Lily Anh wrote: > Thanks for your suggestion. > I have changed ecutwfc (from 20-50Ry every 5 Ry), ecutrho (200-1200 > every 50), k grid. But nothing works. How to change nr1 nr2 nr3, > larger or smaller? nr1, nr2 and nr3 are input parameters in the &control namelist that usually take a default value on the basis of ecutrho but can be specified directly when needed (as in this case) . > The structure should have 4 symmetry operations while the code can not > find any symmetry. I want to reduce the calculation time. the code finds correctly all 4 symmetry operations but discards them (except the trivial identity) because of the incompatibility of the required fractional translation with the FFT grid. By choosing a different origin for your system it could be possible to make the fractional translations simple fractions of the fundamental lattice vectors.
Hope it helps, stefano
