Dear all, Is the "no_overlap=.false." and "lsym=.false." settings implemented in bands.x, when we use the PAW pseudopotentials? In this settings the code should write the eigenvalues in the order that maximises the overlap with the neighbor k-points. This works perfectly fine with the norm conserving pseudopotentials, but I think, it doesn't work with the PAW potentials. The code exits with a "segmentation fault" with some error in s_psi.f90 and in the "punch_band" subroutine of bands.f90 code.
My question is there any fundamental issue with using the PAW potentials in this settings? Or can it be solved by some simple modification of the code? Any help is highly appreciated. Thank you and best regards, Barun Ghosh Senior Research Fellow Department Of Physics Indian Institute Of Technology, Kanpur
_______________________________________________ users mailing list [email protected] https://lists.quantum-espresso.org/mailman/listinfo/users
