[Pw_forum] dE0s is positive which should never happen

[Paolo Giannozzi]

On Sat, 2013-05-11 at 00:37 -0400, Alexandru Bogdan Georgescu wrote:

>      from bfgs : error #         1
>      dE0s is positive which should never happen

I have never seen (and never heard about) such an error message.
Problems in the BFGS algorithm typically signal a mismatch between
forces and energy. This typically happens very close to the minimum,
due to insufficient scf convergence (forces are small, scf error is
significant), as determined by parameters "conv_thr" and "upscale".

> I'm sorry if my question is silly

it is not silly, but it doesn't contain enough data to produce a
better answer

P.

-- 
Paolo Giannozzi, Dept. Chemistry&Physics&Environment, 
Univ. Udine, via delle Scienze 208, 33100 Udine, Italy
Phone +39-0432-558216, fax +39-0432-558222