[QE-users] SCF energy rise up from iteration 2

[Paolo Giannozzi]

The energy is variational, that is: it goes down as self-consistency is 
approached. The energy term that makes the energy variational is however 
approximate and becomes exact only at self-consistency. As a consequence 
it may happen that at the beginning of the self-consistency the energy 
goes up instead of going down. Nothing to worry about as long as the 
energy eventually converges (monotonically from above)

Paolo

On 18/12/2022 19:23, Rameswar Bhattacharjee wrote:
> 	
> You don't often get email from rb1820 at georgetown.edu. Learn why this is 
> important <https://aka.ms/LearnAboutSenderIdentification>
> 	
> 
> Hi Everyone,
> I am doing a single-point calculation on a 1D polymer in QE using 
> PBE-PAW. However, I am seeing an issue in the SCF iteration cycles. The 
> first iteration seems to print the right electronic energy but from 
> iteration 2 the energy suddenly increases significantly and then the SCF 
> ends up with higher energy. I am a little confused about how to 
> stabilize the SCF iteration so the energy does not go up. Any 
> suggestions will be highly appreciated.
> 
> I am providing the output of the first three iterations of the SCF for 
> deep insights.
> 
> ------------------------------------------------------------------------------
> 
> iteration #1 ecut= 120.00 Ry beta= 0.30
> 
> Davidson diagonalization with overlap
> 
> ethr =1.00E-02,avg # of iterations =2.0
> 
> 
> negative rho (up, down):1.348E-04 0.000E+00
> 
> 
> total cpu time spent up to now is 13.7 secs
> 
> 
> total energy=-304.41586797 Ry
> 
> estimated scf accuracy< 2.72932626 Ry
> 
> 
> iteration #2 ecut= 120.00 Ry beta= 0.30
> 
> Davidson diagonalization with overlap
> 
> ethr =3.79E-03,avg # of iterations =1.0
> 
> 
> negative rho (up, down):1.006E-04 0.000E+00
> 
> 
> total cpu time spent up to now is 16.3 secs
> 
> 
> total energy=-304.13553253 Ry
> 
> estimated scf accuracy< 0.76532583 Ry
> 
> 
> iteration #3 ecut= 120.00 Ry beta= 0.30
> 
> Davidson diagonalization with overlap
> 
> ethr =1.06E-03,avg # of iterations =6.0
> 
> 
> negative rho (up, down):5.328E-05 0.000E+00
> 
> 
> total cpu time spent up to now is 19.4 secs
> 
> 
> total energy=-304.26340272 Ry
> 
> estimated scf accuracy< 0.08826578 Ry
> 
> --------------------------------------------------------------------------------------------
> 
> Thanks
> Rameswar Bhattacharjee
> Georgetown University
> 
> _______________________________________________
> The Quantum ESPRESSO community stands by the Ukrainian
> people and expresses its concerns about the devastating
> effects that the Russian military offensive has on their
> country and on the free and peaceful scientific, cultural,
> and economic cooperation amongst peoples
> _______________________________________________
> Quantum ESPRESSO is supported by MaX (www.max-centre.eu)
> users mailing list users at lists.quantum-espresso.org
> https://lists.quantum-espresso.org/mailman/listinfo/users

-- 
Paolo Giannozzi, Dip. Scienze Matematiche Informatiche e Fisiche,
Univ. Udine, via delle Scienze 206, 33100 Udine Italy, +39-0432-558216