Question about read_data add append

LAMMPS LAMMPS Beginners
1

Hi all,

I seem to be misunderstanding the use of read_data with the “add append” keywords, and I can’t figure out where I’m going wrong. I’m trying to create a simulation of a LJ fluid over a substrate. The fluid atoms are created with the create_atoms command, then the substrate is added via a “read_data filename add append” command.

My test input file is:

#----
units metal
dimension 3
boundary p p f
lattice sc 3.840

region liquidbox block 0 7.680 0 7.680 1.920 5.760 units box
region simbox block 0 7.680 0 7.680 -5.702 5.760 units box
create_box 2 simbox

create_atoms 1 region liquidbox
mass 1 5.0

read_data substrate.data add append group S

set group S type 2

mass 2 1.0

dump test all custom 25 “substrate.lammpstrj” id type mass xu yu zu
dump_modify test sort id

run 0
#----

The file substrate.data contains the positions of 16 atoms, all assigned type 1.

I expected this input to create four type 1 atoms with mass 5, then add 16 atoms from the data file as type 2 atoms, since my understanding is that “add append” should increment the atom type found in the data file by 1 (since there is 1 atom type in the current system when the read_data command is reached). Instead, I get 20 atoms, all of type 1.

I can manually assign the 16 substrate atoms to type 2 by uncommenting the “set group S type 2” line, but that has the problem that the read_data command resets the mass of the type 1 atoms.

Clearly I’m misunderstanding something fundamental. Thanks for your time and assistance. The text of the substrate.data file is below.

LAMMPS data file. atom_style atomic generated by CGG on Thu 18 Sep 2025, 02:41 PM
16 atoms
1 atom types
0 7.680 xlo xhi
0 7.680 ylo yhi
-5.431 0 zlo zhi

Masses

1 28.0855 # SI

Atoms # atomic

1 1 1.920 0.000 -4.073 # SI
2 1 0.001 0.000 0.000 # SI
3 1 3.840 0.000 0.000 # SI
4 1 1.920 1.920 -2.715 # SI
5 1 0.000 1.920 -1.358 # SI
6 1 0.000 3.840 0.000 # SI
7 1 5.761 0.000 -4.073 # SI
8 1 3.841 1.920 -1.358 # SI
9 1 5.761 3.840 -4.073 # SI
10 1 5.761 1.921 -2.715 # SI
11 1 3.841 3.840 0.000 # SI
12 1 5.761 5.760 -2.715 # SI
13 1 1.920 3.840 -4.073 # SI
14 1 0.001 5.760 -1.358 # SI
15 1 3.840 5.760 -1.358 # SI
16 1 1.920 5.761 -2.715 # SI

2

It is generally much simpler to first load the substrate from a data file and then add the other atoms with create_atoms.

3

Thank you for the prompt reply. I thought of that, but…

While all my simulations have liquid atoms, they don’t all have substrates (sometimes there’s just an external potential), or the substrate might have multiple atom types. I was hoping to always have the atom types for the liquid atoms start at 1, regardless of whether a substrate is present, and regardless of the number of types of atoms in the substrate. Is there a reasonable way to achieve that, or should I abandon the idea of having stable atom types for the liquid atoms and use something like type labels instead?

Thanks.

4

Type labels are meant to be used that way.