Constrained Moment Calculations
Posted: Wed Aug 03, 2005 4:35 am
The section of the Guide titled "Constraining the direction of magnetic moments" is not very clear to me.
What is the "penalty" function that is added to the energy?
Is it lambda * (M_xm_x+M_ym_y+M_zm_z) where M is the moment on an atom (within RWIGS) and m is the unit vector set in INCAR for that atom?
What is E_p? Should I add E_p to the total energy (e.g. in OSZICAR) to get the total energy in the constraining field?
How would I use the constrained field, for example, to estimate the nearest neighbor exchange interaction, e.g. in bcc Fe.
lambda*MW_perp is said to be the "constraining magnetic field". The subscript, however, suggests it may be a perpendicular (to what?) component.
Why should E_p vanish for large lambda? Actually it initially increases with lambda when lambda is small.
What is the "penalty" function that is added to the energy?
Is it lambda * (M_xm_x+M_ym_y+M_zm_z) where M is the moment on an atom (within RWIGS) and m is the unit vector set in INCAR for that atom?
What is E_p? Should I add E_p to the total energy (e.g. in OSZICAR) to get the total energy in the constraining field?
How would I use the constrained field, for example, to estimate the nearest neighbor exchange interaction, e.g. in bcc Fe.
lambda*MW_perp is said to be the "constraining magnetic field". The subscript, however, suggests it may be a perpendicular (to what?) component.
Why should E_p vanish for large lambda? Actually it initially increases with lambda when lambda is small.