bug (triggered by) create_atoms command

LAMMPS LAMMPS Mailing List Mirror
1

Hello

    I found a bug in lammps-30Oct14 which is set off when I use
"create_atoms", followed by "write_data". I have attached an example
script and molecule template file which causes the error:

ERROR: Atom count is inconsistent, cannot write data file
(../write_data.cpp:150)

    If it helps, to my eyes, it appears that the evil magic occurs
near this line: comm_brick.cpp:837
(which is invoked by write_data.cpp:118)
I say this because at this moment, the numbers in atom->x[1][0],
x[1][1], and x[1][2] no longer look correct to me after this point,
but I could be wrong about this.

Cheers!
Andrew

    (Just to be clear, I have my own customized version of
create_atoms.cpp to submit. But I am not using it for this bug
report. I am using the original version from October30th.)

in.create.molecule (540 Bytes)

molecule.dimer (153 Bytes)

2

this should be because you are inserting an atom outside of a nonperiodic
boundary (z direction), such that the molecule is straddling the boundary

Thanks Jacob!
You are right. There is no bug. I copied the input script from
"examples/deposit", (which uses "boundary p p f") and forgot to set
the boundary conditions to:
boundary p p p

Doing this eliminates the error. Increasing the box size does too.

I apologize.
Please ignore this post

Andrew