ave/chunk and pe/atom

Dear all,

The documentation says compute pe/atom yields the per-atom potential energy (as a per-atom vector). But what exactly is the scalar value printed in the output file when one uses fix chunk/atom as follows?

compute myPE all pe/atom
fix T_slices1 all ave/chunk 100 10 1000 slices1 c_myPE file slices1_equi.prof

Thank you.

Dear all,

The documentation says compute pe/atom yields the per-atom potential
energy (as a per-atom vector). But what exactly is the scalar value
printed in the output file when one uses fix chunk/atom as follows?

the energy of the atoms in each chunk averaged over the chunk and over
the selected time slices.

axel.

See the first lines of the fix ave/chunk doc page:

Use one or more per-atom vectors as inputs every few timesteps, sum the values over the atoms in each chunk at each timestep, then average the per-chunk values over longer timescales. The resulting chunk averages can be used by other output commands such as thermo_style custom, and can also be written to a file.

and also the section on the “norm” keyword. It affects whether these per-chunk

sums are normalized by the number of atoms in the chunk or not.

Steve

Thank you for your answers. Steve, I looked into the norm keyword and it seems only “all” and “sample” are working, “none” produces an error (as mentioned in an older discussion). Anyway I suppose using sample or all and multiplying by the average atom count should give me the chunk’s potential energy.

Jonathan

I fixed this, so that norm none should now work. Will be

in the next patch, later today.

Thanks,

Steve