I am using dpd thermostat. Using dpd pair style and fix nvt will be thermostating the system twice and it may not be right. But if I want to use the potential of DPD with a NH thermostat, will the setting of gamma and sigma parameters to zero will do? As per Warren and Groot paper this is the recipe they gave. I am not aware of whether this is OK to do in LAMMPS?
Also, LAMMPS did not issue a warning (I dont know whether it will because the particles are not time integrated but just thermostatted twice) when I used dpd and fix nvt together.
Pair dpd and fix nvt are two independent commands.
They know nothing about each other. They each
do exactly what their doc page says. Fix nvt
is an intergrator, pair dpd is not. That is shy
you do not get a warning about integrating twice.
Thanks Steve. Just one more doubt.
Please correct me if i am wrong .
(a) if i just want to use the dpd potential (ie, U® = a(1-r)^2, as I asked in the previous question and use a NH thermostat), will it just suffice if I use gamma = 0(which makes sigma zero) and then use a fix NVT since as you said they are independent and one sets the potential and other sets the thermostat.
The reason I thought was because if i turn off both random and dissipative, the probability density function for the Liouville equation becomes identical to that of a canonical ensemble with a potential, V(q) = a(1-q)^2. Now in the canonical ensemble I can maintain temperature by normal NH thermostat. Does this make sense?
Thanks Steve. Just one more doubt.
Please correct me if i am wrong .
(a) if i just want to use the dpd potential (ie, U(r) = a(1-r)^2, as I asked
in the previous question and use a NH thermostat), will it just suffice if I
use gamma = 0(which makes sigma zero) and then use a fix NVT since as you
said they are independent and one sets the potential and other sets the
thermostat.
wouldn't it be easier to just tabulate this potential and use pair style table?