J. K. Dewhurst

Hi, If reducebf < 1 then bfieldc and bfcmt are automatically reduced to zero for all tasks except for 0,2,28,200,208,350,380,420,421,440. In other words, if reducebf < 1 then you don't need to do anything if you want zero fields (i.e. when they are just for symmetry breaking). If the fields are supposed to be finite (like an external magnetic field) then reducebf should be set to exactly 1 and the RT-TDDFT calculation will include them. Regards, Kay.

Hi, You're quite correct: the formula on page 75 is for an earlier version, and also doesn't have the chirp part. I'll fix that for the next release. Thanks and regards, Kay.

Hi Chamara, It's best not to enable both nxoapwlo=1 and lorbcnd=.true. as this can lead to instability in the first-variational eigenvalue problem. Also having rgkmax too high can cause similar instabilities. Here's my input file that works for the structure at high compression: tasks 0 xctype 3 vhighq .true. ngridk 18 18 18 stype 0 swidth 0.001 rgkmax 8.5 gmaxvr 32.0 lmaxi 2 nxoapwlo 1 lorbcnd .false. avec 5.058464396264 0.000000000000 0.000000000000 -0.000000000000 5.058464396264 0.000000000000...

(continued) You can also determine the pressure with task=440 although this is not quite as accurate as fitting to an EOS. Here is the input file I used: tasks 440 wrtdisk .false. deltast 0.001 xctype 3 vhighq .true. lorbcnd .false. nxoapwlo 1 rgkmax 10.0 gmaxvr 32.0 lmaxi 2 ngridk 24 24 24 avec 7.818401918226 0.000000000000 0.000000000000 0.000000000000 7.818401918226 0.000000000000 0.000000000000 0.000000000000 7.818401918226 atoms 2 : nspecies 'Mg.in' : spfname 4 : natoms 0.0 0.0 0.0 0.0 0.5 0.5...

Hi Chamara, This type of calculation has to be converged very well to get reliable pressures. The first and most reliable way of performing the calculation is to calculate an energy-vs-volume curve and fit it to an equation of state (EOS). Here is the input file I used for this: tasks 0 1 1 1 1 1 1 1 1 1 1 batch .true. wrtdisk .false. trmt0 .false. xctype 3 vhighq .true. lorbcnd .false. nxoapwlo 1 rgkmax 10.0 gmaxvr 32.0 lmaxi 2 ngridk 24 24 24 avec 7.818401918226 0.000000000000 0.000000000000 0.01...

Dear Long and Michael, This is quite difficult to converge. I think Michael is correct using mixtype=1 and rgkmax=7.5. I use task=28 to calculate the MAE. This is because it also rotates the spherical covering used for the spherical harmonic transform which eliminates some error. Here is my input file: tasks 28 npmae -1 mixsave .true. wrtdisk .false. epsengy 1.e-6 epspot 1.e-8 spinorb .true. bfieldc 0.0 0.0 0.5 reducebf 0.85 ngridk 24 24 24 rgkmax 7.5 gmaxvr 18.0 lmaxapw 9 lmaxo 8 xctype 20 mixtype...

Elk version 10.9.5 released

Elk version 10.8.16 released