[Wien] More accurate total energy with LSDA+U and lapwso

[XU ZUO]

Dear Colleagues,

I am calculating the magnetic anisotropy, which is usually as small as 1e-7
Ry per unit cell. As the Wien2k code by default is accurate to 1e-6, I need
to increase the accuracy to at least 1e-8. I read the mail list archive and
find how to increase it WITHOUT LSDA+U and spin-orbital (SO) coupling.
However, as the magnetic anisotropy is from spin-orbital coupling and
sensitive to U, it is necessary for my case to increase the accuracy WITH
LSDA+U and SO. Does anybody know how to do it? Thanks a lot.

Xu Zuo
Professor of Electrical Engineering
College of Information Technical Science
Nankai University
94 Weijin Rd., Tianjin 300071, P. R. China
Tel: +86-22-2349-4893
FAX: +86-22-2350-8234
Email: xzuo at nankai.edu.cn