you should be able to get at the information you want.
Paul
I see Paul,
With regards to the short-range cutoff, 10 A is fine with water only
systems. However, in one of my systems, I have Polystyrene Sulfunate (PSS)
in water solution and I'd get this error,
ERROR on proc 1: Dihedral atoms 527 383 384 809 missing on proc 1 at step
1000
I read the documentation and it said that I need to raise cutoff and doing
some trial and error the smallest and most robust cutoff for that system
was 15. So that is why I am setting everything to 15 A cut offs.
Jan-Michael
Unless the 2 ends of your dihedral are more than 10 Ang (plus the skin) apart,
you shouldn't have this problem. If you are using newton on for bonds, then
it's really the distance from one of the central atoms in the dihedral
to the end atom that needs to be less than the cutoff.
Steve