damp fix press/berendsen and fix langevin

Dear all,

I have several questions related to fix press/berendsen command and fix langevin command.

1. In fix press/berendsen command, we need to specify Pdamp and modulus value. From lammps manual of 2016Nov17, the recommneded value for Pdamp is 1000 timesteps, but I don’t know what modulus mean in this command? Do we need to specify modulus value according to the real modulus of the material we simulate? Or we just need to consider the product of Pdamp and modulus to make pressure relax in a reasonable rate? If the latter one is correct, is there a recommend value for modulus?

2. In fix langevin command,

(1) how to specify Tdamp? I don’t find a recommended value for it.

(2) Especially how to specify scale values? I know that the mass and size of different atom types should be accounted for in the choice of ratio values, but how? Suppose I define two atom types in my simulation box. Condtion1, for type1, mass=1, size=1, and type2, mass=1, size=2, what are sacle values for type1 and type2?

(3) Condition2, for type1, mass=1, size=1, and type2, mass=2, size=1, what are sacle values for type1 and type2?

(4) Condition3, for type1, mass=1, size=2, and type2, mass=4, size=1, what are sacle values for type1 and type2?

Thanks and look forward to your reply.

Rui

Dear all,

I have several questions related to fix press/berendsen command and fix
langevin command.

1. In fix press/berendsen command, we need to specify Pdamp and modulus
value. From lammps manual of 2016Nov17, the recommneded value for Pdamp is
1000 timesteps, but I don't know what modulus mean in this command? Do we
need to specify modulus value according to the real modulus of the material
we simulate? Or we just need to consider the product of Pdamp and modulus to
make pressure relax in a reasonable rate? If the latter one is correct, is
there a recommend value for modulus?

please have a more careful look at the documentation for fix

2. In fix langevin command,

(1) how to specify Tdamp? I don't find a recommended value for it.

that is impossible to provide, because it depends on how you are using
fix langevin. you need a different Tdamp when you use fix langevin an
explicit solvent system or implicit solvent (and different solvents to
boot).

(2) Especially how to specify scale values? I know that the mass and size of
different atom types should be accounted for in the choice of ratio values,
but how? Suppose I define two atom types in my simulation box. Condtion1,
for type1, mass=1, size=1, and type2, mass=1, size=2, what are sacle values
for type1 and type2?

again, this strongly depends on the kind of simulation you are doing,
you may not need to set it at all.
at any rate, you need to read the documentation carefully, if you have
trouble understanding the details, you probably also need to read up
on langevin dynamics in general.

please note, that the LAMMPS manual is a "technical documentation",
i.e. it will tell you what the commands and keywords stand for, but
its purpose is not to explain the physics or tell you exactly how to
do your research (and neither is it the purpose of the mailing list).
think of it being like the owner's manual of a car: it tells you where
all the switches, buttons and levers are, but you still have to go to
driving school to learn how to drive.

axel.

Dear Axel,

Thanks a lot for your reply. Every explanation from you helps me to understand commands themselves or to know the direction I should go. Thank you very much.

As for modulus keyword in fix press/berendsen, from lammps manual, the bulk modulus has units of pressure and is the amount of pressure that would need to be applied (isotropically) to reduce the volume of the system by a factor of 2 (assuming the bulk modulus was a constant, independent of density, which it’s not). This means we can’t get an accurate modulus value easily since it’s not a good choice to half the volume of the system and see what the pressure is. Combined with fix nve and fix langevin, I tried two groups of Pdamp and modulus in fix press/berendsen. In the first group, Pdampmodulus=100, while in the seconde group Pdampmodulus=1000. I found that in both groups, pressures reach their average values quickly. The difference is that average value in the first condition is far from the target pressure specified in fix press/berendsen and pressure fluctuates widely, while in the seconde condition, average pressure is very close to the target value and fluctuation is very small. Pdamp should only affect the time when pressure reaches to the target value and the extent of pressure fluctuation , why is the pressure in the first condition far from the target value and doesn’t show any trend to close to the target one?

Thanks for your suggestion. I need to read some materials on langevin dynamics in general to have a better understanding in this command.

Thanks again and look forward to your reply.

Best wishes,

Rui

Dear Axel,

Thanks a lot for your reply. Every explanation from you helps me to
understand commands themselves or to know the direction I should go. Thank
you very much.

As for modulus keyword in fix press/berendsen, from lammps manual, the bulk
modulus has units of pressure and is the amount of pressure that would need
to be applied (isotropically) to reduce the volume of the system by a factor
of 2 (assuming the bulk modulus was a constant, independent of density,
which it's not). This means we can't get an accurate modulus value easily
since it's not a good choice to half the volume of the system and see what
the pressure is. Combined with fix nve and fix langevin, I tried two groups
of Pdamp and modulus in fix press/berendsen. In the first group,
Pdamp*modulus=100, while in the seconde group Pdamp*modulus=1000. I found
that in both groups, pressures reach their average values quickly. The
difference is that average value in the first condition is far from the
target pressure specified in fix press/berendsen and pressure fluctuates
widely, while in the seconde condition, average pressure is very close to
the target value and fluctuation is very small. Pdamp should only affect the
time when pressure reaches to the target value and the extent of pressure
fluctuation , why is the pressure in the first condition far from the target
value and doesn't show any trend to close to the target one?

your expectations would only apply to an ideal gas. for real systems,
and particularly dense system, things are much more complicated and if
you choose a bad combination of parameters, you can, for example, get
a "feedback loop".

specifically, if the time constant is small, the volume adjustments
can be too large, and since it is always easier to quickly expand a
dense system than to rapidly compress a dilute system back to its
dense state at a given average pressure, the average can be easily
off.

axel.