fix_ave_chunk and pppm_tip4p + respa

LAMMPS LAMMPS Mailing List Mirror
1

Dear all,

working with lammps I noticed two little things in the latest source code.

First, in fix_ave_chunk.cpp (Rev 13234), line 730/779:

I was wondering if some conversion factors for computing the temperature might be missing. Earlier, the kinetic enery is computed and then divided by DoF but force->mvv2e/force->boltz are missing.

Second, I am unsure about the compatibility between pppm_tip4p and respa:

pppm_tip4p.cpp requires newton on (see line 55), and checks the flag force->newton.
But if used in a respa level, say ilevel, with newton[ilevel]=0 respa won't comm->reverse_comm() and no error/warning is generated.

Possible fix: respa.cpp (Rev 13234) add after line 312:

2

You are correct about the missing mvv2e factor - it will be in the next patch.

I need to look at the TIP4P/rRESPA issue more closely and confer

with a colleague.

Thanks,

Steve