Dear Siesta, >The amplitude of variations of the sum of forces (around zero), >periodic with the step of the real-space mesh, >does indeed depend on MeshCutoff and can be gradually suppressed >when increasing the latter. >See tutorials on the Siesta web page on this subject. >Best regards >Andrei Postnikov
What I have found on Siesta tutorials page is a random collection of lecture notes, not very suitable for self studies over internet. It would be great if somebody who knows Siesta and is not under threat of unemployment in the near future (ie a person with a permanent position) could offer his/her time for one month to prepare some decent tutorials for Siesta. I have in mind something in line they have in a classical MM package amber (http://ambermd.org/tutorials/), that is excellent stuff. Some issues one could deal with: - What should be tested and how for converged results (with an example) - How to test a pseudopotential (with an example) - What to do if scf cycles do not converge with solution method diagonal and standard settings in the input file (with an example). - When to use orderN method? To which systems it is good/bad. What can be done in case of poor convergence (with examples). I understand that for a program free of charge one should be thankful for what one gets, but anyway... Terveisin, Markus -- --www=http://www.iki.fi/markus.kaukonen --markus.kauko...@iki.fi --office: N102 Nano building FIN-02015 TKK --home: Viinirinne 3 F 12, 02630 Espoo, FIN --tel: h 045-1242068, w 4518694, 050-5112785 --Rikos ei kannata, eika maatalous --Suomessa. (Paimio 1998) ---
