Thank you so much, professor. I really appreciate your scrupulous reply over and over again. In the QE code(subroutine rgd_dyn) of computing the rigid-ion (long-range) term, it is commented that "Only the G-space term is implemented: the Ewald parameter alpha must be large enough to have negligible r-space contribution". But in the following value assignment, alph= 1.0d0, not a very large number. I think such a value will not make the real-space term vanish in the Ewald summation. How to understand this?
> -----原始邮件----- > 发件人: "Lorenzo Paulatto" <[email protected]> > 发送时间: 2020-01-10 17:19:48 (星期五) > 收件人: [email protected] > 抄送: > 主题: Re: [QE-users] phonon dispersion relation from the full IFCs > > > Is it equivalent to the regular adopted Ewald summation method in > > mathematics? > > I think you forgot to attach the paper in question... > > That said, if there is not lo-to splitting, there are no effectiv > charges and no long-range interaction. Than there is no problem doing > Fourier interpolation. > > 2D is a bit special, but the QE code has special techniques to deal with > 2D phonon interpolation, I think it is explained in Phys. Rev. B 94, 085415 > > cheers > > > > > With thanks and best regards ! > > Happy New Year ! > > > > > > -- > > Jian-qi Huang > > > > Magnetism and Magnetic Materials Division > > Institute of Metal Research > > Chinese Academy of Sciences > > 72 Wenhua Road, Shenyang 110016, China > > > > email:[email protected] > > > >> -----原始邮件----- > >> 发件人: "Lorenzo Paulatto" <[email protected]> > >> 发送时间: 2020-01-09 03:56:21 (星期四) > >> 收件人: [email protected] > >> 抄送: > >> 主题: Re: [QE-users] phonon dispersion relation from the full IFCs > >> > >>> Thank you for reply, professor. I understand the regular routine > >>> implemented in QE where the long-range contribution is added in > >>> reciprocal space. My point is can I get the correct dynamical matrix > >>> just by making inverse Fourier transformation of the full(short+long) > >>> IFCs in a large real space? > >>> > >> > >> The dynamical matrix at Gamma is discontinuity with respect to the > >> points nearby, which would make any Fourier transform impossible to > >> converge. > >> > >> I think you should explain WHY you want to do this, and you may get some > >> better answer. > >> > >> In practice, if I was obliged at gunpoint, I would replace the dynamical > >> matrix file at Gamma (typically dyn1) with one computed very close to > >> Gamma, let's say q=0.001,0,0. Edit the file to trick q2r into thinking > >> that it was done at exactly Gamma, and see was comes out. > >> > >> If the material has a non-analytic term (i.e. the long range term > >> depends on the direction), this will definitely not work. Otherwise, you > >> may get something decent. > >> > >> > >> cheers > >> > >> > >> > >> -- > >> Lorenzo Paulatto - Paris > >> _______________________________________________ > >> Quantum ESPRESSO is supported by MaX (www.max-centre.eu/quantum-espresso) > >> users mailing list [email protected] > >> https://lists.quantum-espresso.org/mailman/listinfo/users > > _______________________________________________ > > Quantum ESPRESSO is supported by MaX (www.max-centre.eu/quantum-espresso) > > users mailing list [email protected] > > https://lists.quantum-espresso.org/mailman/listinfo/users > > > > -- > Lorenzo Paulatto - Paris > _______________________________________________ > Quantum ESPRESSO is supported by MaX (www.max-centre.eu/quantum-espresso) > users mailing list [email protected] > https://lists.quantum-espresso.org/mailman/listinfo/users _______________________________________________ Quantum ESPRESSO is supported by MaX (www.max-centre.eu/quantum-espresso) users mailing list [email protected] https://lists.quantum-espresso.org/mailman/listinfo/users
