Hi, everyone, I have a question about the reduced density in lj units.

In lj units, density = mass/volume, where rho* = rho sigma^dim, so if the density is 4.2e3 kg/m3, the diameter of particles is 2e-8 m, so the reduced density is 3.36e-20 kg. It looks weird. So i am not sure whether the ruduced progress is right in lammps. May be rho* should equal to rho sigma^dim /m. Can you give me some suggestions? Best wishes!

Donâ€™t use reduced units if you want real parameters.

Hi, Axel, Thanks very much. I used the lennard jones potential in my brownian dymaics simulation. Can i use the real units in my simulations. Best wishes!

Hi, Axel, Thanks very much. I used the lennard jones potential in my

brownian dymaics simulation. Can i use the real units in my simulations.

you can do correct MD simulations in *any* set of units for as long as

all you input is consistent. that goes for your brownian dynamics just

as well. LAMMPS doesn't care; it will run the computations you ask

for. however, it is not always obvious to people that there are

connections between units. for example, you can choose energy, length

and mass, but that determines the unit of time.

apart from those general consideration, i would also like to caution

that the LJ potential is a rather soft interaction corresponding to

the softness of atoms and thus may not be that good a representation

of larger particles comprised of multiple atoms.

axel.

apart from those general consideration, i would also like to caution

that the LJ potential is a rather soft interaction corresponding to

the softness of atoms and thus may not be that good a representation

of larger particles comprised of multiple atoms.

Axel, thanks very much for your caution. I want to study the sedimentation process of nanoparticles by brownian dynamics simulation. And i need to consider the particle-particle interactions, if the LJ potential is not suitable for my situation, can you give me some suggestions. My model includes 10000 titanium dioxide nanoparticles, which diameters are all of 20 nm. Best wishes!

I suggest you find a paper in the literature that

has modeled a system like this and see what they

used, then how to mimic their model in LAMMPS.

Presumably you want to treat these as large,

coarse-grained single particles. LAMMPS has

many CG options, e.g. see the COLLOID package.

But to use the coarse-grained potentials, you

need to know what you are doing. The literature

is the best place to start.

Steve

Hi, Steve, thanks very much! Best wishes!