Problem with change_box in elastic constants script

Hi all,

I'm a new lammps user and have a question about the elastic constants script. I'm using the 2012.02 version of lammps, and I run into a problem when I use a customized read_data file. I'm simulating amorphous silica using the BKS potential, and I am able to calculate elastic constants using an older version of the EC script that uses the displace_box command. Keeping the entire script the same, I only change displace_box to change_box, but now whenever a strain deformation is implemented, the energy blows up to infinity during the next minimization. My understanding is that change_box absorbed the displace_box commands, but I'm wondering if I'm missing something about how change_box operates (I've read through the documentation). I wrote my own script to calculate elastic constants as well, and the identical problem exists.

Any suggestions about why this might be happening or a solution? I've attached my potential script and my log.lammps output. The diverging energy is at the end of the log.


**Potential Input**

# ---------------------------------- need ewald for BKS2 and BKS3 silica only
kspace_style ewald/n 1.0e-5

# ---------------------------------- silicon

# pair_style sw # this is for silicon
# pair_coeff * * elastic_Si.sw Si

# ---------------------------------- LJ silica

# pair_style lj/cut 7
# pair_coeff 1 2 2.318861e-5 3.233825
# pair_coeff 1 1 7.981161e-8 3.302029
# pair_coeff 2 2 6.738512e-5 3.165553

# ---------------------------------- BKS2 silica

  pair_style buck/coul/long 8.5 8.5
  pair_coeff 1 2 18003.7572 0.2052048 133.5381 # 1 for Si, 2 for O
  pair_coeff 2 2 1388.7730 0.3623188 175.0000
  pair_coeff 1 1 0.0000 1.0000000 0.0000

# ---------------------------------- BKS3 silica

# pair_style buck/coul/long 8.5 8.5
# pair_coeff 1 2 27029.41992 0.19385 148.09909
# pair_coeff 2 2 659.59540 0.38609 26.83668
# pair_coeff 1 1 3150.46265 0.35070 626.75195

neighbor 1.0 nsq
neigh_modify once no every 1 delay 0 check yes

min_style cg
min_modify dmax ${dmax} line quadratic

thermo 1
thermo_style custom step temp pe press pxx pyy pzz pxy pxz pyz lx ly lz vol
thermo_modify norm no

log.lammps (66.2 KB)

I'm using the 2012.02 version of lammps,

I don[t know what version you mean. Can you try
the most current version?

If the problem persists, can you post an input
script (and small data file if needed) that
illustrate something that works with the
old displace box and does not work with the new
change box?


Thanks - I updated to the newest version from the beginning of June and it works correctly now. Looks like there was a bug from a February version that has now been fixed.

Thanks again!