[CP2K-user] sign change of atomic virial between qs_dftb_coulomb.F and qs_dftb_matrices.F

W. Lai wei.... at gmail.com
Thu Feb 6 17:04:08 UTC 2020


Dear cp2k developers,

I have a question about the sign change of atomic virial between 
qs_dftb_coulomb.F and qs_dftb_matrices.F for 7.1 version.  In general, the 
atomic virial is -0.5*r_{ij}*f_{ij} and it seems to be the convention (i.e. 
negative sign) used in qs_dftb_matrices.F.  However, it is the opposite in 
qs_dftb_coulomb.F.

For example, lines 234 to 239 read:
               fi = 1.0_dp
               IF (iatom == jatom) fi = 0.5_dp
               CALL virial_pair_force(virial%pv_virial, fi, fij, rij)
               IF (atprop%stress) THEN
                  CALL virial_pair_force(atprop%atstress(:, :, iatom), 
fi*0.5_dp, fij, rij)
                  CALL virial_pair_force(atprop%atstress(:, :, jatom), 
fi*0.5_dp, fij, rij)
Other places in qs_dftb_coulomb.F also used the positive sign.

I am wondering if the coulombic force was using a different sign or it was 
something else I am missing.

Thanks, Wei
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20200206/a35e78c9/attachment.htm>


More information about the CP2K-user mailing list