Dear All,<div><br></div><div>I am running a test calculation of bcc iron, to check if the relaxation of angle1 (theta) works correct for the non-collinear case. I have copied the file "<a href="http://fe.pen.in">fe.pen.in</a>" from the examples provided on <a href="http://pwscf.org">pwscf.org</a>, and added one more Fe atom (so, I have two Fe atom unit cell). I set angle1(1)=90; angle1(2)=0.0. After self-consistent calculation I expect these angles to be aligned (ferromagnetically). Yet, I end up getting very strange result, i.e. theta(1)=70.356498; theta(2)=19.684944. Does this mean that the relaxation of atomic magnetic moments is not very reliable in case of non-collinear calculation? This is first time I am trying to include non-collinearity in my calculation, therefore I don't have much to compare with. And I did not find any published work on non-collinear systems done with PWSCF. Any help on this issue is very appreciated. I attach my input file which also includes portion from output showing magnetization magnitude/direction.</div>
<div><br></div><div>Sincerely,</div><div><br></div><div><div>Pavel Lukashev, Ph.D.</div><div>Research Associate</div><div>UNL, Department of Physics & Astronomy</div><div>208 Theodore Jorgensen Hall</div><div>855 N. 16th Street</div>
<div>Lincoln, NE 68588-0299</div><div>Office: (402) 472-2400</div></div>