Thank you all for your tips. It turns out the real fix was to upgrade to the latest stable version. Now the program is behaving as I would expect.
Cesar, you mention the parameters wmass and dt...I was wondering if there is a way to get an initial estimate for these values. Unfortunately the documentation on this is sketchy at best. Right now I am running with wmass=0.1, and I didn't set dt, so it must be using the default. Things are converging slowly...the atomic forces are creeping down. Eric >From: cesards at msi.umn.edu >Reply-To: pw_forum at pwscf.org >To: pw_forum at pwscf.org >Subject: [Pw_forum] Re:Re:vc-relax (Eric Abel) >Date: Mon, 11 Sep 2006 20:37:18 -0500 (CDT) > >Hi Eric, > >It is OK, I understand your problem now: Your damped dynamics cicle heve >NOT converged. If dt parameter in &CONTROL, wmass in &CELL namelist and >atomic masses are not set properly damped wc-dynamics takes too long to >converge. If you are not running in damped mode, the keywords tosearch are >"new lattice vectors (alat unit) :" and "new positions in cryst coord". >The program will report those at end of every MD step. > >Also remember to set nstep in &CONTROL namelist to an appropriate number, >depending on the numbers of degrees of freedom, and the fine tunning of >paremeters above. > >If you want to learn how to optimize dt, wmass and atomic masses, take a >look at the several tutorials at > >http://www.vlab.msi.umn.edu/events/tutorials.shtml > > >Yours, > >Cesar R.S. da Silva. > > >_______________________________________________ >Pw_forum mailing list >Pw_forum at pwscf.org >http://www.democritos.it/mailman/listinfo/pw_forum
