[Pw_forum] Problem with constraining an atomic position

[Mahmoud Payami]

Hi All,

I have performed a simple "relax" job for a few-layered Al slab. The input 
file is similar to that in example03 but with an un(?)important difference 
that all z-coordinates are positive, starting from z=0.0 for the first 
layer:

Al  0.00000   0.0000000   0.0000000    0   0    0

I have also constrined this z=0 atom to stay at that position during the 
relaxation. But, at the end all atoms (including the first one) are moved in 
the positive z direction. How can I handle it?
Best regards,
Mahmoud Payami, AEOI, Tehran, Iran