# [lammps-users] the desired density/volume

Dear Lammps users,

The desired length of the box is determined by the concentration(density):

N=# of particle=888=512

Low density for dilute lattice=0.01

Vbox = N / rho = 512 / 0.01 = 51200

Lbox = Vbox^(1/3) = 51200^1/3=37.13

Desired values:

Desired density=0.3

Vdesired_box = N / rho = 512 /0.3 = 1706.67

Ldesired_box =1706.67^1/3=11.95

Is there any error here?

I started with a low density and I used fix_deform to compress the box. I tried to scale the box size every 10 steps by a very small
factor (1% or even less) until I reach the desired density.Finally I run in the NVT
ensemble.

thermo output is (near) INF after 5 000 000 step.

What do I should?
Where are I making mistakes?
İnput file is attachment

Result:
Step Temp TotEng Volume E_pair Press PotEng
0 2 2.9941406 51200 0 0.019960938 0
100 2 2.9941406 51197.917 0 0.01996175 0

4999900 2 2.0278406e+38 1706.7057 2.0278406e+38 9.1971787e+41 2.0278406e+38
5000000 2 1.6266404e+39 1706.4899 1.6266404e+39 8.7761447e+42 1.6266404e+39

Step Temp TotEng Volume E_pair Press PotEng
5000000 2 2.0463684e+39 1706.4899 2.0463684e+39 1.125387e+43 2.0463684e+39
5000100 nan nan nan 0 nan 0
5000200 nan nan nan 0 nan 0

in.ellipse.gayberne (839 Bytes)

If any thermodynamic quantity (pressure, temp) blows
up while you're shrinking the box, then your simulation
is bad b/c the ellipsoids have become too overlapped.

You could be shrinking too fast. You might
try NPT instead. You say you want a density
of 0.3, but for ellipsoids. Are you taking into
account the extra size of an ellipsoid? 0.3 for
spheres and 0.3 for ellipsoids are not the same thing.

2009/9/7 ahmet yıldırım <[email protected]...>: