Reg: Diffusion coefficient of nano-particle and fluid atoms in nanofluid

Dear All,

To capture the physics in the enhanced thermal conductivity of nanofluid (Copper and water), I am interested in calculating the Diffusion coefficient of nanoparticle (Copper) and Self diffusion coefficient of water molecules. For this I divide the system in two groups i.e. group Cu and group water. I have put the group id to compute msd which is shown below. The problem I am facing that while calculating MSD of both the groups by the following command, the values are coming same, i.e. the graph between MSD vs time is same for both solid nanoparticle as well as liquid atoms. I was expecting a very low slope for solid particle with comparison to the liquid particle. Please suggest me the variation in the following code.

compute msd Cu msd com yes
fix msd Cu ave/time 2 100 1000 c_msd[4] file Cu.msd

compute msd1 water msd com yes
fix msd1 water ave/time 2 100 1000 c_msd[4] file wat.msd

thermo 1000
dump 3 all xyz 100 dump.xyz
timestep 1
thermo_style custom step etotal temp vol press density
run 200000

Dear All,

To capture the physics in the enhanced thermal conductivity of nanofluid
(Copper and water), I am interested in calculating the Diffusion
coefficient of nanoparticle (Copper) and Self diffusion coefficient of
water molecules. For this I divide the system in two groups i.e. group Cu
and group water. I have put the group id to compute msd which is shown
below. The problem I am facing that while calculating MSD of both the
groups by the following command, the values are coming same, i.e. the graph
between MSD vs time is same for both solid nanoparticle as well as liquid
atoms. I was expecting a very low slope for solid particle with comparison
to the liquid particle.


​why were you expecting this? have you looked up in publications what order
of magnitude MSD ​you should get for nanoparticles of the size you have?
have you plotted the MSD? are you certain that your results are converged?

Please suggest me the variation in the following code.

that is the wrong question. research with simulations does not work by
manipulating your input until you get the result you want. that would be
doing an animation and that would only make sense, if you plan to work for
pixar in the future. so if you don't get the expected result you have to
*understand* why this is the case and whether your expectations are wrong
or the simulation (or both). after that it should be obvious to yourself
what steps you have to do.

axel.