Hi, you provide too few information. Of course just increasing the number of iterations won't bring you further if there is no convergence. General advises might be as follows:
- Check that you are using diagonalisation and NOT order-N scheme before you fix the things and "know what you are doing". With 235 atoms the chances are, you are in order-N regime by default, and things go astray. - Check your system visually (using visualisation tools from your working XV file) whether its shape is as expected; a huge number of problems result from an error in defining the structure. - Check whether your electronic structure makes sense (the main bands are more or less where they are expected) at very first iterations, before the calculations start to markedly diverge. Do not just look at the convergence, but at other things in the output file. Make sure that the attribution of valence states in the pseudopotential and in the basis matches. Some inconsistencies may cry aloud in the output file, but you need to pay attention... - Reduce the mixing parameter. Reduce it drastically if needed (0.001 is not uncommon in some cases). Start converging with large broadening (electronic temperature), say 600 K, and then reduce it gradually as you enter into a convergence regime. The actual behaviour may depend on whether your system is isolated or periodic, insulator or metal. If your case is periodic and metallic, you may need many k-points as well (in the directions of periodicity, of course). Good luck Andrei Postnikov ----- Le 19 Avr 24, à 4:16, Jyotirmoy Deb <deb.jyotirmo...@gmail.com> a écrit : > Dear Sir/Madam, > I want to optimize a heterostructure consisting of 235 atoms. But I am facing > an > error every time "SCF did not converge in the maximum number of steps". I have > gradually increased the "MaxSCFIterations" however, I have not been able to > solve this issue. Please help me in this regard. > Thanking you > with regards > Jyotirmoy > Dr. Jyotirmoy Deb > C/O Dr. G. Narahari Sastry & Dr. Lakshi Saikia > DST-SERB National Post-Doctoral Fellow > Advanced Computation and Data Sciences Division (ACDSD) > CSIR-North East Institute of Science & Technology, Jorhat-785006, Assam, > India. > Ph. No: +917002140643/ +919435589869 > Email: [ mailto:deb.jyotirmo...@gmail.com | deb.jyotirmo...@gmail.com ] > Webpage: [ > https://urldefense.com/v3/__https://sites.google.com/view/jyotirmoydftphy__;!!D9dNQwwGXtA!QXMXS7L8cCba3hBh8KgcbwyC9C-E5X98KJTAXUguv0CMpKRil5P5cMWK5ziJhwU4-c14ht8eejcbtj_czqW_iUBo$ > | > https://urldefense.com/v3/__https://sites.google.com/view/jyotirmoydftphy__;!!D9dNQwwGXtA!TlNGLjwuIHLvaLUEXMAzZneDEOzZtFJFMRfYAnBQADgcqVNh1tJuG4DMBR8sWhbOVnrRtR-96vIssOhkbvkqvQKGf5LpgmFIQA$ > ] > -- > SIESTA is supported by the Spanish Research Agency (AEI) and by the European > H2020 MaX Centre of Excellence > (https://urldefense.com/v3/__http://www.max-centre.eu/__;!!D9dNQwwGXtA!TlNGLjwuIHLvaLUEXMAzZneDEOzZtFJFMRfYAnBQADgcqVNh1tJuG4DMBR8sWhbOVnrRtR-96vIssOhkbvkqvQKGf5IASVdD0A$ > )
-- SIESTA is supported by the Spanish Research Agency (AEI) and by the European H2020 MaX Centre of Excellence (http://www.max-centre.eu/)
