exciting-0.9.217

exciting-0.9.217
-fixed problem with forces for spin-polarised systems
-forces now work with LDA+U
-electron-phonon coupling now available (experimental)
-calculation of phonon line widths now available (experimental)
-calculation of superconducting TC using the McMillan-Allen-Dynes equation now available (experimental)
-changed phonon calculation DYN_... files: each computer now calculates a separate row in the dynamical matrix
-phonon calculations slightly faster (you also need to run a ground state calculation before the phonon run)
-thermodynamic properties calculated from phonon DOS now available
-included phonon example for silicon (with results)
-S. Sharma updated RDMFT for spin-polarised and finite temperature systems
-calculation of 〈 i,k+G+q | exp[i(G+q)⋅r] | j,k 〉 matrix elements now available, use task=130 (experimental)
-energy loss near edge structure (ELNES) now available, see BN-ELNES example
-changed linear optics calculation from an integration over the Brillouin zone to a direct sum over k-points
-optical relaxation time is now determined by 1/swidth
-fixed problem with MOKE as well as the off-diagonal elements of the dielectric tensor and conductance
-sped up and simplified the diagonalisation of the spin density matrix (Kübler trick) thanks to Lars Nordström
-F. Cricchio, F. Bultmark and L. Nordström fixed minor problem with LDA+U energy
-fixed problem with non-convergence of atomic code (thanks to Stephan Sagmeister)
-conjugate gradient solver for the first-variational problem is now available: use tseqit=.true. (see example H2O-iterative)
-conjugate gradient solver allows for efficient fine-grained parallism with OpenMP (scales linearly with number of processors)
-spherical covering is now more accurate
-code now runs with zero atoms for studying the electron gas (see examples HEG-HF and HEG-OEP for the Hartree-Fock and optimised effective potential homogeneous electron gas)
-Poisson solver is now faster
-nucleus can now have finite size (set ptnucl=.false.)
-energy printed out to more significant figures
-added MgB2 Fermi surface example
-changed variable rlambda to radkpt
-replaced cpu_time with system_clock calls
-lots of optimisations and simplifications