This is the user forum of the DFT FLAPW code FLEUR. It is meant to be used as a place to asks questions to other users and the developers, to provide feedback and suggestions. The documentation of the code can be found at the FLEUR homepage.
Thank your very much for your valuable time and quick reply. The first iteration is finished, but the charge density is too big and seems unreasonable.
l_ss=F,l_mperp=F,l_constr=F,l_disp=F,sso_opt=FFT ----> distance of charge densities for spin 1 it= 1: 115.425684 me/bohr**3 ----> distance of charge densities for spin 2 it= 1: 95.852274 me/bohr**3 ----> distance of charge densities for spin 3 it= 1: 0.828353 me/bohr**3 ----> distance of charge densities for it= 1: 205.064444 me/bohr**3 ----> distance of spin densities for it= 1: 54.498328 me/bohr**3
Here is vca.in file. ################################## 1 0.6 2 0.6 3 0.6 4 0.6 5 0.6 6 0.6 ##################################
After finishing a electronic step, an inwint error has occured.
now 861 instead of 861 now 862 instead of 862 now 863 instead of 863 now 864 instead of 864 inwint forrtl: error (69): process interrupted (SIGINT) Image PC Routine Line Source fleur.x 0000000000AEEB55 Unknown Unknown Unknown fleur.x 0000000000AEC777 Unknown Unknown Unknown fleur.x 0000000000A88224 Unknown Unknown Unknown fleur.x 0000000000A88036 Unknown Unknown Unknown fleur.x 0000000000A0DE76 Unknown Unknown Unknown fleur.x 0000000000A1474E Unknown Unknown Unknown libpthread.so.0 00002ABCDCA01630 Unknown Unknown Unknown
What's wrong with my calculation ? Should be noted that this error occurred for some doping [e.g. the above system with x = 0.6, or x = others], but not for x = 0, 0.1,0.2 and so on. Thank you very much.
I checked the internal -> cartesian of real-space (A) and reciprocal space (B). The A in out file should be the transpose of true real space, while the B is the true reciprocal space. I read the source code of cotra.f. The internal -> cartesian is wrote as :
The real space: ! ********************************************************** SUBROUTINE cotra0(r,x,amat) ! ********************************************************** ! .. ! .. Array Arguments .. REAL, INTENT (IN) :: r(3) REAL, INTENT (IN) :: amat(3,3) REAL, INTENT (OUT):: x(3) ! DO i = 1,3 s = 0.0 DO j = 1,3 s = s + amat(i,j)*r(j) ENDDO x(i) = s ENDDO
I want to know, Is the direction of spiral B*q or q*B ? For fcc, B*q or q*B is the same, while they are different for some lattice. For example, a rhombohedral primitive cell.
If I input the structure as follows, ############################################## 2.4387895018949948 1.4080357754159085 4.7429872621175369 -2.4387895018949948 1.4080357754159085 4.7429872621175369 0.0000000000000000 -2.8160715508318170 4.7429872621175369 1.889726125 1 1 1 ###############################################
The out file will give bravais matrices of real and reciprocal lattices, 4.608644 -4.608644 0.000000 ---- 0.681674 0.393565 0.233673 2.660802 2.660802 -5.321604 ---- -0.681674 0.393565 0.233673 8.962947 8.962947 8.962947 ---- 0.000000 -0.787129 0.233673
Then if spiral direction is along the z direction For B*q, the q will be (0.0000000000 -0.8469595000 1.4264960700) That is ( 0.681674 0.393565 0.233673) ---- (0.0000000000) ---- (0) (-0.681674 0.393565 0.233673) ---- (-0.8469595000) = (0) (0.000000 -0.787129 0.233673) ---- (1.4264960700) ---- (z)
For q*B, the q will be (1,1,1). -------------( 0.681674 0.393565 0.233673) ----(1 1 1) (-0.681674 0.393565 0.233673)--------- =(0 0 z) ------------(0.000000 -0.787129 0.233673)
I want to do a spin spiral calculation involving two magnetic atoms (fleur, v26 version). I have reproduced the result of fcc Fe (one magnetic atom), but I do not know if there is any difference between one magnetic atom and more magnetic atoms. The following are my material system (two magnetic Pd atoms) and my calculation progress.
This is my input generator. ################################################################# PdF3
Here are the bravais matrices of real and reciprocal lattices from "out" file, ################################################################## 4.608644 -4.608644 0.000000 0.681674 0.393565 0.233673 2.660802 2.660802 -5.321604 -0.681674 0.393565 0.233673 8.962947 8.962947 8.962947 0.000000 -0.787129 0.233673 ##################################################################
Then I define the q vector = (0.0, -0.8469595, 1.42649607), which will describe a spin spiral in z-direction [reciprocal lattice * q = (0,0,1)]. The following is the nocoinp file. ############################################################3 atom-type 1,l_relax=F alpha = 0.0000000000,b_cons_x = 0.0000000000 beta = 1.5707963268,b_cons_y = 0.0000000000 0.0000000000 atom-type 2,l_relax=F alpha = 0.0000000000,b_cons_x = 0.0000000000 beta = 1.5707963268,b_cons_y = 0.0000000000 0.0000000000 -- logical parameters -- l_ss=T,l_mperp=F,l_constr=F mix_b= 0.500 qss=( 0.0000000000, -0.8469595000, 1.4264960700) #############################################################3
Finally, I will change the qss from (0.0,0,0,0,0) to ( 0.0000000000, -0.8469595000, 1.4264960700).
Is there a problem with the above calculation process ? Besides, without SOC, I wonder whether there is any difference when q vector is along different axis (e.g. x, y, z axes).
Look forward to your kind reply and thank you very much.