“hybrid full sphere” for “atom_style“

LAMMPS LAMMPS Mailing List Mirror
1

Maybe I was not clear enough about my problem. As you mentioned, I used read-data to read data file and run my simulation.
I used dump just to make sure that lammps assumes the same sequence of columns as in the data file.
for example when I dumped q, I expected to see the seventh column of data file. but all I got was the ninth column of data file. I thought it means lammps considers the last columns as q and the default sequence I had assumed for the columns (“atom-ID , atom-type , x , y , z , molecule-ID , q , diameter , density”) was not actually right! So I tried to put the q in the last column but I got a warning that said “Density can not be negative.”
What is the correct sequence of columns?

Best,
Fatemeh

1.rar (128 KB)

2

for example when I dumped q, I expected to see the seventh column of data
file. but all I got was the ninth column of data file. I thought it means
lammps considers the last columns as q and the default sequence I had
assumed for the columns ("atom-ID , atom-type , x , y , z , molecule-ID , q
, diameter , density") was not actually right! So I tried to put the q in
the last column but I got a warning that said "Density can not be negative."
What is the correct sequence of columns?

it looks like this is not a problem of giving the correct input, but a
problem how lammps reads it.

i'm attaching a test case, that reduces the issue to the absolute
minimum. this input makes valgrid quite unhappy, which is always a bad
sign.

axel.

p.s.: as indicated before. .rar files are not very portable and thus
are not worth attaching.

atom_hybrid_issue.tar.gz (538 Bytes)

3

then it is possibly a bug - I will look at it when
I'm back in the office - first of next week.

Steve

4

This is the bug fix:
in src/atom_vec_full.cpp, these lines should be:

int AtomVecFull::data_atom_hybrid(int nlocal, char **values)
{
  molecule[nlocal] = atoi(values[0]);
  q[nlocal] = atof(values[1]);

I'll post a patch, but it won't be until next week ...

Steve