Dear ELK developers, Apparently, whenever a systam lacks inversion symmetry the code chooses Complex Hermitian eigensolver. Attached I have an example with this situation. It's a bilayer system and lacks inversion symmetry. The code hardly converges after almost 3000 scf loop (Nonetheless, the monolayer and bulk structures converge superfast and the code chooses Real symmetric eigensolver). I want to run the code with Hubbard U and in this case it look it is not going to converge at all even with...
Dear ELK developers, Apparently, whenever a systam lacks inversion symmetry the code chooses Complex Hermitian eigensolver. Attached I have an example with this situation. It's a bilayer system and lacks inversion symmetry. The code hardly converges after almost 3000 scf loop (Nonetheless, the monolayer and bulk structures converge superfast). I want to run the code with Hubbard U and in this case it look it is not going to converge at all even with small mixing parameters. Could it be related to...
Dr. Sharma, I see, so it is included in the 2015 paper but excluded in the 2022 paper. Thank you for the clarification. So, is there any published older version of Elk that comes with this term included? Best Regard, AK
Dr. Sharma, Here is an example that says the Hamiltonian includes the B-field of the laser pulse: https://pubs.acs.org/doi/10.1021/acs.jctc.5b00621 And here is an example that the Hamiltonian does not include it: https://journals.aps.org/prb/abstract/10.1103/PhysRevB.105.134425 Best Regards, AK
Hi Dr. Sharma, Thank you for the clarification and valuable information. Users' participating in the forum and helping others is a nice idea. Regarding the B-field, I noticed in some of your earlier publications it was mentioned explicitly and in some other it was not mentioned at all. I could not find any trace of it in the code and I just wanted to be sure about it. Thanks a lot for the clarification. Best Regards, AK
Why in this forum some questions are answered and some questions are just ignored?! Are there any standards that must be followed when questioning?
Hi, Is the magnetic field of the laser pulse included in the time-dependent Hamiltonian? If yes, would you please let me know the related source files? Thanks! Ali
Non-relativistic spin-splitting does not occur without DFT+U for RuO2. Try this: dft+u 1 1 : dftu, inpdftu=FLL 1 2 0.08084844 0.01469972 : is =Ru, l=2, U=2.20eV, J=0.40eV
Non-relativistic spin-splitting does not occur without DFT+U for RuO2.
Thank you Kay!
Hi, The charge density and magnetization and many other time-dependent quantities are discontinuous after the restart. You can try it with any simple example. Regards, A.K.
Hi, I am trying to get the ground state of a heterostructure of Ni(100) and Pt(100) with noncollinear DFT calculation. The code does not converge while the initial magnetic moments are given from the obtained results from another DFT code. Here I use 20 empty states. I increased it to 30 but could not get convergence. Does anybody have any suggestion? Thanks, Best Regards, AK
Dear Elk community, The 2d charge density of magnetic materials with magnetic moment along z-direction in TDDFT calculations has singularity. I tried for different materials and I increased the accuracy of the calculations; but the singularity did not go away. Bests, AK
Hi, It is <m2,jspn|rho|m1,ispn>. Bests, AK
Hi, It is<m2,jspn|rho|m1,ispn>. Bests, AK
Hi, It is <m2,jspn|rho|m1,ispn>.</m2,jspn|rho|m1,ispn> Bests, AK
Hi, It is <m2, jspn|rho|m1,ispn="">. </m2,> Bests, AK
Hi, It is <m2, jspn|rho|m1,ispn=""> .</m2,> Bests, AK
Hi, It is <m2, jspn|rho|m1,ispn="">.</m2,> Bests, AK
Hi Kay, I did some calculation to see how does the 2D charge density look like for in-plane magnetization. I ran the two attached codes for one single layer and a bilayer of Mn3Sn . Then, I summed the results for each time step (sum over a 20 by 20 grid) and plotted them as a function of time. The plot is attached. Do these results make sence? Bests, AK
Hi Kay, I did some calculation to see how does the 2D charge density look like for in-plane magnetization. I ran the two attached codes for one single layer and a bilayer of Mn3Sn . Then, I summed the results for each time step (sum over a 20 by 20 mesh grid) and plotted them as a function of time. The plot is attached. Do these results make sence? Bests, AK
Hi Kay, I did some calculation to see how does the 2D charge density look like for in-plane magnetization. I ran the two attached codes for one single layer and a bilayer of Mn3Sn . Then, I summed the results for each time step and plotted them as a function of time. The plot is attached. Do these results make sence? Bests, AK
Hi, Here is the code. I attached a sample of output file. The large numbers are present in all output files at the same place. Best, AK !Code tasks 0 450 460 spinpol .true. spinorb .true. ! number of empty states nempty 20 ! use Broyden mixing mixtype 3 tshift .false. ! total simulation time tstime 2000.0 ! Simulation time step dtimes 0.1 ! laser pulse parameters pulse 15.0 0.0 0.0 0.05694 90.0 0.0 800.0 500.0 sppath '/home/elk-8.8.26/species/' scale 1.88972612462577 avec 3.524900000 0.000000000...
Dear Elk community, The 2d charge density of magnetic materials with magnetic moment along z-direction in TDDFT calculations has singularity. I tired for different materials and I increased the accuracy of the calculations; but the singularity did not go away. Bests, AK
Dear Elk community, The time-dependent 2d charge density of magnetic materials with magnetic moment along z-direction has singularity. I tired for different materials and I increased the accuracy of the calculations; but the singularity did not go away. Bests, AK
Dear Elk community, The 2d charge density of magnetic materials with magnetic moment along z-direction has singularity. I tired for different materials and I increased the accuracy of the calculations; but the singularity did not go away. Bests, AK
Hi, I like to calculate the ground state of a two-dimensional structure. When I increase the vertical lattice vector to make it 2D like I notice the number of G-vector along kz is increased very much. The G-vector grid I get is 72 72 192 that gives 396829 G-vectors. It slows down the calculation severely. Is there any efficient way to control the G-vector grid size? Thanks, AK