Category:Electronic minimization: Difference between revisions

The algorithms VASP offers for electronic optimisation (i.e., determining the electronic ground state) can be divided in two categories:

• Iterative matrix diagonalisation + density mixing, aka the "Self-Consistency Cycle" (SCC).
• Direct optimisation of the orbitals.

The Self-Consistency Cycle

1. The SCC starts with an initial guess for the electronic density of the system under consideration.

Direct optimisation

• Electronic minimization: Algorithms used in VASP to calculate the electronic groundstate.
  • Preconditioning of residuals: Preconditioning.
  • Simple Davidson iteration scheme: Davidson iteration scheme.
  • Single band steepest descent scheme: Single band steepest descent scheme.
  • Efficient single band eigenvalue-minimization: Efficient single band eigenvalue-minimization.
  • Conjugate gradient optimization: Conjugate gradient optimization.
  • Residual minimization scheme, direct inversion in the iterative subspace (RMM-DIIS): RMM-DIIS.
• Wrap-around errors in FFT meshes - convolutions: Wrap-around errors.
• Integration of k points: k-point integration.
• Theory of density mixing

How to

• A description to obtain band decomposed charge densities is given here: Band decomposed charge densities.
• k-point projection scheme: LKPROJ.