I am trying to mimic the experimental annealing of polymer using (lammps) DPD simulation , which requires the system being heated at different temperatures. There, I am little stuck and hence, am wondering if you can provide little insight on the following yellow colored parts regarding the green colored options.

Temperature control :

1) pair_style dpd 1.0 1.0 1234567

This is a style how dpd thermostat is used as input in lammps and here kbT/Epsilon (Kb = Boltzman constant and Epsilon = energy) = 1 provides the equivalent temperature of 300K. However, I could not find the source/basis of energy that was used here. I am wondering if I can use the same energy in order to obtain this non-dimensional value for higher temperature.

2) thermo 1000

I found that people used all different numbers like 100, 1000, 2000 for this value. But, I could not detect any explanation with the conversion of these numbers into any real value or what might be any other significance of these values.

Mentioning 'fix NVE' in lammps actually provides NVT DPD :

3) fix 3 all nve

DPD simulation is intrinsically an NVT system (ref- Groot & Warren). However, in lammps we actually use 'fix nve' ( command like # 3 ) in order to obtain fixed temperature simulation. Scavenging throughout the lammps mailing list, I found that they said- in lammps, the temperature is obtained from the kinetic energy and here, 'fix nve' dictates the use of a constant kinetic energy. That's how the system uses this command to dictate the constant temperature. People who tried to use directly 'fix nvt' - reported issues with inconsistent results.

It will be much appreciated if you can provide any insight on this regard or have something to correct me.

Sincerely,

Tuhin