Dear Professors Paolo and and Hari, I appreciate your both help. However, I am still a little bit confused which weights should I chose for lambda.x file. For example: bravais-lattice index = 7 lattice parameter (alat) = 7.9506 a.u. unit-cell volume = 591.7416 (a.u.)^3 celldm(1)= 7.950626 celldm(2)= 0.000000 celldm(3)= 2.354822 celldm(4)= 0.000000 celldm(5)= 0.000000 celldm(6)= 0.000000 scf calculations on the mesh 4x4x4 give me: number of k points= 13 Marzari-Vanderbilt smearing, width (Ry)= 0.0200 cart. coord. in units 2pi/alat k( 1) = ( 0.0000000 0.0000000 0.0000000), wk = 0.0312500 = 2 k( 2) = ( -0.2500000 0.0000000 0.1061651), wk = 0.2500000 = 16 k( 3) = ( 0.5000000 0.0000000 -0.2123303), wk = 0.1250000 = 8 k( 4) = ( -0.2500000 0.2500000 0.2123303), wk = 0.2500000 = 16 k( 5) = ( 0.5000000 0.2500000 -0.1061651), wk = 0.5000000 = 32 k( 6) = ( 0.2500000 0.2500000 0.0000000), wk = 0.1250000 = 8 k( 7) = ( 0.5000000 -0.5000000 -0.4246605), wk = 0.0625000 = 4 k( 8) = ( 0.0000000 0.0000000 0.2123303), wk = 0.0625000 = 4 k( 9) = ( 0.7500000 0.0000000 -0.1061651), wk = 0.2500000 = 16 k( 10) = ( 0.5000000 0.0000000 0.0000000), wk = 0.1250000 = 8 k( 11) = ( 0.7500000 -0.7500000 -0.4246605), wk = 0.1250000 = 8 k( 12) = ( 0.5000000 -0.5000000 -0.2123303), wk = 0.0625000 = 4 k( 13) = ( 0.0000000 0.0000000 -0.4246605), wk = 0.0312500 = 2 (there is no inversion in crystal structure) ph.x calculations are following: Dynamical matrices for ( 4, 4, 4) uniform grid of q-points ( 13q-points): N xq(1) xq(2) xq(3) 1 0.000000000 0.000000000 0.000000000 2 -0.250000000 0.000000000 0.106165127 3 0.500000000 -0.000000000 -0.212330253 4 -0.250000000 0.250000000 0.212330253 5 0.500000000 0.250000000 -0.106165127 6 0.250000000 0.250000000 0.000000000 7 0.500000000 -0.500000000 -0.424660507 8 0.000000000 0.000000000 0.212330253 9 0.750000000 -0.000000000 -0.106165127 10 0.500000000 -0.000000000 0.000000000 11 0.750000000 -0.750000000 -0.424660507 12 0.500000000 -0.500000000 -0.212330253 13 0.000000000 -0.000000000 -0.424660507 Therefore, we can say that both q- and k-meshes are "exactly" the same in scf and ph calculations. However, when I generate k-mesh using kpoints.x, the set is equivalent, but the weights and order are different: *************************************************** * * * Welcome to the special points world! * *________________________________________________ * * 1 = cubic p (sc ) 8 = orthor p (so ) * * 2 = cubic f (fcc) 9 = orthor base-cent. * * 3 = cubic i (bcc) 10 = orthor face-cent. * * 4 = hex & trig p 11 = orthor body-cent. * * 5 = trigonal r 12 = monoclinic p * * 6 = tetrag p (st ) 13 = monocl base-cent. * * 7 = tetrag i (bct) 14 = triclinic p * *************************************************** bravais lattice >> 7 filout [mesh_k] >> TEST enter celldm(3) >> 2.35482 mesh: n1 n2 n3 >> 4 4 4 mesh: k1 k2 k3 (0 no shift, 1 shifted) >> 0 0 0 write all k? [f] >> # of k-points == 13 of 64 13 1 0.0000000 0.0000000 0.0000000 1.00 2 0.2500000 -0.2500000 0.0000000 4.00 3 0.5000000 -0.5000000 0.0000000 2.00 4 0.0000000 0.2500000 0.1061652 8.00 5 0.5000000 -0.2500000 0.1061652 16.00 6 0.0000000 0.5000000 0.2123305 4.00 7 0.2500000 0.2500000 0.2123305 8.00 8 0.0000000 0.0000000 0.2123305 2.00 9 0.5000000 -0.5000000 0.2123305 2.00 10 0.0000000 0.2500000 0.3184957 8.00 11 0.0000000 0.5000000 0.4246609 4.00 12 0.2500000 0.2500000 0.4246609 4.00 13 0.0000000 0.0000000 0.4246609 1.00 Which weights should I use in my el-ph calculations?
-- with regards Arena Konta The Institute of Thermophysics in Novosibirsk Scientific Center _______________________________________________ users mailing list users@lists.quantum-espresso.org https://lists.quantum-espresso.org/mailman/listinfo/users