DPD simulation bonded atoms collapse into each other

Hi, All,

I am new to DPD simulation and have some simple questions to ask.
What I found is that when I use bond-style harmonic to define the spring force between beads with parameter 2.0 and 0.0, the bonded beads tend to overlap each other, i.e., the g of r has a peak close to zero. While non-bonded beads seem to have normal distance distribution with peak at close to 0.85. (with rou =3)
Even I change the harmonic bond parameters to 2.0 and 0.85, I still have a peak close to zero.
I use aii=aij=25 for every pair interactions.

Is it because that LAMMPS “possibly” neglects the repulsive pair-pair interaction when there is a bond between the pair of beads? The conservative force parameter (25) is much larger than bond force parameter (2*2, consider the later 2 is energy). If the conservative force is always considered, I don’t see a reason there is a peak close to zero.

Another question, how do I know that I am running a DPD simulation? Is the pair_style dpd definition enough to tell the program that I am running a DPD simulation? Should I specify the atom style also as dpd?

Thanks in advance for your help.

Best Regards,

Decai Yu

Hi, All,

I am new to DPD simulation and have some simple questions to ask.
What I found is that when I use bond-style harmonic to define the spring
force between beads with parameter 2.0 and 0.0, the bonded beads tend to
overlap each other, i.e., the g of r has a peak close to zero. While
non-bonded beads seem to have normal distance distribution with peak at
close to 0.85. (with rou =3)
Even I change the harmonic bond parameters to 2.0 and 0.85, I still have a
peak close to zero.
I use aii=aij=25 for every pair interactions.

Is it because that LAMMPS "possibly" neglects the repulsive pair-pair
interaction when there is a bond between the pair of beads? The conservative

please see the documentation about the special_bonds command and look
up the description of "1-2, 1-3 and 1-4 exclusions" in your preferred
text book on MD simulations. the default settings in LAMMPS are that
the special_bonds factors are 0.0 0.0 and 0.0.

force parameter (25) is much larger than bond force parameter (2*2, consider
the later 2 is energy). If the conservative force is always considered, I
don't see a reason there is a peak close to zero.

Another question, how do I know that I am running a DPD simulation? Is the
pair_style dpd definition enough to tell the program that I am running a DPD
simulation? Should I specify the atom style also as dpd?

please see the documentation of pair style DPD and compare with the
quoted publications.

atom style dpd is part of the USER-DPD package, and only required when
using styles provided in that package. please see the file
src/USER-DPD/README and the corresponding sections in the
documentation for details.

axel.

Axel,

Thanks for your quick reply.
Does that mean if I don’t specify special bonds in the file, the factors are still 0.0, 0.0 and 0.0?

Decai

yes, that’s what default means.

Also note, as Axel said, that atom_style dpd is not needed
to use pair_style dpd. That atom style is only for the USER-DPD
package, which is not standard vanilla dpd.

Steve

Steve and Axel,

Thanks so much. You guys are awesome. That solved my problems.
I struggled for ten hours to ask the right questions.
Hopefully, this also helps other users of LAMMPS DPD.

Decai