[lammps-users] A strange result for “ fix npt”

Hi all:

I got a deformed box after npt run with style “xyz”. At the beginning, the length abc for the box is “27.350174 27.743213 . After 360000 run, they change to “40.909782 27.702092 .According to the introduction in the Manuel for “fix npt”, the ratio of the box length should keep as a constant. For example, begin with “ 10 10 and end with “ 1 1 . Am I right?

My calculations are:

  1. Command. “fix npt GB npt/asphere 2.99 2.99 0.1 xyz 4.4 4.4 20 dilate partial”

  2. Model. The model is composed with a big box of gay-berne particles and a frozen LJ cylinder inside. The beginning structure is an equilibration model under temperature 1.6 which was obtained by “fix npt aniso”. It is stranger that the round cylinder change to an ellipse based on the deformed box even if I used “dilate partial”.

  3. Some parts of the log file.

fix npt GB npt/asphere 2.99 2.99 0.1 xyz 4.4 4.4 20 dilate partial

Resetting global state of Fix npt Style npt/asphere from restart file info

run 360000

Memory usage per processor = 7.74627 Mbytes

Step Temp a_temp Press Volume PotEng KinEng Lx Ly Lz Pxx Pyy Pzz Pxy Pxz Pyz

0 1.1152402 1.6140644 4.2056315 55429.878 -9.220792 1.6727647 27.350174 27.743213 73.051153 4.0576136 4.1164169 4.442864 0.091552759 -0.033139145 -0.0046028051

30 1.2946492 1.867878 4.6535962 55428.957 -9.0815238 1.9418628 27.350665 27.742776 73.049779 4.4933248 4.5434126 4.9240512 0.089522068 -0.037508084 -0.0081701676

……

157260 2.6269575 3.8154234 3.9850509 79760.434 -2.3971322 3.940211 34.920464 29.689816 76.930756 4.1430258 3.9904494 3.8216776 -0.04386914 -0.036192763 0.06978516

157290 2.6210953 3.8047651 4.0600687 79768.83 -2.3699715 3.9314182 34.922509 29.690554 76.932435 4.1309455 4.118975 3.9302855 -0.019013488 -0.054406976 -0.021494459

…….

359940 2.6564822 3.8150489 4.0817145 79672.389 -2.3678149 3.9844955 40.90657 27.701783 70.308374 4.0459204 4.2123908 3.9868323 0.071709891 -0.0021466466 0.0019907827

359970 2.6299152 3.8167443 3.9938929 79676.102 -2.3939057 3.9446473 40.908165 27.701931 70.308534 4.0254625 4.0933386 3.8628774 -0.18279552 0.014836804 -0.0062076119

360000 2.6175971 3.7787918 4.0704951 79679.935 -2.3715327 3.9261712 40.909782 27.702092 70.308728 4.0878285 4.1204091 4.0032475 -0.046842745 -0.079709734 0.044526091

Loop time of 234162 on 4 procs for 360000 steps with 17492 atoms

Since you did an equil with fix npt aniso (which would change the
box lengths independently) are you sure you started the fix
npt xyz with a equi-sized box? Please print out lx ly lz with your
thermo output. If it is still changing, please post your input
script and a small data file.

Steve

2009/1/27 Kitty Ji <[email protected]>: