The ABINIT Users Mailing List ( CLOSED )

[tony lee <tonymidlee@hotmail.com>]

Dear friends, 

I have a problem in calculating the dielectric function. 

the bulk material is not cube(more like two cube on top of each other), and 
the
three primitive translational vectors are not perpendicular to each other 
(same
length though). I used rfdir 4: 100 rfdir 5: 010 and rfdir 6: 001

but the linear part of optical calc. gives me quite different results in Z
direction, compared with X and Y direction, which is not correct because they
should have similar values ( at least not as big difference as my result).

So what do you guys do when you counter such situation? change rfdir in the
input file?(I tried but gives me error message or wrong results) or choose a
super cell (cube)? (but then you can not choose fine k-mesh).

any help is appreciated!

Thanks for your time!

Tony