Msd of rigid bodies

Dear all,

I am simulating CO2 molecules in water and I want to calculate the diffusivity. I am using rigid/npt to have both the waters and CO2s treated as rigid bodies. I read in the documentation that the msd calculation cannot be used with rigid because of the altered image flags. I am trying to dump somehow the “real” image flags of the molecules in order to create the “outbox coordinates” and calculate the msd but I cannot. I use

Fix npt all rigid/npt molecule temp t1 t2 td iso p1 p2 pd

Fix nptdump all ave/time 1 1 1000 f_npt file npt.dat

But the npt.dat has only the ts and the cumulative energy, as manual says.

I used also

Dump configFileU all custom tdump configurationFile.dat id type xu yu zu ix iy iz

But the image flags are altered as mentioned in the manual. How can I have access to the ix iy iz of rigid bodies?

Best,

Otto

Dear all,

I am simulating CO2 molecules in water and I want to calculate the
diffusivity. I am using rigid/npt to have both the waters and CO2s treated
as rigid bodies. I read in the documentation that the msd calculation cannot
be used with rigid because of the altered image flags. I am trying to dump
somehow the “real” image flags of the molecules in order to create the
“outbox coordinates” and calculate the msd but I cannot. I use

Fix npt all rigid/npt molecule temp t1 t2 td iso p1 p2 pd

Fix nptdump all ave/time 1 1 1000 f_npt file npt.dat

But the npt.dat has only the ts and the cumulative energy, as manual says.

...and why don't you use fix ave/time in vector mode then?

axel.

But the image flags are altered as mentioned in the manual. How can I have access to the ix iy iz of rigid bodies?

They are altered in the dump file, but only slightly. You should be able to
use them to calculate an MSD with a bit of attention to the periodic box.
You can use the image flags of atoms in a molecule to compute its COM
(they are consistent, if unwrapped you have the molecule).

That COM will always be inside the simulation box. So if you track it,
you will know when it crosses a periodic boundary and can adjust its
unwrapped coords for computing the MSD.

Steve