Dear Ziba,probably you'll need to explain better what's your problem.> Once I'm relaxing my structure up to 0.0004 eV/Ang after the first MD step of phonon calculation > without any displacement the maximum force will be 0.002 eV/Ang "The first step without any displacement" is just calculation of your relaxed structure.If it is as you say then either it was not fully relaxed, or you changed some parametersbetween the two calculations. (A guess: either the coordinates are not those from the relaxed case,or the density matrix taken was not converged).> but still there is no negative frequency.- what you mean, "still"? And how do you know "there is no negative frequency" if you've just done "the first step without any displacement"?> This time I've relaxed my structure up to 0.0002 eV/Ang but in the first step I'm receiving 0.015 eV/Ang- see above> which is leading to the negative frequencies.- for what I know, negative frequencies are are hint that the starting point for a phonon calculationwas not at the local minimum, i.e., the relaxation was not good. How negative are your frequencies?> In the successful phonon calculations I just lowered the number of kpoint in the Monkhorst pack for phonon calculation > but this time even the kpoints are the same as relaxation input.- reducing the k-mesh is certainly not a working recipe for getting phonons right; you've just had luck.Best regardsAndrei Postnikov ----- Ziba Torkashvand <zi.torkashv...@gmail.com> a écrit : > Hello everyone,I'm trying to do phonon calculation for some hexagonal structures.Once I'm relaxing my structure up to 0.0004 eV/Ang after the first MD step of phonon calculation without any displacement the maximum force will be 0.002 eV/Ang but still there is no negative frequency.This time I've relaxed my structure up to 0.0002 eV/Ang but in the first step I'm receiving 0.015 eV/Ang which is leading to the negative frequencies.In the successful phonon calculations I just lowered the number of kpoint in the Monkhorst pack for phonon calculation but this time even the kpoints are the same as relaxation input. Any help would be appreciated ThanksZiba
-- SIESTA is supported by the Spanish Research Agency (AEI) and by the European H2020 MaX Centre of Excellence (http://www.max-centre.eu/)