Dear Stefano, Thank you very much for your explanations.
Maybe I didn't make myself clear enough! I understand the general processures of a QHA calculation, but what really bothers me is that many DFT code with phonon functionality simply calculates F-T relationship at a SPECIFIC volume and made a F-T curve. Of course, there is nothing wrong in doing so if no real situation is considered since F IS a function of T and V (or other structural parameters). But this is in fact incorrect when thermal expansion is concerned in the spirit of QHA. Then in order to make a F-T curve, phonon spectrum has to be found at each equilibrium geometry under the given T instead of the single set of spectra. Hope I have made it clearer this time! Thanks W. YU ____________________________________________________________________________________ Get your own web address. Have a HUGE year through Yahoo! Small Business. http://smallbusiness.yahoo.com/domains/?p=BESTDEAL
