Hi Lydia, For phonon calculations, you can use the program fcbuild to generate your supercell. This will also set up a skeleton input file. However, you will need to transfer over additional information from your self-consistent run to insure that the force constant calculation uses the right basis set, k-point mesh, etc.
You can specify the range of atoms that you want to run the force constant calculation on. This is done using by defining MD.FCfirst as the index of the first atom you want to shift, typically the first atom in the central unit cell. and MDFClast as the index of the last atom in the unit cell. For all of these MD.TypeOfRun = FC Since the crystal is periodic, we can get all the information we need by shifting just the atoms in the central unit cell. You can also split these calculations up and do FC constant runs on smaller subsets of atoms in the central unit cell on different processors. When you combine the different FC files, you need to put them in the same order as the atom indices. You also need to remove all the header lines from the combined FC files, except for the first one. Since this is a real space calculation, your phonon dispersion will only be as good as your supercell size! The bigger your supercell, the better you are able to sample your long wavelength phonons. For example, if you run a 3x3x3 supercell (54 atoms for Si), the phonon dispersion has problems along the Gamma-X line. However, if you move up to a 5x5x5 supercell (250 atoms), the phonon dispersion matches well with experiment. The best way to check this is to run a series of calculations where you continue to increase the supercell size until the phonon dispersion converges. The vibra code also does come with a small tutorial discussing the use of the code. Best regards, Derek > Hello! > > I am about to calculate the phonon spectrum with siesta and have several > questions: > > Does anyone knows a publication wher is described what SIESTA and VIBRA > calculates and more important how it does the Calculations? > > I have 6 Atoms in the unitcell and 3x3x3xunitcell=162 atoms in the > supercell. > So far I couldn't finish a FC calculation and there seems no reasonable > restart function. I don't know how many atoms will siesta move? > > It seems to me that SIESTA completly forgets about the symmetries of the > supercell, on the other hand it starts with an atom of the Cell in the > middle > of the block. Can it be that I just need to to Calculate the six atoms of > the > middle unitcell, one after another and assemble the resulting FC-Files > into > one FC-File and Vibra will do the rest for me? > > Can I use order N solution method with Phonon (FC and Vibra) calculation? > I think the orderN algorythm just calculates the gamma point, if it is so, > wouldn't it be useless for an Phononcalculation as I will need a good > k-grid > for reasonable results? > > > Thanks in advanced, > > Best, > > Lydia Nemec > > > -- > Lydia Nemec > ======================================================= > Sector of Biological and Soft Systems > Department of Physics, University of Cambridge > Cavendish Laboratory > J J Thomson Avenue, Cambridge CB3 0HE, UK > Tel. 0044(01223)3-37357 > Email: ln...@cam.ac.uk > ------------------------------------------------------- > Institute for Theoretical Physics > University Regensburg > 93040 Regensburg, Germany > ======================================================= > -- Derek Stewart, Ph. D. Scientific Computation Associate http://www.people.cornell.edu/pages/das248/ 250 Duffield Hall Cornell Nanoscale Facility (CNF) Ithaca, NY 14853 (607) 255-2856
