[Wien] inconsistency between lapw1 and mini?

[Peter Blaha]

> I think I encountered a small inconsistency between lapw1 and mini. It concerns the test for overlapping spheres.
> In one of my calculations mini accepted a structure in which the nearest-neighbor-distance between 2 inequivalent atoms was exactly equal the sum of the muffin-tin-radii of the 2 atoms:
>
> RMT(31)=2.10000 AND RMT(32)=2.10000
> SUMS TO 4.20000  LT.  NN-DIST= 4.20000
>
> So, a new scf-cycle was started, but lapw1 refused this same structure due to overlapping spheres.
> Is it possible that mini tests for sum {RMT} <= NN-distance  and lapw1 for sum {RMT} < NN-distance ?

Yes, I believe you.
As was mentioned before, mini is not in a very good shape and I have seen
these nn problems. (I guess it occurs in non-orthogonal cases ?)

lapw1 should agree with nn and should be correct.

                                      P.Blaha
--------------------------------------------------------------------------
Peter BLAHA, Inst.f. Materials Chemistry, TU Vienna, A-1060 Vienna
Phone: +43-1-58801-15671             FAX: +43-1-58801-15698
Email: blaha at theochem.tuwien.ac.at    WWW: http://info.tuwien.ac.at/theochem/
--------------------------------------------------------------------------