<table cellspacing="0" cellpadding="0" border="0" ><tr><td valign="top" style="font: inherit;"><pre>Dear Prof. <b>Pavel Novak,<br> <span style="font-family: monospace;">So you mean that to have the new structure files<br>using symmetso for 001 and 100, and use the k list generated by one of them<br>in SO calculations for both 001 and 100. In this case one keeps the basis set <br>fixed and just examine how the magnetization direction affects the<br>total energy by running one cycle: runsp_lapw -i 1 -so.<br><br> </span></b><br>Regards<br>Bothina<br><br><br><br><br><br>Dear Bothina Hamad,<br><br>calculation of MAE in 3d metals is tricky and I recommend using the same <br>set of k-points to calculate it. Run symmetso for a general direction <br>(a,a,b) and you can then use the set klist both (001) and (110). If the <br>MAE is small the 'force' theorem make the calculation easier (you converge <br>calculation without s-o, then run
lapwso, lapw2up/dn for two M directions <br>and take the difference of eigenvalue sums). Nevertheless, always check <br>the dependence on the number of k-points.<br><br>Regards Pavel<br><br></pre><br><br></td></tr></table><br>