NSW is the maximum number of ionic steps. When you set it to 0 it means that no structural optimization would happen. Besides, for slabs it's better to freeze bottom atomic layers with "Selective Dynamics" in POSCAR file.

it can use selective dynamics to freeze bottom atomic layers

In relaxation, you will get slightly different energy. but the difference should be very low. One reason could be initial conditions, in CONTCAR file VASP also gives initial velocities, so you have to start with same initial velocities to get the exact same energy value. But, you can a consistent energy for relaxation and then do the scf convergence to use the energy values.

I hope it helps.

Khagehs

Thank you, for your valuable guidance. Khagesh Tanwar Hadi Jabbar Alagealy Ehsan Moradpur-Tari