[lammps-users] Questions about command "velocity" and reduced units

LAMMPS LAMMPS Mailing List Mirror
1

Hello everyone,

I have got two questions here:

  1. When performing simulation using LJ potential, I wrote this line in the input file
    “velocity all create 3.0 …”.
    But the output temperature during the simulation is about 1.6.
    I tried other temperatures, the result temperature is slightly above one half of the previously set value.
    I just wonder why is this happen?

  2. I have read some explainations about the concept “reduced units”, but haven’t yet get a complete understanding of it.
    As there are three basic units used to derive other units.
    The mass unit is the weight of the atom,
    the sigma (length unit) is around the diameter of the atom,
    the value of epsilon(energy unit) divided by k is around the critical temperature of the atom(e.g. Ar or Xe)
    Besides, is there any relationship between the time unit calculated with these units and the timestep employed in the simulation?
    Am I right? Can anyone show me exactly how are the value of the basic units derived?

These two questions maybe quite simple, but are important in the way of my learning.
Any reply is greatly appreciated.

Chen Chen,
Wuhan Univ., China
May 23, 2007

2

Chen,

1. When performing simulation using LJ potential, I wrote this line in
the input file
      "velocity all create 3.0 ...".
      But the output temperature during the simulation is about 1.6.
      I tried other temperatures, the result temperature is slightly above
one half of the previously set value.
      I just wonder why is this happen?

There are several possibilities. I'd have to see your input script to
know which one is the case. Here's some ideas:
1) The simulation initally starts at 3.0, but varies from it b/c it is
NVE ensemble rather than NVT.
2) You might be using shake. This changes the degrees of freedom and
shifts the initial T.
3) You could have an error in your input script.

2. I have read some explainations about the concept "reduced units", but
haven't yet get a complete understanding of it.
      As there are three basic units used to derive other units.
      The mass unit is the weight of the atom,
      the sigma (length unit) is around the diameter of the atom,
      the value of epsilon(energy unit) divided by k is around the critical
temperature of the atom(e.g. Ar or Xe)
      Besides, is there any relationship between the time unit calculated
with these units and the timestep employed in the simulation?
      Am I right? Can anyone show me exactly how are the value of the basic
units derived?

You can look in Allen & Tildesley's "Computer Simulation of Liquids"
in Appendix B where there's a nice discussion of reduced units.

t* = (k T/(m sigma^2))t

Paul

3

dear list

i am interested in simulating the flow past an obstacle

i found one example problem in /examples directory

i am having a problem in creating an obstacle without defining a lattice

as i am reading input data file with positions and velocities, i didn,t define lattice

so when i used "region void1 sphere 20 10 5 3" to create a spherical obstacle, it printed the following error

"Use of region with undefined lattice"

please let me know if there is anyway to create a obstacle if you are reading a data file rather than creating a lattice

thanks and regards

anki

4

Anki,

LAMMPS is actually informing you that you are trying to create a simulation without a simulation box. As an example, if you were to want to create an empty box of dimensions ((-lx,lx), (-ly,ly), (-lz,lz)) you could do the following:

lattice bcc 2.0
region box block -lx lx -ly ly -lz lz
create_box 1 box
mass 1 1

I added in the mass, because LAMMPS wants masses to be defined when you try to run an empty box. Of course, in your case you most likely already have your masses defined, so you can omit the mass line.

Pieter

5

Anki,

Yes, it is doable.

There are two issues here: first you have to create a void space, and
second you have to maintain that void space. To maintain the void
space, you could use fix indent. See:
http://lammps.sandia.gov/doc/fix_indent.html

To create a void space, you can either delete atoms that are in the
way, push the atoms out of the way, or not have them there in the
first place. Presumably the latter isn't an option in your case.
Sounds like you're trying to delete atoms that are in the way, which
might be a bad idea depending on your system. Why not push atoms out
of way if they're sitting in your void space? You could do this by
simply imposing the indent fix and using LAMMPS's minimizer to push
them off, or alternatively, gradually increasing the radius of your
sphere using a series of commands like this: fix indent, run, unfix,
fix indent, run, unfix, etc.

I hope this helps.

Paul

6

dear list

i am trying to simulate couette flow of particles using Dissipative particle Dynamics interactions

i replaced pair_style from lj/cut to dpd style as shwon below

without changing other commands in the input script in.flow.couette given in examples/flow directory

# DPD pair style

pair_style dpd 1.0 2.5 34387
pair_coeff * * 3.0 1.0

after visualisation through xmovie i found that many active particles are going out of the simulation box

i am still trying to learn this DPD technique,

hence i request if anybody used dpd interactions for this problem, please let me know /// complete input script // /examples/flow/

# 2-d LJ flow simulation

dimension 2
boundary p s p

atom_style dpd
neighbor 0.3 bin
neigh_modify delay 5

# create geometry

lattice hex 0.7
region box block 0 20 0 10 -0.25 0.25
create_box 3 box
create_atoms 1

mass 1 1.0
mass 2 1.0
mass 3 1.0

# DPD pair style

pair_style dpd 1.0 2.5 34387
pair_coeff * * 3.0 1.0

# define groups

region 1 block INF INF INF 1.25 INF INF
group lower region 1
region 2 block INF INF 8.75 INF INF INF
group upper region 2
group boundary union lower upper
group flow subtract all boundary

set lower atom 2
set upper atom 3

# initial velocities

temperature mobile flow full
velocity flow create 1.0 482748 temp mobile
fix 1 all nve
fix 2 flow temp/rescale 200 1.0 1.0 0.02 1.0
fix_modify 2 temp mobile

# Couette flow

velocity lower set 0.0 0.0 0.0
velocity upper set 3.0 0.0 0.0
fix 3 boundary setforce 0.0 0.0 0.0
fix 4 all enforce2d

# Poisseuille flow

#velocity boundary set 0.0 0.0 0.0
#fix 3 lower setforce 0.0 0.0 0.0
#fix 4 upper setforce 0.0 NULL 0.0
#fix 5 upper aveforce 0.0 -1.0 0.0
#fix 6 flow addforce 0.5 0.0 0.0
#fix 7 all enforce2d

# Run

timestep 0.003
thermo 500
thermo_modify temp mobile

dump 1 all atom 50 dump.flow
run 10000

thanks and regards

anki

7

dear list

  i am trying to simulate couette flow of particles using Dissipative particle

8

Anki,

DPD particles do not have a solid core. They can move through each
other. So the walls in this example, which consist of particles, no
longer hold the fluid in. You might try adding on one of the wall
fixes to hold the particles in. But overall, this example problem was
not intended to be used with DPD particles.

Paul