In general, a well-posed physical problem converges rapidly; a poorly posed one badly if at all. Most probably your system wants to be spin-polarized, or you have too few k-points, bad RMT's etc. For certain if you have large oscillations for 0.02 you have constructed the problem incorrectly -- unless you are doing a single atom which is different.
On Thu, Jun 26, 2008 at 4:07 PM, Jian-Xin Zhu <jxzhu at lanl.gov> wrote: > Dear Wien users, > > I am running LDA+U calculations for a compound > by executing > > runsp_c_lapw -orb -i 40 -cc 0.0001 > > and find it hard to converge. > > Some of output is pasted below. > > in cycle 38 ETEST: .0029925000000000 CTEST: .0092126 > LAPW0 END > ORB END > ORB END > LAPW1 END > LAPW2 END > LAPW2 END > LAPWDM END > LAPWDM END > CORE END > CORE END > MIXER END > in cycle 39 ETEST: .0129470000000000 CTEST: .0161093 > LAPW0 END > ORB END > ORB END > .... > > The CTEST is oscillating. How can I work around to have it converge? > Though likely less important, I have tuned down the mixing factor to > 0.02 in the MSEC1 mode. > In the case.inorb, I still kept the template setting > PRATT 1.0 BROYD/PRATT, mixing > > > Thanks for the help. > > > -- > ################################ > Jian-Xin Zhu, Ph.D > Theorertical Division, MS B262 > Los Alamos National Laboratory > Los Alamos, NM 87545 > Phone: (505) 667 2363 > Fax: (505) 665 4063 > Emai: jxzhu at lanl.gov > URL: http://theory.lanl.gov > ################################ > > > > _______________________________________________ > Wien mailing list > Wien at zeus.theochem.tuwien.ac.at > http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien > -- Laurence Marks Department of Materials Science and Engineering MSE Rm 2036 Cook Hall 2220 N Campus Drive Northwestern University Evanston, IL 60208, USA Tel: (847) 491-3996 Fax: (847) 491-7820 email: L-marks at northwestern dot edu Web: www.numis.northwestern.edu Commission on Electron Diffraction of IUCR www.numis.northwestern.edu/IUCR_CED