compute chunk/atom for molecule COM

Dear Admin,

Is there a way to do a compute chunk/atom such that I can bin each molecule’s COM and subsequently use compute vcm/chunk together with fix ave/chunk to get the COM velocity gradient of molecular fluid flow?

I thought of using something like:

# Chunk atoms into molecules
compute cc1 all chunk/atom molecule

# Compute COM
compute comChunk all com/chunk cc1

# Bin by COM
compute binCOM all chunk/atom compute c_comChunk z lower 0.02 units reduced

# Compute VCM
compute vcmChunk all vcm/chunk cc1

# Average VCM
fix 1 all ave/chunk 10 100 1000 binCOM c_vcmChunk

Would this work at all considering I am not using per-atom vectors, or is there a way to work around this?

Best Regards,
Jingjie

**Disclaimer** The sender of this email does not represent Nanyang Technological University and this email does not express the views or opinions of the University.

Dear Admin,

Is there a way to do a compute chunk/atom such that I can bin each
molecule’s COM and subsequently use compute vcm/chunk together with fix
ave/chunk to get the COM velocity gradient of molecular fluid flow?

I thought of using something like:

# Chunk atoms into molecules
compute cc1 all chunk/atom molecule

# Compute COM
compute comChunk all com/chunk cc1

# Bin by COM
compute binCOM all chunk/atom compute c_comChunk z lower 0.02 units reduced

# Compute VCM
compute vcmChunk all vcm/chunk cc1

# Average VCM
fix 1 all ave/chunk 10 100 1000 binCOM c_vcmChunk

Would this work at all considering I am not using per-atom vectors, or is
there a way to work around this?

​please set up a simple test system, give it a try and check it out.

axel.​

Dear Axel,

I did give it a spin and it generated the error of “Illegal compute chunk/atom command” at the line “compute binCOM all chunk/atom”, I assumed it was because of not having per-atom vectors when I tried to bin the COMs. Is there a better way to do this?

Best,
JJ

Disclaimer The sender of this email does not represent Nanyang Technological University and this email does not express the views or opinions of the University.

Dear Admin,

Is there a way to do a compute chunk/atom such that I can bin each
molecule’s COM and subsequently use compute vcm/chunk together with fix
ave/chunk to get the COM velocity gradient of molecular fluid flow?

I thought of using something like:

# Chunk atoms into molecules
compute cc1 all chunk/atom molecule

# Compute COM
compute comChunk all com/chunk cc1

# Bin by COM
compute binCOM all chunk/atom compute c_comChunk z lower 0.02 units
reduced

# Compute VCM
compute vcmChunk all vcm/chunk cc1

# Average VCM
fix 1 all ave/chunk 10 100 1000 binCOM c_vcmChunk

Would this work at all considering I am not using per-atom vectors, or is
there a way to work around this?

​please set up a simple test system, give it a try and check it out.

axel.​

Dear Axel,

I did give it a spin and it generated the error of "Illegal compute
chunk/atom command" at the line "compute binCOM all chunk/atom", I assumed
it was because of not having per-atom vectors when I tried to bin the COMs.
Is there a better way to do this?

​i think, you are trying something that is too complex. why not just do a

compute chunk/atom bin1d z lower ...​

and then just use fix ave/chunk over per atom velocities. that should give
you your desired velocity profile.
when you compute vcm, you are essentially averaging over all atoms in the
molecule, so there isn't that much of a difference.

axel.